THOMAS CHARLES ROBINSON, BALANCE-MAKER Barry J. Oliver
THOMAS CHARLES ROBINSON, BALANCE-MAKER
ii THE ‘ROBINSON TULIP’
iii THOMAS CHARLES ROBINSON, BALANCE-MAKER by Barry J. Oliver, C.Chem., FRSC μg MICROGRAMME LEICESTER 2026
iv First published in Great Britain by Microgramme, Leicester, 2026. © Barry J. Oliver 2026. Database right Barry J. Oliver 2026. The author’s moral right to be identified as author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior permission in writing from the publisher, or as expressly permitted by law, or under terms agreed with the appropriate reprographic rights organisation. A catalogue record for this book is available from the British Library. ISBN not needed as there will be no bookshop sales. Printed in Georgia 12-point on acid-free FSC certified paper. Printed and bound in Great Britain by FinaPrint Ltd, Walsall.
v This book is dedicated to the memory of the late PETA DEWAR BUCHANAN, 8th May 1936 ― 25th March 2011, without whose pioneering work on chemical balances in general and on Robinson and Oertling in particular, this book would never have been begun.
vi
vii PREFACE This book on Thomas Charles Robinson (1792―1841) has arisen as a by- product of my extensive researches into Britain’s foremost maker of high-precision balances, Ludwig Oertling. That work resulted in the publication in 2020 of an extensive monograph on him, his company and its products.1 In part it used material from the late Dr Peta Buchanan’s thesis on the development of the chemical balance. In her thesis she devoted a long chapter to both Robinson and Oertling as being the two manufacturers dominating British balance-making in the nineteenth century.2 From this point onwards our subject will usually be referred to as T.C. Robinson or just plain Robinson. It must be understood from the outset that this book is not only a biography of T. C. Robinson, in so far as that is possible considering the scant quantity of resource material available, but it may also be read as a companion volume to the Oertling monograph. This is because Robinson was one of the three great influences on the young Ludwig Oertling, the others being his elder brother August Oertling in Berlin, and the London instrument-maker Robert Brettell Bate in whose workshop Ludwig became a journeyman in 1842. The history of R.B. Bate (as he is usually known), has already been covered by the late Anita McConnell.3,4 August Oertling was given thorough coverage in the Oertling monograph. This third book therefore fills in the gap, but concerning Robinson only as a maker of balances. Irrespective of their other instrument-making activities, the triumvirate of Robinson, Bate and Oertling exerted the greatest influence on British high-precision balance design for the rest of the nineteenth century and beyond: Robinson on laboratory and reference balances, Bate on bullion balances and Oertling on laboratory, reference and bullion balances. Naturally other great British instrument-makers existed and created balances, men such as Fidler, Field, T. Jones, W&S Jones, Nairne, Newman, Ramsden, Sisson, Troughton, and others, but the few balances they made were almost always specials, not routine production items and certainly not as a significant part of their outputs. Many readers will be familiar with T.C. Robinson as a maker of fine mathematical, optical and philosophical instruments. He was particularly strong in astronomical telescopes, microscopes and surveying instruments; he also created some devices for research into magnetism. His balances occupy only a very few lines in his catalogues, because he made only one basic production model in three different sizes, and a very 1 Oliver (2020). 2 Buchanan (1982), Chapter 5. Robinson is covered in pp.208-216, plus the references on pp.225-226. 3 McConnell (1993). 4 McConnell (2004).
viii few bespoke balances and early scales. As a consequence, Robinson’s balances and their importance in the development of high-precision weighing in England have been largely overlooked, indeed ignored until recently, despite there being a clear progression of design within them. I hope to rectify this oversight. This book will be concerned solely with his balances and their influence on later balance-makers. His better-known optical and magnetic instruments will not be covered here. T. C. Robinson’s productive years as a London maker of scientific instruments are usually quoted as 1821―1841. He was one of the first English makers to produce a wide range of scientific instruments on a commercial basis, by which is meant they were standard designs constructed according to a printed and priced catalogue. Benjamin Martin was one of the first to do this, as early as 1757, followed by W&S Jones from 1814 onwards, quite a few years before Robinson’s earliest known analogous catalogue of 1828. Finally, the presence of the Oertling Addenda and Errata pages here in the appendices of this book needs to be justified. Robinson died in London in 1841, just one year before Ludwig arrived here. Robinson had a great influence on Ludwig Oertling’s design philosophy, by either or both of two different routes: 1) Ludwig saw them in brother August’s workshop in Berlin whilst he was being trained there, either as imports by August for sale in Europe, or as existing old ones coming into the workshop for servicing or repair; 2) or more probably, he saw Robinson balances going into the Bate workshop for servicing or repair. There would have been a fair number of Robinson’s balances on active service during Ludwig’s early years in London. Furthermore, a few fresh balances in the Robinson style were being constructed by other makers for some years after Robinson’s death, as will be discussed later. So, because of Robinson’s influence on Ludwig, he was considered at some length in the Oertling monograph. Therefore, as said earlier, this book may be seen as a companion volume to the earlier one, by taking the examination of Robinson’s balances very much further. Although well over half of the original print run of 200 of the Oertling monograph have now been sold or otherwise placed, there is clearly not a market for a second edition. However, numerous new findings have occurred since 2020, which need to be inserted into the Tables, and one major error and a few minor errors have been found in the original text. These revisions were printed on an errata sheet tucked into most of the books sold to date, but a few more minor errors have been found. As it is hoped that the two books will be read together, it seems only prudent to place the addenda and errata in this present volume, rather than let them go unrecorded and lost for ever. The contents of the original list have been incorporated in this new listing, so there is only one list to search here. Having said all of the above, this assessment of Robinson may perfectly well be read entirely alone without reference to the Oertling volume. I hope you find it a pleasant, informative and helpful read. Barry J. Oliver, LEICESTER, DECEMBER 2025.
ix CONTENTS Preface. vii List of Illustrations. x Note to the Reader. xiv Chapter 1 The Story so Far. 1 Chapter 2 T.C. Robinson, The Man. 5 Chapter 3 T.C. Robinson, The Business. 9 Chapter 4 English Fine Balance Design circa 1800―1820. 17 Chapter 5 The Typical Robinson Balance. 55 Chapter 6 The Origin of Robinson’s Balance Designs. 65 Chapter 7 The ‘Swan-Neck’ / ‘Early’ Robinson Scales. 71 Chapter 8 The ‘Wollaston’. 81 Chapter 9 The ‘Miller’. 85 Chapter 10 The ‘German’ Robinsons. 89 Chapter 11 The ‘Used by Miller’. 93 Chapter 12 The ‘Cambridge University’. 97 Chapter 13 The ‘Rothamsted’. 99 Chapter 14 The ‘Ex-Edinburgh University’. 103 Chapter 15 ‘Dr Moll’s Balance’ alias The ‘Robinson―Olland’. 107 Chapter 16 The Robinson & Barrow ‘Miller 1843―1845’. 121 Chapter 17 The Robinson & Barrow ‘ex-Stokes Collection’. 123 Chapter 18 The Robinson & Barrow ‘ex-British Museum’. 127 Chapter 19 The ‘J.B. Dancer’ Robinson, Query. 133 Chapter 20 The ‘Adie’ Robinson Copy. 135 Chapter 21 The Robinson Precursors, Query. 139 Chapter 22 The Styling of Robinson’s Balance Cabinets. 149 Chapter 23 Robinson’s Weight Sets. 159 Chapter 24 Robinson Revived in the 20th Century. 165 Chapter 25 Robinson’s Balance Designs, a Final Assessment. 177 Chapter 26 Valediction. 183 Appendices 185 1) Thomas Charles Robinson’s Will, Transcribed. 186 2) Thomas Charles Robinson’s Will, the Original. 187 3) Hydrostatic Weighing. 189 4) The Form of Knife Edges. 195 5) Philosophical and Scientific Instrument-makers. 196 6) British Pre-Decimal Currency. 199 7) A Pseudo-Robinson Scale. 201 8) Addenda and Errata for the Oertling Book. 205 Acknowledgements 213 Contacts 214 Bibliography 215 Index 221
x LIST OF ILLUSTRATIONS FRONTISPIECE: The Robinson Tulip, the most obvious characteristic structure in a T.C. Robinson balance. Photo by courtesy of Rothamsted Research / Lawes Agricultural Trust. FRONT COVER: A typical Robinson balance. REAR COVER: A Robinson early scale with a steel beam. IN TEXT: 1 Map: 37 and 38 Devonshire Street. 7 2 Map: Devonshire Street and Portland Place. 8 3 The Robinson catalogue of 1828, first page only. 10 4 The Robinson 1829 catalogue, first page. 10 5 The Robinson 1829 catalogue, second page. 11 6 The Robinson 1842 posthumous catalogue. 13 7 The Robinson & Barrow catalogue of 1842. 14 8 A rare, early photograph of the Ramsden on its long legs and board. 20 9 Benjamin Martin coin balance. 21 10 Benjamin Martin coin balance, beam end. 21 11 Whitehurst balance V22456, Boerhaave Museum, Leiden; general. 23 12 Whitehurst balance V22456, Boerhaave Museum, Leiden; details. 23 13 The Sisson Balance Sci. Mus. 1927-2040. 24 14 Ditto, with door raised, plus handle. 24 15 The Magellan balance of 1781. 26 16 Troughton 18″ cone-beam balance and assay balance. 29 17 Troughton cone-beam balance recorded by Benton. 31 18 The Avery Museum Troughton 01.02.17. 32 19 The Troughton 01.02.17 beam. 32 20 The ‘Railway Arch’, beam, CG bob and flag. 33 21 The ‘Railway Arch’ and fixed arms. 33 22 W&S Jones cone beam balance. 34 23 John Field assay balance. (Doorslide removed for clarity.) 36 24 Close-up, on release. (Note also the platinum chains.) 37 25 Arrested. (The Field engraved inscription is losing its black filling.) 37 26 The John Field balance of 1826. 39 27 The oval cartouche on the Field balance. 39 28 The Rothamsted Bate. 46 29 The stand and cut-out. 46 30 The Tralles beam of 1808. 52 31 HSM 17549 front view. 54
xi 32 Rothamsted BA104 without its lantern. 57 33 Rothamsted BA104 with its lantern in place. 57 34 The ‘Robinson Tulip’, chart and pointer. 58 35 The Robinson production beam. 58 36 The Robinson production beam and arrestment frame. 59 37 The Robinson production beam end. 59 38 Released. 60 39 Arrested. 60 40 End-on view of the arrested beam end. 60 41 The cradles on the arrestment frame, and their adjustments. 61 42 The finger bolts and clear agate centre plane. 61 43 Arrestment lever and ball & socket. 62 44 Arrestment lever in use. 62 45 The drawer contents. 64 46 Children’s sketch of the original Kater-Robinson scale. 67 47 Kater’s original sketch of the ‘Wollaston Pan Stabilisers’. 68 48 The ‘Wellcome’. 70 49 The ‘Steel Scale’ Beam and Pointer. 72 50 The ‘Steel Scale’ complete. 72 51 The ‘Steel Scale’ minus lantern, front view. 73 52 The ‘Steel Scale’ minus lantern, overhead view. 73 53 The ‘Steel Scale’ stand base. 74 54 The ‘Steel Scale’ chart. 74 55 The Specific Gravity pan. 75 56 The interior of the ‘Steel Scale’ drawer. 77 57 The ‘Wellcome’. 78 58 The ‘Wellcome’ beam and Tulips. 79 59 The ‘Wellcome’ beam and the pairs of Tulips. 79 60 The interior of the ‘Wellcome’ drawer. 80 61 The ‘Wollaston’ minus its lantern. 81 62 ‘Wollaston’ end plane, underside. 82 63 Ditto, side view. 82 64 The ‘Wollaston’ with its lantern. 83 65 ‘Miller’ J-hook. 85 66 The ‘Miller’. 86 67 The ‘Miller’, with an SG pan and a normal pan suspended. 87 68 Hass 1, front view. 89 69 Hass 1 overhead view. 90 70 Hass 1 minus its lantern. 90 71 Hass 1 drawer. 91 72 Hass 1 stand base. 91 73 Hass 2 stand base. 91 74 Hass 2. 92 75 Hass 2 with lantern removed. 92 76 ‘Used by Miller’ 10½″ Robinson. 93 77 An engraving of unknown date of a Robinson 10½″ balance. 95 78 Rothamsted Robinson 8″ partially dismantled; general view from rear. 99 79 Ditto, 8″ beam and pointer, pans, cords and J-hooks. 101 80 The Rothamsted Robinson 8″ beam end, top view. 101 81 The Rothamsted Robinson 8″ beam end, end view. 102 82 The ex-Edinburgh University Robinson. 103
xii 83 The ex-Edinburgh University Robinson drawer. 104 84 Two parts of the weight set. 105 85 Dr Moll’s rebuilt Robinson balance in its new cabinet (case). 111 86 The Robinson beam. 112 87 The Robinson beam end. 113 88 The Robinson beam end. 113 89 The Robinson beam end, side view. 113 90 The ‘waisted’ centre knife. 114 91 The ‘waisted’ centre knife. 114 92 The Olland pan, hanger and suspension piece. 114 93 The suspension piece plus extension. 115 94 Olland pointer plus mounting plate. 115 95 The beam and new pointer in position. 115 96 The Olland chart. 116 97 The Robinson engraved baseplate. 116 98 The Robinson and extra baseplate. 117 99 The rack and pinion. 117 100 Arrestment frame and push rod, arrested. 117 101 Ditto, released. 117 102 The microscratches. 118 103 The ‘ex-Stokes’ Robinson & Barrow 10½″ with lantern in place. 123 104 The ‘ex-Stokes’ in demonstration circa 1976. 124 105 The stiffening screws from below. 125 106 The Robinson & Barrow 10½″ general front view, complete. 128 107 The Robinson & Barrow 10½″ minus beam and pans. 128 108 The Robinson & Barrow 10½″… to show parts more clearly. 129 109 The chart. 130 110 The ‘Dancer’ Robinson. 133 111 The ‘Adie’. 136 112 The Adie drawer contents and trade card. 136 113 An Adie beam end. 137 114 The Miller and Adie high capacity balance with a steel rhomboid beam. 138 115 MHSO 157 (HSM 18579) with doorslide in place. 141 116 MHSO 157 (HSM 18579) with doorslide removed. 141 117 Sci. Mus. 1927-1138, general view. 143 118 Sci. Mus. 1927-1138, minus doorslide. 144 119 Sci. Mus. 1927-1138, beam and pointer. 144 120 Wh.1446 without doorslide. 146 121 Wh.1446 with doorslide. 146 122 The centre self-locating knife and ring, and pin S3. 147 123 The lever in a ball joint in front of the stand. 147 124 The gap beneath the doorslide which almost touches the lever. 148 125 The ‘Wollaston’ and its lantern. 151 126 ‘Hass 1 with all doors open to show their details. 151 127 A typical Robinson case for a 10½″ balance. 152 128 An 18th century chemical lab with two balances on the windowsill. 153 129 Oertling model 2* in a Robinson-style case plus front doorslide. 155 130 The Duffey American copy of a Robinson 10½″ balance. 156 131 Avery 01.02.17 left front drawer. 157 132 The weight set belonging to BA104. 160 133 The Rothamsted BA104 drawer weights. 161
xiii 134 The BA104 ‘fractional’ weights box and lid. 161 135 The usual five-recess fractions box. 161 136 The Avery 01.02.17 weight set. 163 137 The Webb White Monarch. 166 138 The White Monarch beam, front, with pointer and flag in position. 166 139 The White Monarch beam, rear, with suspension pieces in position. 166 140 The formal specification for the White Monarch. 167 141 The two agate centre bearings. 167 142 A few White Monarch parts. 167 143 Upper: The beam of an original Webb White Monarch. 168 144 Lower: No usable image of a Webb slit-end beam, so a J10 is shown. 168 145 The final product (1): the Oertling J10 laboratory scale. 170 146 OJ11 beam and pointer, push rod and frame. 171 147 OJ10 oil damper pot, pointer and vane. 171 148 A J10 Robinson slit. 171 149 The final product (2): the Oertling OJ11 pharmacy dispensing scale. 172 150 The OJ11 beam. 172 151 Part view of the Queensland Webb White Monarch. 173 152 The mysterious serial number. 174 153 An Oertling J10 in its plywood case. 175 154 Horseshoe weights from Buchanan (2002). 192 155 A set of variations of the Mohr principle; Hartner catalogue, c.1912. 193 156 The Westphal SG balance as imported by Oertling; Jackson cat. 1908. 194 157 A Pseudo-Robinson scale. 201 158 The stand, released. 202 159 The stand, arrested. 202 160 The beam box-end and cradle. 202 161 Michael Crawforth’s sketches. 203 162 The drawer with contents in place. 204 163 Map of Cowcross St. in 1823. 208 164 Map of Cowcross and Turnmill Street, early 1870s. 209 165 Turnmill Works in the Nucard period, 1962. 210 166 Map of Cowcross and Turnmill Streets at the start of the 21st century. 211 167 Photo of Cowcross Street and Farringdon Station, 2007. 211
xiv NOTE TO THE READER Abbreviations used throughout the book: # is serial number, as is s/n. BM is British Museum. CG is centre of gravity. DGS is De Grave Short Ltd. EQM is Equilibrium, the journal of ISASC. ISASC is the International Society of Antique Scale Collectors. k-k is the beam length measured from end knife to end knife. ‘L’ refers to the left and ‘R’ to the right, balance pan, knife, arm etc. Lab. The abbreviated term ‘lab’ is used instead of ‘laboratory’ quite often because this is how it is generally referred to these days as the full word appears to be falling into disuse, and it also saves space. It is interesting to note that the supposedly full word is itself a contraction of the 17th to 19th century term ‘elaboratory.’ MD is Managing Director. NPL is National Physical Laboratory. ‘R’ refers to the right and ‘L’ to the left, balance pan, knife, arm etc. RGO is Royal Greenwich Observatory. Sci. Mus. is the Science Museum, London. s/n is serial number, as is #. WAW Ltd is William A. Webb Ltd. WWI is World War I, the Kaiser’s War, 1914―1918. WWII is World War II, Hitler’s War, 1939―1945. A few names occurring frequently throughout have at times been reduced to initials: TCR Thomas Charles Robinson. HB Henry Barrow. R&B Robinson & Barrow. PDB Dr Peta Buchanan. O. Ludwig Oertling, or L. Oertling Ltd., not August Oertling. JADO August Oertling of Berlin.
CHAPTER 1 The Story So Far HE FIRST SCHOLAR to realise that T.C. Robinson was an important and pioneering maker of high-precision balances was the late Professor John T. Stock (1911―2005).1 Stock was Professor of Chemistry in the College of Liberal Arts and Sciences, University of Connecticut. Circa 1960 he read in Child’s well-known book The Tools of the Chemist that the first high-precision balances used in the USA were by the London maker Robinson.2,3 He was intrigued by this and yet more so after having failed to find even a single Robinson balance in the USA. He later found that not even the Science Museum in London had one at that date, but there were two here in the UK at the Whipple Museum in Cambridge, plus one belonging to Edinburgh University. There was also an incomplete and damaged one in Cambridge University. To date it seems that there is still no known surviving Robinson balance in the USA. To anticipate the story slightly, the author has found that at least a few Robinson balances were indeed imported into the USA and were widely copied there; this is taken much further in Chapter 25. The late Peta Buchanan (1936―2011) followed Stock and delved more deeply into the life and work of T.C. Robinson in the course of her ground-breaking investigation of the evolution of the chemical balance.4 At the same time she also investigated the immigrant Ludwig Oertling who, from 1846 onwards, effectively filled the gap in the commercial London high-precision balance market created by the death of Robinson a few years earlier. The author joined forces with Buchanan for some years to extend her work, later writing a major monograph on Oertling after Buchanan’s decease.5 In the course of these researches the author realised that Robinson was even more important and influential in the development of high-precision balances than hitherto supposed, but this aspect of his work has been totally ignored by academics and collectors alike. To date Robinson is known as a noted maker of many other types of scientific instruments, mainly optical and magnetic (telescopes, sextants, theodolites, dip circles etc.). He was among the first to offer a wide range of instru- ments from a printed, priced catalogue, in other words, on a fully commercial basis. Professor Stock’s understanding of Robinson’s balances was limited by the fact that initially he had only three complete ones to examine, but later discovered the makers Adie, Dover and Barrow (the last trading as ‘Robinson & Barrow’ and later as ‘Henry 1 Stock (1968); Stock (1969); Stock (1971); Stock (1973); Stock (1978); Stock (1986). Obit Bull. SIS 87 (2005), 38. 2 Stock (1986), 239, 244-245. 3 Child (1940), 81. 4 Buchanan (1982), 208-216; 225-226. 5 Oliver (2020). T
2 Barrow’) continuing production of Robinson’s designs. He made important observations in his first two papers.6 He essentially created and defined the subject, putting this previously forgotten and ignored aspect of Robinson on the map of English quality balance-making. Buchanan was both persistent and lucky, and discovered more Robinson balances to bring the known total in 1982 to eleven. Four more recent discoveries, plus the inclusion of several possible predecessors and copies, brings the current total to be considered in this book to twenty. To date nobody in the 21st century has considered the Robinson balances as a group. Now that we have well over a dozen to work on, it should be possible to make further deductions and, above all, to understand the progression in the designs, which is considerable. However, on a casual glance the final models bear only a passing resemblance to the earliest, the most obvious common factor being the Robinson Tulip, shown in the frontispiece. The Tulip styling was never taken up by later makers, but many of Robinson’s other innovative designs were highly influential upon later makers, as the author hopes to demonstrate. The clear progressive advancement in detail design is really the only way of putting most of these survivors into a provisional date order. It is unfortunate that not even one Robinson balance has a definite date assigned to it in any way. Neither Ludwig Oertling nor Robinson nor R.B. Bate engraved serial numbers on their balances, although one Robinson does bear what appears to be a true serial number, but no production records exist. The working lifetime of R.B. Bate (1782―1847) overlapped that of T.C. Robinson, meaning they were rivals at the top end of scientific instrument-making in London for some twenty years. Both made balances, but because of his experience in creating the well-known Kater Balance alias Bushel Balance, Bate concentrated on bullion balances with massive wooden beams rather than on fine balances for laboratory use such as those created by Robinson.7 Four of the five known surviving Bate balances are bullion balances; the other is his sole surviving chemical balance. That balance, which was never completed by Bate, has a stand dominated not by a Robinson Tulip, but by a cruder cut-out in the stand, in an old English tradition. The rest of it is mostly by Ludwig Oertling completing the balance for the customer; this was when Ludwig was a journeyman at the Bate workshop 1842―1845. This odd instrument is featured in Chapter 4 within a review of the English fine balance-making scene by makers of the period just before Robinson. This review is solely to establish the state-of-the-art when Robinson entered this field and changed it forever. There are only two mentions of Robinson in McConnell’s biography of Bate, both in connection with the prototyping of the Imperial Standards in 1824.8 Her biography contains very little on the normal Bate products, as her assignment was something of a rush job, a fee-earning contract which was to be completed within one year.9 Whether the late McConnell’s ‘box of Baterie’ (her term) held any further facts on Robinson, or on Bate production records (if any), we shall now never know. What was clear, she said, was that she had not seen any evidence of Bate serially numbering any of his instruments. A few words are necessary on the nomenclature of balances, both current and that of the Robinson period. The terms ‘balance’ and ‘scale’ are frequently confused, even by auction houses which ought to know better, and even by some academics in their 6 Stock (1968); Stock (1971). 7 Oliver (2020), 11-18. 8 McConnell (1993), 19, 21. 9 Personal communication with the author, November 2008.
3 writings. ‘Scale’ is short for ‘beamscale’, meaning an equal-armed two pan weighing device, the traditional item used in the past to weigh at a crude level in shops and kitchens. Here the bearing surfaces, both at the centre and the beam ends, are in constant contact. A balance is defined as ‘a fully relieved beamscale’ which means all three knives and planes are lifted clear of each other when not in use. The balance in its relieved condition is said to be ‘arrested’ and when working it is ‘on release.’ Also confusing is the use of the term ‘a pair of scales’ which literally means two pans, but it suggests the crude scale that one’s grandmother used for weighing in the kitchen. It would be best to avoid ‘a pair of scales’ in the context of precision weighing. With their steel knives and planes in continuous contact, the old kitchen and shop scales suffered from wear leading to low precision, but were quite adequate for weighing kilograms of fruit, vegetables, sugar and flour. As we shall see later, fine balances originally also had steel knives and planes, but at the higher end of the market steel planes were increasingly being replaced by agate, and later still agate knives were introduced too. Robinson specified ‘crystal’ planes at one guinea extra in his catalogues. Currently, the terms ‘fine balance’, ‘high-precision balance’ and ‘laboratory balance’ may be taken to mean the same thing: a sensitive balance used primarily in labs for weighing relatively small amounts of chemicals or physical objects, merely a few grams, as opposed to kilograms of produce in a shop. The specific definitions of the different types of fine balance relate mainly to the sensitivity rather than to the capacity or size of the balance. The classifications are in order-of-magnitude steps of sensitivity, and are set out below. Note that the antiques trade and museums tend to specify balances by beam length but this, whilst helpful in a general way, is not a sensible mode of classification, due to the radical shift from long-beam to short-beam balances in the late 19th century. This is especially true in the case of Robinson, who pioneered short-beam balances far ahead of Paul Bunge of Hamburg, who published his seminal paper and produced his radical short-beam prototypes in 1867.10 Bunge’s analytical balance beams were a mere 5″ (130 mm) instead of the then conventional 14″ (355 mm) or 16″ (405 mm) or even 18″ (460 mm). He is usually credited with the highly influential introduction of the short beam, with other makers following his lead immediately, but Robinson in London was there first some forty years earlier. At around the turn of the 19th century ‘chemical balance’ meant any laboratory balance, and their capacities and sensitivities were specified in grains, not grams, although the metric system was rapidly gaining ground in scientific circles in continental Europe. The term has become subdivided into the clearer classifications listed below, and have been expressed here in metric units for clarity to modern readers. The changeover from Imperial plus metric specifications to metric only in balance catalogues occurred during the 1880s, and was essentially complete by 1890. Chemical Balance is in modern lab terms a ‘rough’ balance, with a sensitivity of, say, 10 mg, 1 mg at best, capacity at least 200 g, more usually a half or one kilogram, even two. One milligram is three decimal places of a gram, so these are usually called ‘3- place’ balances. Thus the better ones are included under the ‘fine balance’ heading. Analytical Balance is the correct term for the routine laboratory balance with a sensitivity of 0.1 mg and a capacity of at least 100 g, usually 160 or 200 g, even 500 g. 0.1 milligram is four decimal places of a gram, and the usual practice is to call these ‘4-place’ balances. Those with a capacity of 500 g are termed ‘heavy duty’ analyticals. 10 Bunge (1867). To the best of the author’s knowledge no English translation is available.
4 Assay Balance is a term reserved for a very delicate balance having a capacity of usually no more than two grams, five at most, originally with a sensitivity of 0.1 mg, later 0.01 mg, i.e. 4-place or 5-place balances. These were intended principally for rapid and precise weighing in the assaying of precious metals. The following four terms also need to be specified here. Hydrostatic Balance is nowadays usually regarded as a relatively minor part of the subject, but was very important in the 19th century. It is a particularly important type here, for it is in this area where Robinson began balance construction. Hydrostatic weighing is a subject in its own right with a long pedigree, and is far too large to be covered here. Essentially it is the adaptation of chemical balances and weighing technique for the determination of the density of solid objects by weighing them in air and again in water, noting the apparent loss in weight, and applying Archimedes’ Principle. To this end many makers, including Robinson, supplied some models with a short third pan, used instead of the normal sample pan, for hydrostatic weighing as will be shown in Chapter 6. By an inverse logic, the density of liquids may be determined by weighing a solid object in water and again in the liquid under test, but this does require an appreciable volume of test liquid to be available. Bullion Balance is not a useful term in balance classification. In colloquial usage it refers to a large, very strongly built balance, with a beam length from 30″ to 48″ and intended for the weighing of bullion, with capacities from 1,000 to 10,000 Troy ounces (56 lbs/30 kg to 500 lbs/300 kg).11 Contrary to popular perception and their robust appearance, these really are true balances, not scales. However, many lesser bullion balances were also made; these were small sturdy items with capacities from 50 to 300 Troy ounces. Some small bullion balances are technically scales, in that the end knives and planes are not relieved. Inspectors’ Balance is also not a useful term in balance classification; it merely refers to one use to which any suitable fine balance, from an assay to a bullion, may be put ― the official checking or ‘Verification’ of masses. It means the balance has been critically tested and adjusted, i.e. ‘Verified’, and bears a plate to declare this fact. Precision Balance is a loose, ambiguous term and is best avoided. In Britain it refers to all of the above, but in Germany and for historical reasons it refers to beamscales and simple balances with sensitivities from 2 mg to 100 mg. The term was originally used to indicate only that the balance was more precise than any scales, but as time passed so the fine balances in labs became more precise than the ‘precision balance’ which causes confusion today, hence the avoidance of the term. Many of these so- called ‘precision balances’ were in fact simply superior scales. There are other specified types of fine balance, such as the microchemical balance, but these 20th century types are not relevant to a study of Robinson. A full working knowledge of the above nomenclature will be assumed from here on. These terms were rather elastic, some models standing on the borderline of two classes, and makers have often stretched the definitions in their catalogues. Before critically examining Robinson’s balances, their development and their influence upon later makers, we should look first at what little is known of the man himself, secondly, at his instrument-making business, and then the state of English fine balance-making circa 1800―1820 just before Robinson entered the field. These three backgrounds are to be covered in the next three chapters. 11 One Troy ounce = 31.10347 grams. Twelve Troy ounces = one Troy Pound = 373.242 grams.
CHAPTER 2 T.C. Robinson, The Man HOMAS CHARLES ROBINSON, son of Henry Robinson and wife Mary, was born on September 3rd 1792 in Marylebone, London, and baptised in St Mary’s parish church, Marylebone, on September 30th. Henry Robinson was a hosier, haberdasher and linen draper at 37 Devonshire Street, Portland Place, London, until his death in 1809. Thomas Charles was the third of their four children. He never married and he predeceased his widowed mother. After Henry’s death the family moved to No.38 next door, from where the widow and son Thomas both traded, she in haberdashery and he, from an unknown later date, as a mathematical instrument- maker. He appeared in Pigot’s Directory of 1825, listed as ‘optician’ as was often the case with instrument-makers who made or sold at least a few optical devices. His mother really needed a good window display on the street, so he worked at the rear, probably with a difficult-to-find access from the narrow Clark’s Mews round the corner, which may explain his absences in some directories e.g. Post Office 1841. Clark’s Mews was partly named Upper Wimpole Mews, and later partly renamed again as Dunstable Mews.1 See the map Fig. 1 at the chapter end for the locations of Nos. 37 and 38. The two premises filled that narrow Devonshire Street frontage between Beaumont Street and the Mews, but how far back they went is unknown; they no longer exist. There is a second map Fig. 2 on the following page showing the relationship between Devonshire Street and the wide thoroughfare Portland Place, which becomes potentially important in Chapter 6. Of the four Robinson children all three sons were gifted. The children were: Henry Jnr (1788―1846), pictorial artist, hung at the Royal Academy several times from 1818 onwards; Mary Jnr (1789―1875), spinster, of whom very little is known; Thomas Charles (1792―1841), mathematical instrument-maker; Charles Robert, c.1803―c.1834), pictorial artist. Nothing is known of Thomas Charles Robinson’s education; likewise nothing of his training has come to light, but he did have a clock-making cousin, Henry Blundell, of whom little is known. Henry Blundell was later one of Robinson’s executors. Robinson’s early collaboration with the well-known Captain Henry Kater was critical in his initial designs, from where a clear progression of thought may now be perceived up to and including the ‘Robinson & Barrow’ and other late versions. His working dates are usually recorded as 1821 to 1841, but 1821 as a starting date seems 1 Stock (1968), 256; London A/Z (1998). ‘Mrs Mary Robinson, haberdasher’ is at No.38 in the Post Office Directory of 1841. T
6 improbable. By that date Robinson would have been aged twenty-nine, and had presumably been working for some years. Buchanan found a note by Kater suggesting Robinson could have been working for or with Kater as early as 1818, but it only suggests a possibility, no more.2 Thus Robinson’s education, training and start-up as an instrument-maker are all totally obscure, and likely to remain so. Having retired to Ealing Green, Ealing, where he appears together with his sister Mary in the 1841 census (8th June), he returned to his mother at Devonshire Street to die of chronic bronchitis aged only 48 on 25th July 1841.3,4 He was buried on 1st August in the same grave as his brother Charles Robert in the graveyard of the church of St Mary at Finchley.5 The inscription on the headstone reads: Endowed with the highest intellect and devoted to science he attained to eminence in his profession and was highly esteemed. No portrait of him is known to exist. His will of April 1841 (PROB. 11/1950/295) provided mainly for his mother and yields only two items of relevance to this study: 1) I direct that as soon as conveniently may be after my decease and as may occur to my Executors consistent with its advantageous disposal that my executors do make sale of my stock in trade tools and every other saleable property belonging to me excepting my wearing apparel books and articles appertaining to the person either by private contract or public sale or both and I hereby empower them to carry on my business for such short period as may seem necessary for the completion of orders in hand the rendering fit for sale goods in the course of manufacture and the advantageous disposal of my property. I hereby will that after my just debts and funeral expenses testamentary bequests and other contingent expenses are paid that the whole rest and residue of my property be invested in the public funds in the names of my said executors Thomas Charles Blundell and Henry Blundell in Trust for the hereinafter mentioned purposes. [These ‘purposes’ were provisions mainly for the welfare of his mother, also for his brother and sister.] 2) I will and bequeath to my Apprentice John Dover in testimony of my approbation of his zeal and integrity the sum of five pounds to be paid to him when he attains the age of twenty one years. [£5 in 1841 translates to about £600 in today’s currency; a most generous bequest.] This John Dover (1824―1881) continued to make balances in the Robinson style for some years, as did Henry Barrow (c.1790―1870) who bought Robinson’s business in 1842.6 Both were competing with Ludwig Oertling in his start-up years, and Dover exhibited a Robinson-style balance at the 1851 Great Exhibition. Robinson’s full will is to be found in Appendices 1 and 2, in both its original form and transcribed into modern English. It is worth noting that Robinson appointed as his executors his cousins Thomas Charles Blundell and Henry Blundell; the former was a rope, twine and mat-maker, and the latter a clock-maker. In his will another cousin Mary Blundell is mentioned. Note the frequent recurrence of the same fore-names in the two families. It seems clear that the Robinsons and the Blundells were closely related, which the author takes to be suggestive that Robinson most probably learnt instrument-making from clock-maker Henry Blundell. This is far from proof, but it is the only potential source for his training yet discovered. 2 Buchanan (1982), 225, quoting from Phil. Trans. (1818), 40. This is reconsidered in Chapter 6. 3 1841 census ref. HO107/689/9, Ealing Green, Ealing. 4 Death Certificate, 27th July 1841, Marylebone, sub-district Cavendish Sq., entry 363. The informant was his brother Henry. 5 Stock (1968), 256; Burial Register for Finchley, 1841, entry 1498. 6 See Chapter 3.
7 Despite extensive efforts it has proven impossible to find in what respect T.C. Robinson is a cousin of the three named Blundells. The core problem is that his mother’s maiden name has not been found anywhere; his parents’ marriage certificate cannot be found and his entry in the relevant 1792 baptismal register does not, naturally, give the mother’s maiden name. Now consider this point: his father died in 1809, when Robinson was only seventeen. This perhaps was the point at which Henry Blundell took Robinson in hand and began to teach the lad clock-making, but whether as an official apprentice or not is unknown. Alternatively Blundell may have taken young Robinson on at the more conventional starting age of 14, i.e. as a school-leaver. Either way, Henry Blundell would have taken pity on his relative the widow Robinson and saved her much worry, time and expense looking after the lad. But these are merely speculations. FIG. 1. 37 AND 38 DEVONSHIRE STREET. The two properties between them cover the Devonshire Street frontage of this narrow block, but how far No.37 returns down Beaumont Street and how far No.38 returns down the Mews is unknown. No.37 was the original home of the Robinson family; No.38 was their later home and Mary’s shop, with Thomas’ workshop to the rear with an entrance off the Mews. No.37 (left) and No.38 (right)
8 FIG. 2. DEVONSHIRE STREET AND PORTLAND PLACE. The true significance of the location of the termination of Portland Place and its possible relevance to Kater and Robinson will become apparent in Chapter 6. Nos. 37 and 38. Original termination of Portland Place.
CHAPTER 3 T.C. Robinson, The Business HE FIRST KNOWN REFERENCE to Robinson as a maker of fine and original balances occurs in 1821, implying that he is in business as an instrument- maker, but this is not made entirely clear.1 These early instruments were made to a design by Captain Kater of an ‘economy’ hydrostatic balance. Robinson and Kater then developed this basic design into a superior chemical cum hydrostatic balance. For readers unfamiliar with hydrostatic weighing, see Appendix 3. The fact that Robinson’s later balances were highly regarded is exemplified by their use in the adjustment of the Imperial Standard masses in 1825, 1829 and their replacements post-1834. By these dates Robinson clearly had a thriving business. Three catalogues survive, showing that he was making a very wide range of scientific instruments, mainly in the areas of optics and magnetism. There is a strong emphasis on astronomical instruments, with balances appearing to be a minor area. This may be an illusion, created by the fact that having devised a successful balance design, he continued to produce these in quantity, but requiring no novel catalogue entries. There are suggestions that Robinson balances existed in considerable numbers in the latter half of the 19th century, both from the frequency of mentions in the literature, and that the Oertling ledgers and day books were recording Robinson balances coming in for repair as late as the 1870s. This also shows that the balance owners held their Robinson balances in high regard, such that they were considered to be still worth servicing when up to fifty years old. From his catalogues of 1828, 1829 and the posthumous one of 1842 we know he was routinely offering only one standard balance design in three sizes as tabled below, but there were incremental refinements and genuine improvements made as time progressed, some initiated by Kater and some by Robinson. These figures are taken from his priced catalogues of 1828 and 1829: BEAM LENGTH, INCHES CAPACITY, GRAINS CASE SIZE, INCHES, w x d x h PRICE, GUINEAS 5½ 400 9 x 5 x 10 Eight 8 1000 13 x 6 x 13 Eleven 10½ 2000 18 x 7 x 15 Fourteen Their metric equivalents in grams: 400 gr = 26 g; 1000 gr = 65 g; 2000 gr = 130 g. 1 Kater (1822) double reference; to be considered in detail later, in Chapter 6. T
10 Fig. 3 shows the first page of his 1828 catalogue, appearing in English in a German astronomical journal.2 The balances are placed first, and occupy a very tiny portion of the whole. The rest of this 1828 catalogue has been omitted here. FIG. 3. THE ROBINSON CATALOGUE OF 1828, FIRST PAGE ONLY. Below and opposite is the full Robinson catalogue from the same journal in the following year, 1829.3 The journal has laid out both Robinson catalogues in its unusual house style of two columns per page, with references made according to column number, not page number. The 1829 catalogue has been completely re- ordered with astronomical instruments first, then surveying instruments, and finally philosophical instruments, with balances leading that section. They are defined as ‘Balances for Chemical, Philosophical and Assaying purposes.’ FIG. 4. THE ROBINSON 1829 CATALOGUE, FIRST PAGE. The sparse balance notes of the 1828 version have been most helpfully expanded in the 1829 version, thus: 2 Astronomische Nachrichten 7 (1828), coll. 231-232. 3 Astronomische Nachrichten 8 (1829), coll. 47-50.
11 ‘The above balances have steel knife edges acting with agate plane surfaces, with crystal knife edges instead of steel each size 1 Guinea extra. Larger sizes are also made to order their arrangement embraces several important improvements by Captain Kater, Dr Wollaston and the Maker, all the size[s] are furnished with platinum weights to 100 Grains, and upwards of brass. They will indicate .001 of a Grain with 1000 Grain.’ FIG. 5. THE ROBINSON 1829 CATALOGUE, SECOND PAGE. There is much to consider here. First, in both, Robinson’s claim that his balances could resolve to 0.001 grains in 1000 grains (0.065 mg in 64.8 g) or one part in a million, was a mighty claim at that date for routine production items. His balances
12 had steel knives working on agate planes, but ‘crystal’ knives instead of steel were available to order for one guinea extra. ‘Crystal’ simply means agate in this context.4 The use of all-agate bearings shows Robinson’s concern with maximising the lifetime of his knives, thereby maintaining his ‘one part per million’ claim. Clearly Robinson saw his products as being of very high quality, and at the leading edge of the possible. Secondly, larger sizes were also available but made only to order, such as an 18″ balance for Dr Moll of Utrecht, in use in 1831 for his inter-comparison of British, French and Dutch National Standard masses, and subsequently lost. Whilst this book was being written Dr Moll’s balance was rediscovered in the Netherlands and has been returned to the University of Utrecht; see Chapter 15. A 15″ Robinson-style balance by Robinson’s successor Henry Barrow was later used by Professor Miller of Cambridge, and is currently missing. A third ‘special’ has recently been noted in the literature, made for the Natural Philosopher H.C. Schumacher; this too is missing. Thirdly, the brass masses supplied in the drawer of the 8″ version were 400/300/ 200/100 grains, with platinum ‘fractions’ (smaller masses) exactly as specified above. It is interesting to note that Robinson’s masses are in the ‘old’ grains 4-3-2-1 system, rather than the more usual ‘modern’ grains 6-3-2-1 sequence.5 Much more is said on Robinson’s weight sets in Chapter 23. It is arguably true to say that Robinson was the first balance-maker, certainly in England, to offer routine production high-precision balances with all the necessary fine adjustments built in, by which a competent user could ‘fine-tune’ the balance to his exact requirements in terms of precision, sensitivity and speed; this will be demonstrated in a later chapter. Robinson’s lead in this regard was followed rapidly by August Oertling in Berlin and shortly after by Ludwig Oertling in London. Robinson was one of the first English instrument-makers to offer high-precision balances on a fully commercial basis, meaning standard models from a printed, priced catalogue. Benjamin Martin was probably the earliest, his first known catalogue issued as early as 1757. W&S Jones offered hydrostatic balances in their catalogue of scientific instruments and demonstration apparatus from 1814 onwards. This catalogue was frequently updated and reissued over succeeding years, in 1817, 1818, 1825 and 1827, all before Robinson’s first known catalogue of 1828. W&S Jones competed with Robinson throughout his working life, for they carried on until 1855. Most of this competition is not examined here, but it is interesting to see how the two companies differed in their attitudes to balance design; to be pursued later. As a mathematical instrument maker Robinson offered a great range of scientific instruments in his priced catalogues, as shown above. He used a few laboratory supply houses as agents, such as J.J. Griffin and John Ward. He was exporting at least a few balances to Europe; certainly his designs were known there and well respected, witness Moll and Schumacher above. At the 1844 Berlin Exhibition, Reimann exhibited nine chemical balances of which one was in the Robinson style. Even as late as 1855 the maker Oechsle of Baden was advertising chemical balances made in the Robinson style alongside ones in his own style and those of others too. The extent to which the balances contributed to Robinson’s turnover is unknown, but in the light of the numerous appearances in the literature, they must have been at least a steady and reliable bread-and-butter line. 4 For a detailed account of agate/crystal and other bearing materials, and the origin, extraction and preparation of knives and planes from agate see Oliver (2020), 840-845. 5 For the various sequences used by weight-makers in grains and other units, see Oliver (2020), 864-866.
13 Returning to his will from Chapter 2, Robinson clearly envisaged his executors would allow the business to complete orders already in hand, then to close it down and sell the remaining stock and tools. He seems not to have envisaged the sale of the business as a going concern, but that is what happened. It was bought from the executors in 1842 by Henry Barrow (c.1790―1870) who declared the business would continue there as ‘Robinson & Barrow.’6 Nothing has come to light concerning this sale to Barrow, nor is it known if the widow Robinson remained at No.38 or not. It is most likely that she went to live with her spinster daughter Mary in Ealing Green. Mary the daughter remains an elusive figure, but has made a fleeting appearance via a copy of a letter from the Scottish Natural Philosopher James D. Forbes. Only a brief mention is made in a Forbes letter book, thus:7 JDF to Miss Mary Robinson 31st January 1843 Sends £28-17-6. “… I have not quite made up my mind about the vertical vane [ ? ] Instrument.” Note the date! This was eighteen months after the death of her brother, and the Robinson business had been bought in 1842 and was being run by Henry Barrow. Do we infer from this note that Mary was working for Barrow, managing the paperwork side of the business, i.e. sales and accounts? If so, perhaps she acted similarly, earlier and unreported, in her brother’s business too? It is a possibility to consider. The reason for the posthumous catalogue of 1842 shown in Fig. 6 overleaf was, presumably, that Barrow was keen to sell off all the remaining stock instruments which were engraved with Robinson’s name and address, before introducing his own catalogue of ‘Robinson & Barrow’ instruments, shown in Fig. 7. The posthumous catalogue contains the odd anomaly of the 5½″ model now appearing as being 6″. This may be just coming to terms with reality, for several Robinsons examined so far are somewhat over 5½″, and the ‘Miller’ (Chapter 9) is only just under 6″.8 The Robinson & Barrow catalogue contains much the same selection of astronomical, surveying and philosophical instruments as before, although the balances get only one line, the last, for ‘Chemical and Philosophical purposes’, with assaying omitted. The original Robinson balance designs continued to be made by Barrow, marketed as made by ‘Robinson and Barrow’ and at the same address but only until 1843. In a manuscript letter to the Astronomer Royal Sir George Airy, dated 29 September 1843, Barrow announced that he had now moved the business to 26 Oxendon Street, Haymarket.9 Here Barrow and the company were making not only Robinson & Barrow balances but many other types of scientific instruments, with a distinct emphasis on surveying instruments and on magnetic instruments such as dip circles and compasses. The move to 26 Oxendon Street was definitely between May and September 1843, but appears to be in 1845 according to the Post Office directories. The directories are notorious for the time lag on their updates.10 So, who was Barrow? His earlier career made him very suitable as a successor to Robinson. Henry Barrow, c.1790―1870, was an outworker in the mathematical and philosophical instruments trade in London in the 1820s, serving various prestige 6 Stock (1971), 387; McConnell (1985), 46; Chaldecott (1989), 160, noting Barrow’s letter to this effect sent to the Astronomer Royal, Sir George Airy, 30/8/1842, ref. RGO 6/718, f.269r. A first Robinson & Barrow price list was date-stamped 31AU1842 by the RGO; ref. RGO 6/718, f.270r; see Fig. 7. Therefore it presumably accompanied the 30/8/1842 letter. 7 J.D. Forbes letter book III, item 476, in St Andrew’s University library. 8 At these dates the exact overall beam length was not critical; anything within ± ¼″ of nominal in all sizes is quite common. 9 Chaldecott (1989), 162, noting Barrow’s letter of 29/9/1843 on relocation, ref. RGO 6/718, f.277r. Oxendon Street and Haymarket are much closer in to the heart of London, near Piccadilly Circus. 10 Chaldecott (1989), 167-168, giving a brief analysis of the delays and his deductions concerning the reasons for them.
14 workshops such as those of Nairne, Dollond and others. After meeting Sir George Everest, the Surveyor-General of India in 1829, he joined the Trigonometrical Survey there, based in Calcutta. His hope was to make enough money to set up later on his own in London. His principal activities were the maintenance and repair of the surveying equipment out in the field, and the setting up of newly acquired instruments there. In these activities he demonstrated a remarkable ability to produce excellent work under trying conditions and without a permanent workshop and with few facilities. This is reminiscent of Captain Kater’s spell on the Everest Survey of India some years before Barrow, in 1799―1808. FIG. 6. THE ROBINSON 1842 POSTHUMOUS CATALOGUE.
15 FIG. 7. THE ROBINSON & BARROW CATALOGUE OF 1842. Sir George Everest and Henry Barrow were both strong-willed characters with very different personalities; this to such an extent that they fell out with each other so seriously that Barrow was eventually dismissed and sent back to England in 1839. It appears that he must have made enough money to achieve his ultimate goal of a workshop of his own. Rather than creating a new workshop he next surfaces in history by purchasing the Robinson workshop in 1842. The business quietly flourished under his command, making surveying instruments. These included some for the Indian survey, for his workmanship there had never been in question, only his behaviour towards Everest. The workshop also made magnetic dip circles, and compasses, and continued the production of the three standard Robinson balances.
16 For more on Henry Barrow see references.11, 12 Not only did Barrow engrave his balances ‘Robinson & Barrow’ but his invoices circa 1850 carried the subheading ‘Successors to the late T.C. Robinson, from 38, Devonshire Street.’13 After 1850 all instruments were made under Barrow’s own name, with the business appearing in the directories of 1850 to 1864 as ‘Henry Barrow’ or ‘Henry Barrow & Co.’ The subsequent history of the business is unclear. In 1865 and 1870 the Post Office directory entry for 26 Oxendon Street reads ‘T. Owen, mathematical instrument maker.’ From 1871 to 1877 the entries refer to ‘Barrow & Owen.’ One assumes that Owen used Barrow’s name in the same way as Barrow used Robinson’s name, to enhance their prestige. So, in one form or another, Robinson’s original enterprise of the early 1820s continued until well into the 1870s, for over fifty years. The apprentice John Dover (1824―1881), noted in Robinson’s will, also continued to make balances in the Robinson style for some years until at least 1851, as well as instruments for research in magnetism. Dover exhibited a Robinson-style balance at the 1851 Great Exhibition. Regrettably no Dover balance is known to have survived, although a few of his magnetic instruments have. See reference for more on Dover’s career and his instruments.14 It should be made clear that neither the Devonshire Street nor the Oxendon Street premises now exist. 38 Devonshire Street was later altered, but when Stock went to examine it in 1967 he found it had been recently demolished. The current 38 Devonshire Street is in a completely different location some 100 yards away from the original site.15 26 Oxendon Street has been rebuilt.16 It is to be regretted that neither address is illustrated in Tallis’ London Street Views 1838-1840. Finally, having seen the development of the Robinson instrument-making business as recounted in this chapter, one has to ask exactly why the seemingly little-known Robinson was employed by Captain Kater circa 1821 to make his ‘economy’ hydro- static balances. He may have been merely a maker of instrument parts to Troughton and other workshops at this period, and Kater spotted him that way. Another possible answer is that the more obvious well-known makers, in particular Troughton, were already over-loaded with work. But what was Robinson doing in his twenties? Possibly he was not as unknown as he appears to be now, but was then already noted for his optical and magnetic instruments, all undated as far as the author knows.17 Yet Buchanan was unable to find any notices of a Robinson business in the trade directories of the period.18 Following his 1821 collaboration with Kater, his earliest-known catalogue of 1828 with its comprehensive range of scientific instruments cannot possibly have arisen out of nothing; there must have been earlier production on a smaller scale, with or without formal catalogues. It is unlikely now that we shall ever know exactly how Robinson began his instrument-making career. 11 Insley (2004). 12 McConnell (1985). 13 Stock (1971), 387. 14 Stock and Laurie (1980). 15 Stock (1968), 256. 16 Stock (1971), 387. 17 The only precisely dated Robinson instrument known to the author is a portable barometer of 1832, too late to be of any relevance here. The item is in the National Museum of Scotland, accession no. T.1984.52. 18 Buchanan (1982), 209.
CHAPTER 4 English Fine Balance Design circa 1800―1820 HE CAPACITIES AND SENSITIVITIES of fine balances were expressed solely in grains in England at around the turn of the 19th century.1 This was despite the fact that the Metric system was rapidly establishing itself amongst researchers in continental Europe. It was also before the traditional English units of mass, length and volume were rationalised and standardised in an Act of 1824, and became styled ‘Imperial’ units. The 1824 Act was entitled ‘An Act for ascertaining and establishing Uniformity of Weights and Measures’ 5 Geo. IV c. 74.2 R.B. Bate was heavily involved in the making of Model (prototype) sets of the new Imperial Standard weights and volume measures for the government, as required under the Act. His part in this endeavour, under the guidance of Captain Kater, will be considered at the chronologically correct point in the narrative below. This chapter reviews the English fine balance-making scene before Robinson and Kater began to influence it, in so far as this is possible given the few balances surviving from this era. The roll-call of English instrument-makers creating fine balances is long but, as said before, most of their balances were one-off specials or only a minor part of their output. The customers for these specials at the opening of the 19th century were mostly minor nobility and other gentlemen of independent means, pursuing scientific research in whatever areas interested them. The words ‘science’ and ‘scientific’ were unknown then, and these researchers termed them- selves Natural Philosophers, and had done so throughout the 17th and 18th centuries. Many were Fellows of the Royal Society (FRS), and published papers in learned journals, most notably the august Philosophical Transactions of the Royal Society of London, usually referred to nowadays simply as Phil. Trans. As the 19th century progressed so these researchers were increasingly becoming salaried professional academic scientists, e.g. Professors Daubeny and Maskelyne at Oxford and Professor Miller at Cambridge. Others became fee-earning professional consultants. The independent Natural Philosophers appear to have effectively died out by the end of that century. In this period of an expanding British Empire the fine instrument-making industry survived largely on Government contracts, such as supplying telescopes and sextants in quantity to the Royal Navy; likewise similar contracts for the merchant navy and, inland, surveying instruments for the military, for mines and for builders of canals and, later, railways. After 1824 many were supplying sets of Imperial standards of 1 One grain = 64.7989 milligrams; 1 gram = 15.4324 grains. 2 For the background to the Act and its implementation, see Connor (1987), 249-278. T
18 mass, length and volume to local authorities and the colonies, as did Bate after having completed the prototypes. This bread-and-butter work meant that the fine instrument-makers were available to supply Natural Philosophers’ needs, ordering custom-made specials, which were therefore termed ‘philosophical instruments’. What did the fine balance field look like as T.C. Robinson entered it? A good starting point is Henry Cavendish, for his indirect but great influence on Robinson and other instrument-makers, indeed on Natural Philosophy itself. He and Jesse Ramsden are treated first and at some length because it is frequently necessary in the rest of the book to refer back to these two and their innovations. The remaining influential makers and Natural Philosophers are listed as far as possible in chronological order, with some biographical data on each, but only that which is relevant to fine balance- making or balance usage. For more biographies see their DNB entries and the other given references. The Honourable Henry Cavendish, FRS (1731―1810) was a reclusive and highly eccentric nobleman, but with a brilliant scientific mind.3 Educated at Peterhouse College, Cambridge, he left without a degree, but established important connections for himself amongst the Natural Philosophers. He investigated all manner of topics of interest to the Natural Philosophers of his day, but is probably best known for his attempts in 1797―1798 to determine the universal gravitational constant, G, and the mass of the Earth. He used a delicate torsion balance consisting of a 6-foot wooden pole carrying two small lead balls acting under the influence of two large nearby weights. The historical importance of this work is not so much in the values he obtained, but the inordinate lengths to which he went to eliminate every possible source of error: variations in temperature, humidity and air pressure, plus vibration caused by nearby human activity. It may be argued that it was his extreme attention to detail in these experiments which marked the start of experimental rigour in scientific research. This extreme attention to sources of error was also present when he designed a philosophical balance at some unknown date 177x, and it was clearly intended to be of a high capacity and high sensitivity. Its intended purpose is unknown, but was possibly for his investigations on the composition of water in 1781―1784. It is large and is initially unimpressive, having a crude wooden case, a heavy duty and very complicated arrestment system, and a tapered hollow beam of sheet iron 472 mm (18½″) long k-k, 19½″ overall. The original is in the Royal Institution, and careful inspection of this and the Science Museum copy discloses several important refinements novel at that date.4 All three bearings are knife and plane, in steel. The arrestment frame has cradles at each end which catch not only the end planes, but the beam too, quite separately, in a theoretically correct manner. Each end plane has a very small area but the principle is there, and is far superior to any hook and loop arrangement. There are small screws holding the end knives allowing for adjustment of arm lengths and parallel. The centre knife is adjustable vertically to ensure the three knives are coplanar. Cavendish reported it to be sensitive to one tenth of a grain, about 6 mg, and this was confirmed by Stock in 1965 and Buchanan in 1976. The left end of the beam carries a small mass on a threaded rod for the fine adjustment of poise (zero), and the right end of the beam carries a pointer running over a vertical ivory or bone chart attached to the case and graduated 13―0―13. A 3 This section is based largely on: Schaffer’s DNB 2004 biography; Barclay (1935 supplement), s4-s6; Stock (1969), 11-12; Buchanan (1982), 120-124, 361. 4 The original is in the Royal Institution, accession no. 15. The Science Museum, London, copy has accession no. 1930-770.
19 screw-threaded small post with a gravity bob projects from the underside of the beam, for fine adjustment of the sensitivity. Each pan is suspended from its end plane by a universal joint, but the current pans and wires are not original. The first true panstops are seen here, i.e. operating as an integral part of the release action. There is also a strange pair of movable vertical posts beside each pan, intended to reduce pan-swing during release. As will be seen later, in Chapter 5, this feature is usually attributed to someone other than Cavendish or his instrument maker. At the top of the stand there is a brass frame bearing the divided centre plane and two spirit levels set at right angles (‘crossed levels’). To complement the levels the table beneath the case carries four brass feet height-adjustable by rotation. Jesse Ramsden, FRS (1735―1800), after having tried the cloth trade, was bound in 1756 at the relatively late age of twenty-one to a mathematical instrument-maker named Mark Burton. Ramsden soon acquired a reputation as a mathematical scale engraver, and began trading under his own name in 1763 in a workshop in the Strand, later moving to Haymarket, and in 1780 to 196 Piccadilly. Ramsden was a perfectionist and was regarded as the finest instrument-maker of his day. He is now best remembered for his astronomical and surveying instruments, his dividing engines and the massive cone-beam balance he constructed for the Royal Society in 1788, and which now bears his name; it is currently on permanent loan to the Science Museum, accession no. 1900-166.5 The most important gifts of the Ramsden balance to posterity are, firstly, its pioneering use in this country of agate as a plane bearing material and, secondly, its use by Charles Blagden in his ground-breaking work of 1789 for the Royal Society on density measurement as a possible basis for levying Excise Duty on spirits.6 It resembles the Cavendish balance in that it is a curious mixture of the mostly excellent but some indifferent, in both design and construction. The centre bearings were 80° steel knives upon agate planes, but the end bearings were only steel knives and loops, yet it managed to achieve a sensitivity of one hundredth of a grain, 0.6 mg. The four pillars, see Fig. 8, supported a fixed frame holding the agate planes, with the central push rod lifting a pair of notched arms to raise the beam out of action. This was therefore an early form of descending arrestment frame. It is worth noting that the arrestment is not entirely gravity-driven, as there is a compressed spring hidden at the top of the push rod, ensuring a consistent descent into action. The instrument was fitted with a pair of panstops controlled independently of the release. The highly original cone-beam was 61 cm k-k, 24″ exactly, maximum diameter about 3″. The sensitivity was adjusted by an internal mass rising or falling on a screw- threaded adjustable flag post, thereby providing both sensitivity and zero adjustment. The cone-beam concept was originally devised by Ramsden to solve the problem of building a balance of high capacity and high sensitivity. He was already using the principle of folding sheet metal into cones as structural elements in his telescope and theodolite mounts, providing strength with lightness. This worked very well where heavy stresses were shared between multiple cones. Here, he conceived the idea of joining two brass cones base to base to create a light but strong beam. Whilst each cone has great inherent strength, the difficulty of the concept is that of joining two together and keeping the joint absolutely rigid. 5 For the full biography of Ramsden see McConnell (2007). For the balance see McConnell (2007), 178-180. 6 Oliver (2020), 141-142, for a critical assessment of this balance. For the construction of the balance see Rozier (1788); for its use see Blagden (1790).
20 His solution was to have a heavy casting over which each cone overlapped and was soldered, the ugly soldered joints being covered by a brass band, as seen in Fig. 8. There was then the simultaneous difficulty of mounting the centre knife on this centre casting plus joints. There were similar but lesser considerations at the beam ends too. Further, each cone was reinforced with a casting part way along its length. The result was a heavy beam, as those who have held the Ramsden cone beam in one hand will testify. However, because the Royal Society had over-specified it, the balance was very successful and greatly lauded. The beam-end steel pointers run over curved ivory vertical charts graduated 45―0―45. Much more could be said, and much more research is needed on the internal construction of the beam. Endoscopy via the tiny covered inspection port on the underside is obviously the way forward. When necessary the Ramsden could be made suitable for hydrostatic work by having the conventional four brass feet replaced by legs some 9 or 10″ long raising the cabinet high off the bench together with long hook-ended steel suspending rods.7 For readers unfamiliar with hydrostatic weighing, see Appendix 3. FIG. 8. A RARE, EARLY PHOTOGRAPH OF THE RAMSDEN ON ITS LONG LEGS AND BOARD. McConnell in Appendix 1 of her Ramsden biography reports on page 299 of his making assay balances of high sensitivity, but the author has not been able to verify this. Ramsden is also reported as making folding coin balances, of no relevance here. 7 Now resident in the Science Museum, London, but unfortunately the four legs, the big board to receive them, the two sets of pans and their suspension pieces, the steel suspending rods, the two spirit levels and other parts have remained at the Royal Society and now appear to be lost.
21 Benjamin Martin 1704 [bap.]―1782) was initially a teacher, writer, publisher and travelling lecturer in experimental philosophy.8 Not until in his fifties did he set up a permanent shop in 1756 at 171 Fleet Street.9 From here he sold numerous types of philosophical instruments as well as performing lecture-demonstrations. He issued various catalogues of instruments from 1757 onwards, in which he claimed various inventions and improvements, some rather dubious. His genuine improvements were in optics and especially in his microscopes. [DIANA CRAWFORTH-HITCHINS. FIG. 9. BENJAMIN MARTIN COIN BALANCE. [DIANA CRAWFORTH-HITCHINS. FIG. 10. BENJAMIN MARTIN COIN BALANCE, BEAM END. 8 This Martin section is based mainly on Millburn (1986). 9 At the sign of the Visual Glasses. Fleet Street was not numbered until July 1766.
22 He was more of a retailer rather than an instrument-maker, despite many surviving instruments bearing his name. He was selling all types of optical instruments as well as barometers, orreries, globes, etc., but it is not possible to say if some items were made by him or merely sold by him. He was one of the first to offer philosophical instruments off-the-shelf, and the catalogue of the auction of his posthumous stock in 1782 shows an astonishing quantity and range of instruments. These included 180 ‘Balances’ most of which were probably the ivory steelyard coin balances, serially numbered, of which he made about 5,000. This is an early instance of balances in Britain serially numbered in the manner of the clock-makers of the period. Beside the coin balances he also sold hydrostatic balances in mahogany boxes. He was making philosophical balances and not common trade scales, but because of their delicacy, coin scales/balances were considered to be philosophical devices. There is one surviving coin balance (in a private collection) which demands serious attention; shown in Figs. 9 and 10. For detailed reasons its most probable date is 1774―1778, but it cannot be before 1766. This is a hand-suspended coin balance, but clearly made by an instrument-maker rather than a trade scale-maker. There is no maker’s name or serial number, so it may not actually be made by Martin despite the presence of his trade card. The beam length is 172 mm (6¾″) and it is built with a delicacy befitting an instrument-maker. The brass pans are suspended on dark green plaited silk cords. The pointer is exceptionally long and fine and, above all, the swan neck ends are each horizontally screw-adjustable for arm length; precision adjustment has replaced brute force. The end of each swan bears a knife edge, from which hangs a double loop in the form of a shackle; see Fig. 10. On current knowledge these seem to be the first screw-adjustable swans, and not reappearing until those of Robinson c.1821. George Adams Snr (1709 [bap.]―1772) and George Adams Jnr (1750―1795) were noted makers of fine instruments in quantity, but of very few balances. The only two known to the author are hydrostatic scales of c.1760, with plain, solid brass beams with poor non-adjustable end fittings, mounted on decoratively-turned brass stands, hardly to be taken seriously as philosophical instruments.10 The Adams can be ignored as having made no real contribution to the development of the balance. John Whitehurst, FRS (1713―1788) was a noted watch and clock-maker; indeed his early reputation was based on his ability to make extremely accurate clocks. He was a member of the Lunar Society and a man of wide interests and capability.11 He became in 1774 the first to hold the office of Public Stamper of Weights at the Royal Mint. These ‘weights’ are coin weights, not ordinary weights. For the weighing of Mint production coins and the reference coin weights very delicate and sensitive balances were needed, and these Whitehurst had to build. He also made two weight sets for Joseph Priestley in 1785 and at least two assay balances for Matthew Bolton in 1768 and 1785. Others have been mentioned with notable sensitivities, one of 1/100 grain (0.65 mg) and of 1/2000 grain (0.03 mg). Three balances, of which nothing definite is known, were made by him for the office of Public Stamper of Weights. On current knowledge only two complete Whitehurst balances survive. a) One in Birmingham Assay Office, accession no. 978, is an assay balance said to have been in use there from 1773 to c.1860.12 Another source suggests it may be the one 10 Warner (2018), 34; also Whipple Museum Wh.1871, which has some good accessories, in Brown (1982). 11 For a full and excellent biography see Craven (1996). 12 Holbrook (1992), 31 (photo) and 92 (text). Holbrook reference B10.1. Unfortunately the only available photograph, in Holbrook, is too poor to show any useful detail.
23 he made for Matthew Boulton in 1785, which seems a more probable date.13 Neither Buchanan nor the author has examined this instrument. It is of a very traditional scale-maker’s form, with a suspended slender steel rod beam with swan-neck ends, hoisted aloft via pulleys in a turned wooden pillar, and held suspended by a heavy counterweight on the cord. The pans are small brass dishes suspended on fine silk cords. It has a glazed mahogany case with a curious front door that is hinged at the top, opening outwards and upwards; a most awkward arrangement for the user. But on the progressive front it has brass adjustable pan tables, moved seemingly by a large horizontal action handle at the case front, i.e. panstops. This seems to be the second recorded instance of panstops following Cavendish’s first. Presumably they were made to descend in order to give room for a greater arc within which the beam could oscillate, because the hoist distance via the big counterweight is rather small. b) one in the Boerhaave Museum, Leiden, inventory no. V22456; see Fig. 11. The assay beam is a slender steel rod 30.7 cm (12″ exactly) between knives, suspended from the top of the ornate case. Each end bearing is a pair of loops in the form of a shackle, rather as Martin’s, above, but over a unique form of fixed steel knife; see Fig. 12. [BOTH: NATIONAL MUSEUM BOERHAAVE, LEIDEN. FIGS. 11 AND 12. WHITEHURST BALANCE V22456, BOERHAAVE MUSEUM, LEIDEN; GENERAL AND DETAILS. This knife style has not been repeated by other makers as far as is known. The silvered chart is also suspended, ensuring horizontality, and is graduated 25―0―25 in steps of 5. Thus we have for the first time a wide and easily readable central chart, allowing the ‘method of swings’ for speeding up weighings to be used, although it is uncertain whether this technique was used at this early date. His signature is 13 Craven (1996), 124 f/n 55.
24 engraved as Whitehurst DERBY on the rear of the chart. The pan tables are each on three long and adjustable legs. There is a complicated hoist mechanism with which to lift the beam and thereby lift the pans from the pan tables; the control for this is hidden at the lower front of the case behind the lockable centre flap. This is quite a large instrument, the case dimensions being 72 cm high, 50 cm wide and 33 cm deep. Access is via a conventional front doorslide, a great improvement of the preceding model a). As to whether this is the first practical doorslide, see the discussion in Chapter 22. A date of c.1765―1775 has been suggested for this instrument. In a possibly relevant aside, among Whitehurst’s many published philosophical papers, there is one which covers his ideas on the measurement of length, capacity and weight from the measurement of time, anticipating Kater’s work in this area.14 Jeremiah Sisson (c.1720―1789), son of the greatly acclaimed instrument-maker Jonathan Sisson (?1690―1747 or 1694―1749). Jeremiah was famed mainly for his astronomical and other optical instruments. At least one balance by him is known, currently resident in the Science Museum, inventory No. 1927-2040 (was 1927- 1947). It is officially a part of the George III Collection and numbered E12. It has a 12″ solid brass beam of lozenge cross-section, with a superior type of box ends; it is therefore technically a scale, as seen in Figs. 13 and 14. Two steel cones perform the function of a centre knife, working on two circular steel planes when on release. The arrestment is of a rack and pinion type, worked by a detachable front handle in the form of a clock key. The stand consists of four pillars enclosing rods supporting a complex structure, part of which functions as an arrestment frame, and part as the holder of the twin centre planes. The beam carries a 2″ pointer which moves in a very short arc above a very short chart held on top of a brass post inside the stand as shown. The chart post rises with the rising plane, keeping the pointer and chart in their correct spatial correlation. [BOTH: P.D. BUCHANAN. FIG. 13. THE SISSON BALANCE SCI. MUS. 1927-2040. FIG. 14. DITTO, WITH DOOR RAISED, PLUS HANDLE. The whole instrument contains an excessive number of precision adjustments which are not fully understood. It is clearly made to a very high standard. The glazed mahogany case is a high quality item too, with a single, wide, fitted drawer and four brass adjustable offset feet. The full-width front doorslide has no counterpoise but can be latched. The 1″ diameter circular spirit level is not shown in the photographs, 14 Whitehurst (1787).
25 being hidden behind the plinth of the stand. The entire lantern is held onto the drawer base by eight small screws; a most unusual form of case construction. No date is available for this instrument other than pre-1789. The instrument is signed: J. Sisson, London on a brass strip inset in the wood just in front of the stand. Jean Hyacinthe de Magellan, FRS (1722―1790), born João Jacinto de Magelhães, was a Portuguese Natural Philosopher and a man of very international outlook. He lived in London from 1763 onwards. He travelled widely and frequently throughout Europe, was a member of numerous leading philosophical societies, moved easily in elevated social circles and seems to have known almost all of the leading Natural Philosophers of his day, as well as many leading instrument-makers. He is said to even have had connections with the Lunar Society in Birmingham. Because of his multiple connections he was often supplying English-made instruments to foreign customers; this was at the time when the London instrument-makers were held to be the best in the world. He also devised improvements to a wide variety of philosophical instruments, and some of his ideas were used by leading makers such as Ramsden and Nairne.15 Magellan published widely on many topics, but only one need be noted here: a strange, long and possibly relevant paper from 1781 upon a novel assay balance.16 This appears to be the ultimate origin of the flat rhomboid lattice beam; see Fig. 15. Magellan anticipated many of the ideas of Field and Bate, and also those of the next generation, Makins and Oertling.17 The artist has clearly had trouble with parts of the drawing, but between the text and drawing the meaning becomes mostly clear. It seems to be a scale in that all bearings are in constant contact, but ‘release’ is effected by a horizontal rotary item labelled vv, which merely drops the panstops. The paper was published in a French journal, partly because of his international outlook, and partly because his French was greatly superior to his poor English. The letters on the drawing are not at all clear on the printed page, but using a lens helps to some extent. The author has translated only the sections concerning the beam, ignoring the main sections on making precise fractional weights, and ‘how to use an assay balance.’ His ideas seem very progressive for the date. He proposed: 1) a slim, steel, flat, rhomboidal lattice beam, greatly cut away to reduce weight yet conserve strength; 2) the use of two screw-threaded individually adjustable steel pins on the beam resting on two agate planes as the centre bearings; 3) the use of two screw-threaded and individually adjustable steel pins locating in a groove in a ring acting as a suspension piece as the end bearings; 4) single wire pan hangers; 5) panstops height-adjustable by rotation. In item 1) his skeleton beam anticipated both Robinson and Makins. He made his reasoning clear, using steel instead of brass to maintain rigidity. 15 For more on this multi-faceted man see Home’s excellent biography in DNB (2004). 16 Magellan (1781), mainly 43-44 and the plate. 17 Anticipating many of the ideas of Makins c.1851, leading to the revolutionary Oertling model 12 assay balance, which changed the design of assay balances for ever; see Oliver (2020), Chapter 12, especially pp.200-202.
26 FIG. 15. THE MAGELLAN BALANCE OF 1781. In item 2) the idea of adjustable pins on agate centre planes was radical, but would not work as drawn, there being no means of locating the beam; it would have skated during loading despite the pan tables. This is the same concept as that of Sisson, above. The Sisson date is vague but is at about this time, 1781. Was this a Magellan idea developed by Sisson, or a Sisson idea lifted by Magellan? The adjustability of the centre pins was clearly for setting the sensitivity. It is quite clear from a long footnote that Magellan had an intuitive grasp of beam theory. In item 3) he anticipates the Oertling dimple and groove end planes. In item 4) he is pointing the way toward the complete single rod pan hangers by Haas & Hurter c.1793, below, but did not quite get there. The long piece uy was a thin rod with a hooked end at y. The pans were suspended from y by ‘extremely fine metal wires’ but the artist clearly had trouble drawing them. Nowhere is the metal of the rods, pans and wires specified. They are probably of brass, as this instrument is just a little too early for platinum; see Wollaston, below. In item 5) he has anticipated the adjustable panstops in the Field balance of 1826. This paper is possibly of far greater significance than it has been given credit for hitherto. It has long been regarded as simply an exercise in balance design, yet the author has found within it that Magellan refers to practical work done by him on a set of five versions of this beam. He has also created a new system of weights, making the production of small weights rapid and accurate. He quotes examples and mentions an error which was determined ‘on my balance.’ This implies that he engaged one of the fine balance-makers of the time to construct five balances to his design, but which maker is unknown; and may well have been French or English.
27 Or did he? He is known to have been a shameless borrower from others in the field of scientific instruments, and passing them off as his own. Here he says 18 I have seen this set of different beams of which the very small appears to be 8 or 10 inches in length, and the very large nearly 3 feet in English measure.’ He would know the length of the assay beam if it was his own design. No beam length is given in Fig. 15, which may imply that it really is the design for a matching set of different sizes, later noting there are five beams in the set. He says ‘this set’ but has not mentioned it earlier. There is an implication that ‘this set’ of beams was used by Mr Harris of the English Royal Mint. This and his other references to the Royal Mint open up fresh possibilities: was there some connection between Magellan and Field via the Royal Mint, and did Field see a Magellan balance there? The Mint has no knowledge of cooperation with Magellan. Or, did Makins ever stumble across a Magellan balance at the Mint? One also wonders whether Field, Kater, Makins, Robinson, Tralles and the De Graves were aware of Magellan’s paper, which is long and complex and now really needs a complete accurate translation and assessment. It may produce significant information on assay (and other) balance evolution. It is too far removed from Robinson to justify pursuit here; the author leaves it as a Known Unknown for future researchers. However, the difficulties of the artist in delineating what he saw, the fact that Magellan was not writing in his own language plus the subsequent translation problems, may make an exact understanding impossible. Finally, we should note Peta Buchanan’s views on Magellan, she having looked at him in some depth. She says:19 Magellan is described as one who spread news and ideas through his extensive circle of correspondents and as such performed a valuable role in conveying information between the Royal Society of which he was a member and the Académie Royale des Sciences, of which he was a correspondent. It is now very difficult to disentangle the ideas in his writings which were solely his, from those which he collected elsewhere and passed off as his own. With all these uncertainties it seems safest to regard the ‘Magellan’ ideas not as his own but as those which were ‘in the air’ at this date, using the term ‘Magellan’ as a useful handle and nothing more. But this begs the question: who, if not Magellan, devised that forward-looking balance design? This needs investigation. Robert Fidler (?―1819?) made a large cone-beam balance which was presented to the Royal Society in 1879, and it is in most respects a very close copy of the Ramsden, at beam length k-k 60 cm (23½″).20 This too is on permanent loan to the Science Museum, accession no. 1900-158. It was fitted, like the Ramsden, with a pair of panstops controlled independently of the release, and with steel end knives and loops. After repairs it was said to be accurate to 1/100 grain (0.65 mg). In 1976 Buchanan found it sensitive to 1 mg. Fidler made a similar balance for the Royal Institution in 1802, but at only about two-thirds the size, 44 cm (17¼″). This also had end pointers, steel knives, and a pair of panstops controlled independently of the release, but the centre planes were of steel instead of agate. One notable step forward here by Fidler was his use of roof- type suspension pieces instead of loops on steel end knives. It was used by Humphry Davy and Thomas Young, and was reputed to turn easily with one five-hundredth of a grain (0.13 mg), but no longer does so. It is currently RI balance no.14. 18 Magellan (1781), 48. 19 Buchanan (1982), 42. 20 This account is based mainly on Buchanan (1982), 129-131, 275-276, 360, and Benton (c.1940), plate 14, record 302.
28 Fidler is not well documented. A workshop was first recorded in 1800; at Wigmore Street in 1808; in the 1812 London Directory at 32 Wigmore Street; then Foley Street until 1816 when his business failed. It would appear that this was due to a lack of business acumen rather than any decline in his ability as a fine instrument-maker. Edward Nairne, FRS (1726―1806) was both Natural Philosopher and instrument- maker.21 In 1741 he was apprenticed to the optical instrument-maker Matthew Loft, whose business he eventually took over, and set up in business at 20 Cornhill. In 1744 he took his one-time apprentice Thomas Blunt into partnership, and the business traded as Nairne & Blunt up to Blunt’s death in 1822. The workshop moved to 22 Cornhill in 1791. Nairne built a very wide variety of scientific instruments and in large numbers. He achieved an international reputation as one of the foremost makers of philosophical instruments of the 18th century. These were mainly optical instruments (telescopes, microscopes and quadrants), electrical and magnetic instruments (electrometers, generating machines, batteries, magnets, dipping needles and compasses), air- pumps, thermometers, drawing instruments, and several one-off devices, as well as publishing papers on some of these topics. A few hydrostatic balances have been mentioned indirectly; none appears to have survived. Amongst his many workmen he notably employed Jesse Ramsden. His main contribution to weighing may well be only that Ramsden learnt much of his mastery from him, and that he was Henry Cavendish’s preferred instrument-maker, but not of the Cavendish balance, above. Edward Troughton, FRS (1753―1835) was trained by his elder brother John (c.1739―1807), a scientific instrument-maker who was particularly skilled in circle- dividing by hand. 22 Later John built his own dividing engine after Ramsden, and later still Edward built one too although he had also become skilled at circle dividing by hand. John and Edward entered into a formal partnership c.1788 with their workshop at 136 Fleet Street. Their primary product areas were astronomical and surveying instruments, and these were greatly in demand, generally rated as being of excellent design and construction, and unrivalled in accuracy after Ramsden’s death in 1800. After John died in 1807 Edward carried on alone until he entered into formal partnership with instrument-maker William Simms (1793―1860) in 1826. The business continued to flourish as Troughton & Simms for many years. In addition to his surveying and astronomical instruments, Edward is recorded as having made only a few balances. In the 1790s he made a large cone-beam balance for Sir George Shuckborough-Evelyn, which was essentially a larger copy of the Ramsden, with a 27″ beam instead of 24″.23 Only the drawings now exist, from which minor improvements can be seen, but not the most interesting one. This is reported in the literature as being a pair of tiny springs which attempt to steer the beam in its descent such that the knives contact the centre planes at the exact same location every time. It was reported as having a sensitivity of 1/100 of a grain (0.65 mg) when loaded with six pounds Troy (2.238 kg), i.e. one part in 3½ million. The period of swing of the balance when correctly adjusted was 40-50 seconds. The end knives were adjustable; for height at the L end and for arm length at the R end, so that the three knives could be made co-planar and the arm lengths equal. The angle of the centre knife was only 30°. The sensitivity was adjustable as in the Ramsden. 21 This account is based mainly on Schaffer’s DNB 2004 biography, and R.S. Webster’s DSB (1974) biography. 22 This account is based mainly on Buchanan (1982), 131-142, McConnell’s DNB 2004 biography, Stock (1978) and Stock (1969), 14-18. 23 Shuckborough-Evelyn, Phil. Trans. 88 (1798), 133-182. The sensitivity claimed in the paper was only one part per million. The guiding springs were presumably a failure, not being taken up by later makers.
29 At some unknown later date, possibly c.1815, Troughton made an important small cone-beam balance, now resident in the National Museum of Scotland, accession no. T.1996.48. It is said to have a beam length of only 15″, whilst the very extensive description in the Edinburgh Encyclopaedia gives it as 18″.24 The engraving, Fig. 16, is mostly self-explanatory, but observe the feature which Buchanan describes as a ‘railway arch’ fore and aft of the beam. These Troughton ‘railway arches’ will appear again and again in later balances by other makers. Possibly Troughton was inspired by the rather similar structure atop the pillar of the Ramsden balance, see Fig. 8. FIG. 16. TROUGHTON 18″ CONE-BEAM BALANCE (1-3) AND ASSAY BALANCE (4). 24 Brewster III (1830), 199-201 and Plate LI. In the first American edition, pages 199-201, but 213-215 in the second, Parker and Delaplaine edition of 1832. Kater (1831) 320-322 also gives details of this instrument.
30 Here each arch carries a small agate plane for the short steel knife on each side of the beam. The centre knife edge is again of only 30°, and the contact length upon each agate is only 0.05″, so he is clearly intent on minimum contact area throughout. The key G engages a rack-and-pinion mechanism which raises a tube inside the stout pillar. The top of the tube bears the inverted arch passing under and beyond the beam, such that the action of notches in the inverted arch is to lift the knives clear of the planes for arrestment. It is therefore an odd form of a descending arrestment frame. The beam has the advanced separate adjustments he used on the Shuckborough-Evelyn above: at the L end horizontal to make the two arm lengths equal; vertical at the R end to ensure that all three knives are coplanar. Here he is following in the footsteps of Cavendish but in a much more elegant manner; this was later taken up by other makers. The capacity of this large balance is not stated, but a weight set with a head weight of 8000 grains is recommended, indicating a capacity of some 16,700 grains, or about one kilogram. The exact nature of the end bearings is unclear from the description and is hidden by outer plates. There is a gravity bob hidden inside the beam, worked by the screw at the top; here he is following Ramsden, which he also does with panstops controlled independently of release. See the footnote 24 for more detailed descriptions. The lowest engraving in Fig. 16 shows a most interesting and advanced assay balance by Troughton, said to be for hydrostatic and general chemical weighing as well as for assaying. It has a simple rod beam, with centre and end bearings very much as above. This is the earliest balance in this examination of English design to have a descending pointer running over a curved chart fixed to the front of the pillar, plus the first conventional flag post with gravity bob and flag. In the engraving, if correct, there appears to be no relief of the centre knives, which seems strange in view of the advanced features. Technically this is therefore a fine scale, not a balance. In these two instruments Troughton has made numerous steps forward in fine balance design. In 1812, Troughton made a balance for the Swiss/American Ferdinand Hassler, who came here as First Superintendent of the Coast Survey and of Weights and Measures, purchasing from Troughton surveying equipment for the very early United States. He ordered a balance suitable for pioneering work on a national system of weights and measures, i.e. of top quality, to be based on a Troy pound which had been imported from London. In this case Troughton abandoned the cone beam; the beam here was of solid blued steel, much shorter at about 12″, ¼″ thick, and tapering from 1″ depth at the centre to ½″ at the ends. The centre knives were of pippin steel, but specially profiled to minimise the contact area with the steel (not agate!) plane, which was domed slightly to reduce the contact area still further. Hassler insisted on a glazed cabinet for it; see Chapter 22. The details of the rest of the balance were much as that made for Shuckborough-Evelyn or for one he made at about the same date (certainly before 1816) for Sir James South. South’s balance has survived.25 It features a solid mahogany beam a little over 24½″ long which turns on steel conical points bearing on two tiny agate discs, relieved by a sophisticated arrestment system. Here we have two conical points acting as a centre knife again, as suggested by Magellan, and used by Sisson. A chart is fixed to the right-hand arm of the beam, and moves over a fixed pointer, reversing the usual arrangement but to no obvious advantage. Buchanan gives much detail on three of 25 Now resident in the Science Museum, London, accession no. 1876-227.
31 Troughton’s balances, the Shuckborough-Evelyn, the Hassler and the South, but not on the Edinburgh one.26 Thus we have here three different attempts by Troughton to create a beam both rigid and lightweight: long hollow metal, short solid metal, and in long solid wood, plus his developments of several Cavendish and Ramsden ideas. According to W.A. Benton (1927―1942), founder and first curator of the Avery Historical Museum (AHM), Troughton made a large cone-beam balance very much as the Ramsden and Fidler above, but with the centre plane carried on Buchanan’s ‘railway arches’ fore and aft of the beam, as seen in Fig. 17, thus:.27 FIG. 17. TROUGHTON CONE-BEAM BALANCE RECORDED BY BENTON. Here is a very close copy of the Ramsden, but with the old horizontal release lever augmented by the new ‘railway arch’ and inverted arch of the Edinburgh Troughton. Which came first? Presumably it was Benton’s, being so much nearer to the Ramsden design with the awkward horizontal release lever and the stand of four pillars. The other introduced the rotary front release and a simple stout pillar. Although the pans, unspecified, are still in place, it is otherwise shown as set up for hydrostatic weighing, with the long legs, a glass vessel and one long suspension rod lying as yet unused. The current location of this balance is unknown. 26 Buchanan (1982), 131-142. See also Stock (1969), 18. 27 Benton (c.1940), plate 12, record 236.
32 Recently an anonymous 14″ cone beam balance in the Avery Historical Museum, accession no. 01.02.17, has been identified as another Troughton, showing clearly an evolution in his design thinking; see Figs. 18 to 21. It features his ‘railway arch’ and the broad band covering the joint of the two cones. It has a pioneering front release knob (‘milled head’ 40 mm, 1½″), as opposed to his key G in Fig. 16, but still driving a rack-and-pinion arrestment. In this model, the ‘railway arches’ carrying the tiny agate planes are as before, but the inverted arch has vanished. Here the two arches rise to engage the short (15 mm) 30° steel knives on the beam which is supported in notches in the fixed vertical front and rear arms; it is a rising plane balance, the inverse of his earlier designs. The beam carries a flag post and flag with what seems to be a gravity bob at the top, but which currently refuses to move. The bright steel end knives are very hard to see, being shrouded in sophisticated shackles developed from the simple shackles of Martin and Whitehurst, for which see Figs. 10 and 12. [© ANDREW P.B. LOUND. FIG. 18. THE AVERY MUSEUM TROUGHTON 01.02.17. [© ANDREW P.B. LOUND. FIG. 19. THE TROUGHTON 01.02.17 BEAM.
33 [BOTH: © ANDREW P.B. LOUND. FIG. 20. THE ‘RAILWAY ARCH’, BEAM, CG BOB AND FLAG . FIG. 21. THE ‘RAILWAY ARCH’ AND FIXED ARMS. There is no pointer; instead the front face of the R shackle carries a fine horizontal line which aligns perfectly with the mid-point of the vertical ivory chart on the end of the long rod arm when the beam is truly horizontal. Edward Troughton was one of the first instrument-makers to make serious use of the platinum made available by Smithson Tennant and William Hyde Wollaston (see Wollaston below) from 1804/5 onwards. He used it mainly to replace the corrodible silver then used on graduated angular scales in sextants, mural circles and the like, but not on any of his balances.28 Unfortunately, the pans on this one are non- magnetic steel replacements for the lost originals, possibly of platinum. They are 42 mm in diameter, held on green plaited silk cords. They rest on stylish pan tables, which bear a triangle of metal points to locate and hold the pans. That is a nice touch, but the tables are not easily height-adjustable, rather spoiling the design. Instead each pan table is held on the baseboard by a push fit small wooden peg. A weight set from 2000 grains down, in the old 4-3-2-1 grains system, is still present but the ‘fractions’ are missing. These were presumably in the empty R drawer. More will be said on the cabinet and on the weight set in chapters 22 and 23. The overriding impression from the above examination of Troughton’s known balances is that he has had a far greater influence on fine balance design than anyone has suspected to date. It is unfortunate that, like so many others, he followed Ramsden’s cone-beam construction for so long. Jacob Bernard Haas (1753―1828) and Johann Heinrich Hurter (1734―1799) were in business together as ‘a manufactory of philosophical instruments’ in London from 1782 until 1795 when the partnership was dissolved. How the partnership began is unclear, but Hurter was the principal partner and Haas was the instrument- maker. Hurter was a society painter who had European trade contacts very useful to the business. The partnership was noted especially for its superior air pumps and portable barometers. One very important small precision balance signed Haas & Hurter, London is in the van Marum collection in Teylers Museum, Haarlem, but no useful image is currently available. It was a very advanced design for its time, and is probably the one ordered by van Marum in 1789. 29 In summary this balance is very progressive with its flat, open, rhomboid 300 mm (12″) beam with a large circular central feature carrying a steel centre knife, defying the then current conical beam convention for high precision work. The arrestment has the essence of a descending 28 Chaldecott (1987). 29 See Bryden (2025) and Bull. SIS 100 (2009), 32 for more biographical detail, for data on the Teylers Museum balance and data on the sale of the stock in trade in 1795. A Haas case-suspended fine scale in the NMS seems irrelevant to this research.
34 arrestment frame with cam and camshaft construction. On the front of the shaft is a knurled rotary release knob; this seems to precede the Troughton balance above as the first recorded occurrence of the conventional ‘milled head’. The centre plane is polished steel atop a pillar strangely constructed of two brass plates. It has easy- access 40 mm pans on single slim rod hangers, and a thoroughly protective case of an unusual construction. The hook-and-eye end ‘knives’ were a backward step. After 1795 Haas ran his own instrument-making business from No.4 Silver Street, Golden Square, from which an undated catalogue survives. There are two lines of particular interest, implying the routine production of fine balances: Hydrostatic balances, OF A PECULIAR CONSTRUCTION WITH Apparatus adjusted with great accuracy, from £10-0 to £20-0. Assaying balance, peculiarly constructed, which turn with the addition of a 200th part of a grain, when charged with an ounce in each scale, fixed in a glass-case; from £10-0 to £20-0. Haas sold an even more sensitive balance to Matthew Boulton, of which more later. Haas moved to Lisbon in 1800 at the request of the Portuguese government. William Jones (1762―1831) and Samuel Jones (1770―1859) were sons of the instrument-maker John Jones.30 William was apprenticed to John in 1776. When John retired in 1791 the two sons took over the business together as W&S Jones at 135 Holborn, moving to 30 Holborn in 1798. William was forced by ill health to retire early, dying in 1831. Samuel thereafter ran the business himself, but continued trading as ‘W&S Jones.’ This had been their house style from 1791 and continued to be so until 1859, when the company ceased trading at the death of Samuel. The company was one of a very few pioneers in printing priced catalogues of their wares, which enabled them to introduce what was effectively a mail order service for instruments.31 A total of twenty-five such catalogues were issued between 1784 and 1855. These were quite substantial, with fourteen pages on average, plus a couple of pages on books for sale. From just under 300 instruments advertised at the start, by 1855 this had risen to over 500. According to the catalogues W&S Jones made a very wide range of philosophical apparatus. From 1814 up to 1830 they made a range of hydrostatic balances from 2½ guineas up to 9 guineas, and a range of cone beam balances from 12 guineas to 50 guineas. By 1838 the 9 guinea hydrostatic was now 10 guineas. In the 1843 and 1850 catalogues both the plain and conical balances are listed under a new section entitled ‘Hydro- statics.’ More detail was given: conical balances from 30 guineas to 50 guineas came complete with agate planes, plus platinum pans and weights. 3 [P.D. BUCHANAN. 3FIG. 22. W&S JONES CONE BEAM BALANCE. 30 This is based largely on the Morrison-Low biography of William in DNB 2004, which includes Samuel and W&S Jones, and on the W&S Jones section of Morrison-Low (2007). 31 Following closely on the heels of Benjamin Martin and George Adams in issuing what we would now call ‘trade catalogues.’ It is interesting to note that John Jones and Son bought much of the books and prints offered at the auctions following the ending of Benjamin Martin’s business in 1787. Such works must surely have influenced their design thinking to some extent.
35 The cone beam was essentially obsolete by c.1815, but W&S Jones continued making them up to 1850, so there must have been continuing demand despite the very high prices. They were approximately half the size of the Ramsden, and incorporated some of Troughton’s amendments. Very few of W&S Jones’ large output have survived, but there is one of their cone- beam balances in the Science Museum, accession no. 1928-1024 (Fig. 22), having a beam length 32.5 cm (12¾″) k-k, end pointers and a steel centre knife working on an agate plane, the latter held by a Troughton-style ‘railway arch.’ The steel end knives support captive hooks. The arrestment is inferior in that the rising plane elevates the beam but not the end charts, which have to be manually adjusted to make the pointers align with the zeroes on the charts when the unloaded beam is fully operational. Presumably this messy adjustment is rarely if ever needed once set. Panstops are present and operated separately from the beam by ugly visible pivots in a manner not repeated by other makers. The case has three adjustable feet and two spirit levels. Does this sole survivor represent the many W&S Jones cone-beams? John Miller (?―1825) and Alexander James Adie (1775―1858) need to be mentioned briefly here, although working in Edinburgh. They built a large heavy- duty balance with a steel rhomboid beam carrying a heavy steel knife bearing on a steel plane, somewhere in the period 1804 to 1825. The beam length was not specified, but the capacity was 650 g. See the final paragraph and engraving at the end of Chapter 20 (Adie). This may have influenced Kater in his initial thoughts for a hydrostatic balance, particularly in the use of steel for a very open rhomboid beam. John Field, (1764―1843) was apprenticed to Nairne in 1779. He was made free of the Spectacle-Makers Company in 1787 before setting up his own workshop at 74 Cornhill in 1787/8.32 He sold it in 1795 and joined the family business, specialising as a ‘Russia merchant’ (fur trader). He joined the Royal Mint in 1804, and in 1815 became Weigher and Teller to the Mint, and additionally as Stamper of Weights in 1821.33 He therefore followed Whitehurst in those Mint offices, with one person of lesser importance between them. He reports in his examination by the Government select committee on the Royal Mint that he also began the care and adjustment of the Mint balances on the basis of his ‘having for some years applied myself to mechanics, as an amusement’ but, curiously, he did not mention his instrument-making business. ‘As an amusement’ possibly means keeping his hand in whilst being a Russia merchant. In 1818 he introduced an ‘improved balance’ also described as ‘a beam on a new principle’ ― unspecified ― for the routine weighing of gold ingots. This was of far greater reproducibility than the old beams, and saved much time. It was clearly stated that he had no assistance in making and repairing the ‘assay balances’ but this seems to be in the context of the heavy beams for weighing gold ingots. (The weighing terminology throughout is distinctly imprecise, and the committee report seems to be confusing the delicate ‘assay balances’ with the large balances for bullion weighing.) It appears to confirm that he alone was making the balances/scales, large and small, but it is entirely unclear as to whether he merely designed or partly made balances. At one point in respect of a new beam and a counting machine he says: ‘I expended a great deal of money in both, and making a full sized model of the beam, and money to workmen to expedite its completion’. 32 Bull. SIS 100, 31; Buchanan (1982), 58-64, 101. See Waylen (1897), 74-75 and Clifton (1995) for two very different biographies of Field. He was of the House of Cromwell. 33 For the minutes of the select committee 4th May 1837, see Buchanan (1982), 434-438.
36 This is all very unsatisfactory and unclear. It seems best to ignore all of this select committee testimony in view of the confused reporting. He retired in 1842, only one year before his death. The Whitehurst balances which Field inherited were by his time becoming unreliable. That Field made fine balances for the Mint, following Whitehurst, is known for certain because two survive, which now follow. The John Field Assay Balance is of unknown date but somewhere in the range 1818 to 1826.34 See Figs. 23, 24 and 25. This is a curious, small, lightweight instrument, clearly intended as an assay balance. It contains some very basic and simple scale-maker’s features plus several forward thinking ideas pointing towards his magnificent creation of 1826. Therefore this simple rising plane balance is presumed to be pre-1826. [J. WATERS. FIG. 23. JOHN FIELD ASSAY BALANCE. (Doorslide removed for clarity.) It might be thought to be one of those he made earlier ‘as an amusement’ were it not for the fact that it bears an integral brass plate engraved in a rather odd manner thus: J. FIELD FECIT His Majesties Mint . It also carries a wide ivory chart, engraved 25-0-25, which rises and falls with the planes, keeping the pointer and chart in their correct relationship. The chart is of the anti-parallax type, but has had the back-screen broken off. According to Buchanan, 5 mg gives full-scale deflection, implying a sensitivity of 0.2 mg/division. When on release, the contact line of the long central knife and the tiny planes is very small, but as this is an assay balance, the low masses involved will induce little wear. The release is by the vertical lever at the front of the case, and is simply a primitive 34 Science Museum accession no.1925-40. They date it to 1818 ‘or very near.’
37 spring-loaded see-saw device, which has the most unfortunate effect that the operator has to keep the lever depressed whilst weighing. The lever has lost its ivory or wood knob. [J. WATERS. FIG. 24. CLOSE-UP, ON RELEASE. (Note also the platinum chains.) [J. WATERS. FIG. 25. ARRESTED. (The Field engraved inscription is losing its black filling and is now rather faint.) The thin brass pans are each suspended on three platinum chains, each chain linked to the pan via a small brass ring passing through a hole drilled in the pan, very close to the periphery. This was the standard mode of platinum chain suspension from this point onwards. Platinum wire had been perfected by Wollaston by this date, as shown in his section below. Note there is also an extra tiny hole beside each currently used hole in each pan, suggestive of an earlier suspension on fine silk cords. The fine mahogany case has front and rear doorslides, and two solid side doors which are concave on the inside (why?). Note the very unusual pan tables which are adjusted for their height by three screws. A similar three-screw adjustment of pan tables was noted above in the Whitehurst balance in Leiden; see Fig. 11.
38 The John Field Balance of 1826 resides in the Royal Mint Museum, Llantrisant. It has no accession number. See Figs. 26 and 27. 35 The stand consists of a cast brass bridge across two widely spaced brass pillars, within each of which is a push rod for the arrestment frame. This is a complex and sophisticated balance, and very advanced in most respects for its date, and needs detailed examination. It is coeval with Robinson and Kater’s work and far superior. The capacity is unknown, but the use of a solid beam and heavy chains for the pans suggest a capacity of at least 250 g, probably 500 g, maybe as high as a kilogram. Briefly, the tapered 16½″ (422 mm) steel beam of lozenge cross-section carries three knives of pippin steel, all fully relieved. The 2¼″ (57 mm) centre knife rests on a divided centre plane of steel carried on the brass bridge. The descending arrestment is notably important, having a small rotary knob at the front which releases both the arrestment frame and the panstops simultaneously. The stroke (drop) is remarkably large at 27 mm. The frame carries two tall V-pieces which engage each end of the centre knife, plus two curled features at the ends of the frame proper, which engage the beam with a firm grip just over half way along each arm. There is therefore a four point arrestment, which should give a perfect reproducible placement of the beam on the centre plane. Each arm of the arrestment frame carries a brass bolt-on extension which engages the end suspension pieces, as shown in Fig. 26, giving full relief of the steel end planes too. It is unclear whether this feature is an original option, or an added afterthought.36 When so fitted, the placement of the three knives on their planes is optimal and far better than anything created by Makins or Oertling at their best. This follows Cavendish’s thinking instead, although with a different mechanical solution to that of Cavendish’s balance. The nominal separation of the two pillars is five inches, allowing a wide ivory chart to be fitted, divided 22 ― 0 ― 22. This is developing Whitehurst’s wide chart, then Field’s own assay balance chart, to give for the first time truly easily readable central charts, allowing the ‘method of counting swings’ for speeding up weighings to be used. The concept of a wide chart with well-spaced graduations was highly influential, being copied by many later makers, both British and European. The beam is fitted with a flagpost bearing a gravity bob and flag, and there is an ultrafine zero adjustment on the steel pointer. The panstops are height-adjustable by rotation and fitted with locking rings, and even carry pan-brakes to reduce swinging, actuated by the vertical lever at the centre front, between the drawers. The locking rings and panbrakes were not copied by later makers. It is built like a research-grade balance and is in the author’s opinion of a more complex and better design and construction than any of its predecessors above, including the much more famous Ramsden of 1788. The mahogany (?) case is beautifully made, with a pitched roof and a sash front doorslide counterweighted behind covers in the rear corners. It is most regrettable that the underworks are completely boxed in and totally hidden, as it hides a sophisticated central locking system set by a single slider just inside the doorslide, itself lockable with a key. Pages could be written on the multiple refinements present in this instrument. On current evidence several developments in the design of fine balances occur here first, and there are a few points of this balance which are not yet fully understood; it needs thorough re-examination. 35 The metric weight set by Becker’s Sons Rotterdam currently associated with this balance is not the original. 36 The end knives are not readily measurable with the suspension pieces in place and, considering all the refinements, it seems odd that they are seemingly not adjustable for arm length and parallel. This needs to be checked.
39 [FIGS. 26 AND 27 BOTH BY COURTESY OF THE ROYAL MINT MUSEUM. FIG. 26. THE JOHN FIELD BALANCE OF 1826. FIG. 27. THE OVAL CARTOUCHE ON THE FIELD BALANCE. (‘Invented and Made for His Majesty’s Mint by John Field, Stamper of Weights, 1826.’)
40 To the best of the author’s knowledge nothing has been published before on this balance. The only negative points which can be raised are the use of steel instead of agate for the planes, the apparent lack of adjustment of the arm lengths, and the lack of adjustable feet, although the latter may simply have been lost. Field’s two surviving balances were clear influences on later designers, most notably upon Makins who lifted several of Field’s ideas and incorporated them in his very influential assay balance design for Ludwig Oertling in 1851.37 De Grave should be mentioned briefly here, at the turn of the 19th century, as highly respected makers of assay balances of a very traditional form and high quality. They also made simple coin, diamond and apothecary scales, Inspectors’ beams and local Standards. The complex history of the De Grave company has been covered else- where in so far as it is currently possible.38 At around this period the principals were Charles de Grave I and II, Mary de Grave and Edward Abraham de Grave. Their delicate assay balances featured a slender steel beam, usually a rod, with swan- neck ends, rather in the manner of the rod beam in the Field assay balance above. No image of these is readily available, but they probably differed little from the assay balances in their later De Grave, Short and Fanner (DGS & F) period 1845―1871. Each of these archaic designs is in a tall glazed wood case to accommodate its upstanding needle pointer. There is a fine example in the Science Museum and another in the Avery Museum.39 The former has an elegantly tapered 10½″ solid steel beam of a lozenge cross-section. The latter has a slender steel rod beam 10″ long plus a great deal of decorative brasswork on the stand and arrestment frame. Each has a relieved centre bearing and a semi-relieved end bearing with crudely adjustable pan tables. From the earliest times and up to and including the DGS & F era, the company enjoyed a high reputation as the English maker of Reference and assay balances and Inspectors’ beamscales. Before Ludwig Oertling arrived on the scene De Grave was therefore the maker of choice. They began to lose their way against Oertling and foreign competition shortly after the 1851 Great Exhibition. William Hyde Wollaston, FRS (1766―1828) was a Natural Philosopher and not an instrument-maker. His interests covered a wide field, medicine, botany and chemistry, but he is best known for his success in preparing the recently-discovered (1735) metal platinum in a form sufficiently pure as to make it malleable.40 He was educated at Charterhouse, then Gonville and Caius, Cambridge where he studied mainly medicine, with some astronomy, botany and chemistry. Achieving MB (1788) and MD (1793) he practised medicine until 1801. But whilst at Cambridge he met the chemist Smithson Tennant, and in 1800 they set up a business partnership in London intent on making platinum sufficiently malleable as to be industrially useful. They had eventually perfected a method by 1804/1805, such that by 1815 they had purified some 1½ tons of the crude metal, and gained a fortune thereby. This work was conducted in great secrecy, for which Wollaston was heavily criticised, as being contrary to the spirit of open scientific research. The whole process was eventually revealed in full detail in his Bakerian lecture to the Royal Society shortly 37 Oliver (2020), 183-189. 38 Oliver (2020), Chapter 31. 39 Accession numbers: Science Museum 1925-41 and Avery Historical Museum 01.02.29. 40 Chaldecott (1987) emphasises the pioneering work of Troughton and Wollaston in making practical use of platinum in instrument-making.
41 before his death in December 1828. 41 It needs to be pointed out that their process was not quite as original as it is usually presented, for they were building on the critical step devised in 1800 by the forgotten chemist Richard Knight (1768―1844).42 They simply paid much more attention than Knight did to vital details. The first commercial uses for the pure platinum were for the critical parts of flintlock pistols and for small items of laboratory ware. Larger items for the chemical industry later became possible as their experience grew. Their project also became a research success in that they discovered four more of the platinum metals: Tennant claimed osmium and iridium, and Wollaston palladium and rhodium. Smithson Tennant had visited the Swedish chemist J.G. Gahn, who was adept at chemical analysis on a small scale. He seems to have transferred this ability to Wollaston, who greatly surprised Berzelius with his small scale analytical manipulations. This ability was essential to both of them in their separation and characterisation of the other platinum metals. Thus Wollaston was now also a philosophical success as well as a wealthy industrialist. He went on to perform original research in crystallography, optics and physiology. He was awarded his FRS in 1793, and became Secretary of the Royal Society 1804―1816. There are two aspects of Wollaston’s work of particular relevance to high-precision laboratory balances including those of Robinson, but not widely acknowledged. First, he perfected an existing extrusion process capable of drawing very fine gold wires into one for drawing platinum wires down to as low as 0.0001″ in diameter, which is one tenth of one thou.43 This meant that balance pans could thereafter be suspended on platinum chains created from platinum wires, replacing the trouble- some traditional silk cords. Such extremely fine diameters were not needed in this application, but balance-makers were soon suspending their pans on 6-thou. or 8- thou. platinum chains, such as W&S Jones, John Field and T.C. Robinson here in London and even K.P.H. Pistor in Berlin. Robinson and other balance-makers also began to make weights out of platinum, as being much more resistant to corrosion in chemical laboratories; likewise making balance pans of platinum in the more expensive models for the same reason. Some makers plated brass or copper pans with platinum for a cheaper mode of reducing chemical corrosion of pans in chemical labs. Such corrosion was becoming a real problem because serious chemical research was expanding rapidly in Britain and Europe as the 19th century progressed. Secondly, the so-called ‘Wollaston pan stabilisers’, first appearing in the Cavendish balance, and very much a feature of Robinson balances, may not have been invented by Wollaston at all. William Hyde Wollaston was too young to have devised these when the Cavendish balance was being built in 177x. Buchanan has suggested that the name refers to his father, the Reverend Francis Wollaston, FRS (1731―1815), who was a contemporary of Cavendish and deeply interested in science.44 There can be no doubt that Kater (see below) was aware of the ‘Wollaston pan stabilisers’ via their correspondence. 41 See Williams’ excellent DNB biography: Williams (2004). See also Chem. Brit. 2 (1966), 525-527 and Chem. World 2 (2005), 50-53. Wollaston’s own very detailed Bakerian lecture account is to be found in Wollaston (1829). There is a very fulsome and informative memoir of him in The Annual Register for 1829, 521-526. 42 See Hunt and Buchanan (1984). Knight was primarily a retailer of scientific instruments and chemicals. 43 Wollaston (1813). One tenth of a thousandth of an inch. Such very fine wires were used as crosshairs in telescope eyepieces, replacing traditional spider thread. 44 Buchanan (1982), 122, 215.
42 Thomas Jones, FRS (1775―1852) was apprenticed 1789―1796 to Jesse Ramsden, who considered Jones to be ‘the most ingenious and the most assiduous’ of his many apprentices.45 In his own workshop from 1803 onwards, he became an important instrument-maker who described himself as ‘a pupil of the late Mr Ramsden’ and he was one of the makers favoured by Captain Kater for pendulums and other instruments; see Kater below. He was also favoured by Edward Troughton who, when he later declined commissions on the grounds of old age, recommended Jones instead. After Thomas’ death the business was continued by his son, also Thomas, until 1861. Their workshop addresses were: 13 Panton Street, Haymarket, 1803-1806 (very uncertain); Mount Street, Berkeley Square, 1806-1811; 21 Oxendon Street, 1811-1816; 62 Charing Cross Road, 1816-1850; 4 Rupert Street, 1850-1861. Thomas Jones made many types of scientific instruments, principally those needing graduated scales, such as mural circles, transit telescopes, theodolites and various magnetic and meteorological instruments. He is known to have made at least one cone-beam balance, now in the Mineralogy Department, British Museum (Natural History) as ‘No.5.’ The beam length k-k is 37 cm (14½″). The centre plane is of agate, and this rising plane balance has the refinement that the ivory chart rises as the centre plane rises, keeping the pointer and chart in their correct spatial correlation, unlike the W&S Jones balance noted above. Was this copied from Sisson? From the 21 Oxendon Street address on the trade card in the drawer it can be dated to 1811―1816. The case carries two spirit levels and four adjustable feet, plus a second drawer at the rear fitted for the beam. A 36.5 cm plain, solid, brass beam by the same maker is also present, engraved ‘Thomas Jones & Son 62 Charing Cross Road’ thereby dating it to not earlier than 1816. Captain Henry Kater, FRS (1777―1835) was a Natural Philosopher and not an instrument-maker. We need to consider Kater in some depth because of his enormous influence on many English and European instrument-makers, including Robinson; indeed it may fairly be said that he became in effect an instrument designer as a natural part of his extensive researches. He had an enormous influence on British metrology in the 19th century, as will become apparent later.46 As a young army officer he was deeply involved in the Everest survey in India 1799 ― 1808. This is where he revealed a talent for care and precision in the use and maintenance of surveying and other scientific instruments. These were important abilities when using instruments thousands of miles from the instrument-maker’s workshop. He also developed an aptitude for devising improvements to instruments, an ability which stood him in good stead in his subsequent career. He returned from India in 1808 on health grounds and retired from the army in 1814, the same year as he was awarded his Fellowship of the Royal Society. In 1818 he was appointed to a Royal Commission alongside Wollaston and other Royal Society members for the creation of a new system of Weights and Measures for Britain. This led ultimately to his critical work in implementing the 1824 Act entitled ‘An Act for ascertaining and establishing Uniformity of Weights and Measures’ 5 Geo. IV c. 74, establishing the ‘Imperial’ standards of mass, length and volume. The practical work on the construction and adjustment of the masses and the volumetric 45 Based on McConnell’s DNB 2004 biography and two brief sections on Thomas Jones in Buchanan (1982), 134 and 426. 46 This brief account is based mainly on Holland’s DNB biography, Buchanan (1982), 208-210, and Kater & Lardner (1831).
43 measures was carried out by Robert Bate under the guidance of Kater; see the Bate section below. Inter alia this involved the building of the ‘Kater’ alias ‘Bushel’ balance for weighing the prototype bushel measures. The construction and the weighing operations were reported in a major paper.47 Kater subsequently became a frequent consultant to the Government on various metrological matters. Kater researched in many areas, publishing on astronomy, magnetism and surveying as well as several aspects of metrology. One of his particular interests was the creation and use of reversible pendulums beating seconds very precisely, first for astronomical timekeeping and secondly for the determination of the exact shape of the Earth. He devised numerous instruments in all these areas, employing makers such as Thomas Jones, Edward Troughton and T.C. Robinson to bring his ideas into reality. His experiences on the creation of knives and planes for the pendulums must have been highly relevant to the design of knives and planes for the Robinson balances. It should be noted that Whitehurst before him was also working on measurement of mass, length and time, and the true length of pendulums.48 The Royal Society was the focus of much of his research. It is important to note that he had a gift for making friends, both scientific and literary. Not only was Kater acquainted with many of the fine instrument-makers of his time, he also knew many of the Natural Philosophers too, and had a close friendship with William Hyde Wollaston. He was therefore very much at the centre of British scientific research for many years, before, during and after his collaboration with Robinson in the matter of hydrostatic and chemical balances which are at the heart of this book. As well as research papers on these topics Kater also wrote an important treatise on balances and pendulums.49 Kater was a prolific letter writer, and a great archive of letters to and from the leading Natural Philosophers and instrument-makers of his time accumulated. Alas, this archive has become greatly dispersed and much has been lost. He was in correspondence with Berzelius, Herschel, Lardner, Moll, Schumacher, Troughton, Wollaston and others. He also corresponded with leading social reformers and literati of the day, such as Elizabeth Fry and Maria Edgeworth. To return to the great Bushel balance, although Kater reported his weighings on it of the various masses and measures in meticulous detail, and some details of the beam, he says almost nothing on the balances used on the lesser masses; indeed it is not entirely clear how many different balances were involved. His first reported weighings were comparing copies of Troy pounds, made by Bate, on a 10″ balance made by Robinson; it had agate planes and knives, with the ultimate angle of the centre knife set at 120°. With an avoirdupois pound in each pan, its reported sensitivity was 0.01 grain (0.6 mg).50 The Troy pound is 5760 grains, the avoirdupois 7000 grains, but Robinson’s largest commercial balance was his 10½″ model of 2000 grains capacity. So what was this balance, and was 10″ an error for 10½″? With a capacity of some 2½ times his highest capacity 10½″ production balance this must have been a larger balance; for now we will call it Robinson/Kater A. Later this was replaced by another Robinson balance which Robinson considered as being ‘superior’ and which was also sensitive to 0.01 grain. No other details were given; for 47 Kater (1826). 48 Whitehurst (1787). 49 Kater (1831), the long chapter 21 only. Full of sound advice on balances on 320-329, there is an unfortunate confusion on 324; where having described a Robinson ‘10″’ he says ‘A balance of this construction’ was used to weigh a Troy pound. 50 Kater (1826), 11.
44 the present we will call this one Robinson/Kater B.51 Later in the proceedings, Bate prepared a balance strong enough to weigh gallon weights, which was delivered to Kater’s house on 16th March 1825, but nothing further is said about it.52 For now we will call this one Bate/Kater C. Kater later states that for comparisons of the avoirdupois pounds ‘Mr. ROBINSON’S balance’ was used, but which one? Presumably the ‘superior’ one, Kater B, but it is left unclear.53 Later in this section Kater suddenly reports:54 After the adjustment of the avoirdupois pounds was completed, Mr. BATE used the large balance to make an approximate adjustment of the weights of the imperial gallon of distilled water. This was interrupted by an accident which rendered it necessary to repair the knife edges of the beam; and notwithstanding every care, the balance afterwards proved less certain in its indications than it was before the accident. This is the first specific mention of a ‘large’ balance and he is obviously referring to Bate/Kater C, but it should have been stated more clearly. All three, A, B and C, must have been ‘large’ balances in view of their elevated capacities. Omissions of this type are frequently found in scientific reports involving weight determinations, where the masses and the procedures are specified in great detail, but the balance is barely mentioned. The reverse also applies, where trials of balance performance frequently assume that the masses are all perfect, or at least appear to be so. Robert Brettell Bate, (1782―1847), usually known nowadays as R.B. Bate, has a biography by McConnell, and is also covered elsewhere.55 His workshop produced a very wide range of scientific instruments, of which balances were a very minor part. He is probably best known for his prototyping of the ‘Imperial’ volumetric standards. This immensely demanding prototyping work was conducted under the supervision of Captain Kater (see above) throughout 1824 and 1825, as recounted in outline by McConnell, and in full technical detail by Kater in his 1826 paper. As to why the new Imperial standards of mass, length and volume were required, and for their subsequent renewal in 1834, see reference.56 It was all clearly intended to be at the very limit of the then-possible in both precision engineering and in weighing. Bate even built a double-glazed and air-conditioned balance room for the weighings, perhaps the very first temperature-controlled and air-conditioned balance room. 57 Several sets of Model Standard masses and measures were to be made, and multiple problems had to be solved. It was already known that the corrosive atmosphere of London was harmful to brass, the obvious material of choice at that date for both the masses and the measures. Bate devised a more resistant alloy, consisting of 576 parts of copper, 59 of tin and 48 of zinc, presumably classifiable as a bronze. Whilst various problems were discovered at each level of mass or volume, the most notable problems encountered were on the Standard Bushel Measure and the huge balance upon which to weigh the bushel measure when full of water. Recall, one bushel is eight gallons; consequently the specification for the beam of this balance was truly formidable; it had to be capable of swinging 250 lbs per side, yet resolving to one grain or better at that load. Even the redoubtable Kater was at first stuck for a solution to this challenging design problem. Eventually, having rejected 51 Kater (1826), 14. 52 Kater (1826), 19. 53 Kater (1826), 23. Kater (1831), 324 has the same confusion; a Robinson 10″ balance used to weigh a Troy pound; ref. 49. 54 Kater (1826), 26. 55 See McConnell (1993) and McConnell (2004a) for the biography; also Oliver (2020), 12-18 and Buchanan (1982), 142-146. 56 Connor (1987), 249-278. 57 McConnell (1993), 19-28.
45 cast iron, he came up with the notion of a mahogany beam, as the only material he could think of with the relevant strength and rigidity, yet light enough to give the required sensitivity. Possibly he was guided to this by Troughton’s mahogany beam made for Sir James South a decade earlier; see Troughton above. There are a few helpful notes by Kater on the construction of the beam, knives and suspension pieces in his paper, but not much more.58 Bate had to build this, venturing into completely uncharted territory. Indeed, it seems unlikely that Bate had ever built any balance at all before this date. He was selected by Kater because, as an accomplished maker of hydrometers he was, as Kater put it, ‘accustomed to nice operations in weighing.’ Fortunately Bate had the clear-sighted Kater to guide him in the construction of what ultimately became known as the Kater or Bushel Balance. What all the above shows is that Bate was capable of working to the very highest standards, and had a thorough understanding of the problems involved in achieving truly precise and accurate weighings. The beam was 68″ k-k, 72″ overall, 2¼″ thick, a huge size. The literature is confused as to whether it was a 250 or 300 lb balance; it had to swing 250 lbs weighing the full bushel, so presumably built to 300 lb capacity to be safe. The Kater was a research balance; in modern terminology it was a Comparator Balance, designed to determine very accurately small differences between two very similar large masses. Kater reported that he was very pleased with it, saying:59 The performance of this balance fully equalled my expectations. With two hundred and fifty pounds in each scale [pan], the addition of a single grain occasioned an immediate variation in the index of one-twentieth of an inch, the radius being fifty inches. The Bushel Balance is now resident in the Museum of London, accession number NN7347. Presumably the success of the great balance suggested to Bate that the making of bullion balances would be a natural extension to his repertoire. As early as February 1825, in his circular letter advertising the new Local Standard weights and measures, he added:60 I am also about to make Scale Beams on an entirely new construction, and shall be happy to supply you with such as may be required… On his trade card he again referred to this ‘entirely new construction’ but clearly envisages it applied not only to Weights & Measures work but also to the following, although not specifically mentioning bullion weighing: 61 Scientific Experiments… or any other purpose requiring extreme accuracy. Quite apart from its inspirational effect on Bate, it has also functioned additionally as the prototype of all Oertling bullion balances weighing up to 56 lbs thereafter, from 48″ down to 20″, with other makers also replacing the wooden beam with one of gunmetal or bronze. Thus bullion balances were definitely in production in Bate’s workshop in his later years, and to date Bate is not known to have made any bullion balances with cast gunmetal beams, but only with wooden ones, of which only four are known to have survived.62 There is only one known surviving chemical balance by him, the ‘Rothamsted Bate,’ which is of great importance in the understanding of the development of chemical balances in general, and those of Oertling in particular. This balance is still in its 58 Kater (1826), 35-38. 59 Kater (1826), 37. 60 EQM (1998), 2250. [Annex to Buchanan (1998).] 61 McConnell (1993), 38, Fig. 10. 62 See table in Oliver (2020), 18.
46 original home at Rothamsted Research, near Harpenden, inventory number BA40. Bate was not a healthy man, suffering from both chronic bronchitis and occasional mental disturbances.63 Possibly this balance had been in the making when Bate had one of his frequent bouts of illness, and Oertling with his Berlin experience was the man best fitted to finish it. In reality foreman Ladd would have assigned jobs in the workshop, and he gave this one to Oertling, who took the opportunity to show what he could do; his sections of the balance are clearly far superior to Bate’s. Although this event was in the 1840s, a few years after Robinson’s death, Bate’s basic design was in the traditional English form; see Figs. 28 and 29. One of the design features of this balance, Bate’s sleeving of the pillar in a Robinson style, had a very long-term effect, reappearing fully perfected in Ludwig Oertling’s classic long and short beam models years later. But what is most important here is the design of the stand made and signed by Bate. First, he used the Troughton ‘railway arch’ to support his centre plane. More importantly the pointer swung in a very short arc over a very short chart contained within the crude cut-out in the pillar. The traditional English style of fine balance construction, as opposed to traditional English scale construction, seems to have the pointer always working in a short arc contained within the stand. The final and most elegant realisation of this configuration is the Robinson Tulip. This is the first balance in this chronological chapter to feature an original riderslide. This was definitely by Oertling, installed in the case at its construction and was not a retrofit. The rider graduations on the beam were poorly executed and were most probably made by Bate for a manually placed rider, to which Oertling added the riderslide in the manner he had learned in his brother’s workshop in Berlin. [BOTH PHOTOGRAPHS BY COURTESY OF ROTHAMSTED RESEARCH / LAWES AGRICULTURAL TRUST. FIG. 28. THE ROTHAMSTED BATE, BA40. FIG. 29. THE STAND AND CUT-OUT. 63 McConnell (1993), 5.
47 A note must be made of Kater’s commendation of Bate’s work in all of the phases of their collaboration on creating and adjusting the Imperial standards. Kater gives full credit to Bate for designing and making the arrestment and the gravity bob to adjust the sensitivity of the Bushel Balance in particular, and more generally thus:64 I cannot conclude without bearing testimony to the unwearied perseverance, ability and accuracy, which Mr. BATE has shown in the course of a work attended with no common difficulties, and to the perfect execution of which he has devoted, for a long period, the whole of his time and attention. John Newman, (1783 [bap.]―1860) is best known as the philosophical instrument- maker to the Royal Institution (RI).65 He was apprenticed in 1799 to Philip Brock, a little-known mathematical instrument-maker. After freedom in 1807 Newman set up his own business at 11 Windmill Row, Camberwell, and from 1809 as a ‘chemical and philosophical instrument-maker’ at No.8 Old Lisle Street, Leicester Square until 1827. Thereafter his workshop was at 122 Piccadilly up to 1859. Circa 1809 he began to supply the RI with chemicals and instruments, and later began acting as a lab assistant there, not only cleaning and repairing the instruments, but even preparing the lecture theatre for Davy and Brande. He continued supplying instruments to various philosophical customers as well as the RI for many years, seemingly mostly one-off specials and others in small numbers. His surviving instruments are described as being of ingenious and imaginative design, and of first- rate quality. He won a Council medal at the 1851 Great Exhibition for an air pump, and is known to have made a few hydrostatic ‘balances’ of which two have survived. The first is an uncased, simple, rising plane scale with a box-end 10″ beam. Buchanan noted it as MHSO No.324. The beam has a single end pointer moving over a long vertical chart. The pillar consists of two concentric tubes emerging from a heavy circular brass base. Only the presence of a sleeved pillar makes it of any possible relevance to Robinson. 66 The second is a strange, complex and utterly incomprehensible instrument. It appears to be a drawer-mounted but unfinished experimental balance with the drawer full of parts, but nobody has been able to make sense of it all. Some of these probably do not belong to this balance, whilst others are missing, a chart for instance. It has a Robinson Tulip as part of the stand, and a single demountable Robinson cradle. Being made at the Piccadilly address it is too late to have had any influence on Robinson; instead Newman is simply copying Robinson’s designs. There is nothing in the way of a chart within the Tulip; this and many other features strongly suggest it was never completed. What is obvious is the high quality of execution and finish to each of these mixed parts. The baseplate of the stand is elegantly engraved in three lines: J. Newman / 122 Piccadilly / London. No commercial instrument-maker could possibly have afforded the time to create an unfinished object as complex as this, which suggests that it was one of his specials for the RI. Maybe it was abandoned because the researcher who commissioned it lost interest, or left, or died, or he and Newman realised they had come to an over-engineered dead end. It dates most probably from 1827 to c.1841, but Robinson up to 1841 had shown simpler and better ways of devising hydrostatic balances. This item now resides in the Boerhaave Collection of the Rijksmuseum, Leiden, with accession number V24640. 64 Kater (1826), 49. 65 This account is based mainly on Dawes’ DNB biography. 66 See Buchanan (1982), 337.
48 Another Set of Interactions. That ends the listing of all the relevant English makers and Natural Philosophers of which the author is aware, circa 1800―1820, including those earlier ones having an influence down to this period. Whilst each individual above has been treated completely separately, yet noting where one has clearly trained or otherwise influenced another, recent discoveries have indicated a web of much more complex interactions between certain instrument-makers and Natural Philosophers of the period. These occurred between various members of the well-known Lunar Society of Birmingham and their visitors. Matthew Boulton (1728―1809) and James Watt (1736―1819) seem to be at the centre of these productive interactions. For the Lunar Society generally and its membership see the references. 67, 68 Boulton’s letter books still survive, and are now in the Birmingham Library archives, MS3782/12. They record an extensive correspondence between Boulton and Troughton from 1782 to 1785. Boulton supplied instrument parts not only to Troughton, but seems to have been making parts for various members, such as brass tubes to Ramsden for telescopes and microscopes, whilst simultaneously buying parts and instruments from Ramsden; likewise, technical advice was flowing in both directions. Thus there was a letter of January 11th 1782 from Jesse Ramsden to Boulton concerning the supply of tubes. On another front, Ramsden was working with Benjamin Martin’s son on methods of gilding and silvering copper tubing; Boulton’s opinion was requested.69 And so it goes with multiple interactions driving scientific and engineering progress, with Boulton usually somewhere in the web. Watt is well-known as the inventor of the separate condenser, the greatest single step forward in the development of the steam engine, but it is less widely-known that he was initially a mathematical instrument-maker with a little shop in the precincts of Glasgow College. Here, in the winter of 1763/4 he had to repair a model of the Newcomen steam engine, clearly constructed by a fine instrument-maker, possibly by Sisson, for it was being repaired later in London by Jeremiah Sisson in 1760. It was entirely by this fortuitous event that Watt became involved with steam engines. In his Glasgow shop Watt met Dr Joseph Black (1728―1799) and the two became lifelong friends. In his discovery of latent heat, Black was weighing water. Whether the surviving ‘Dr Black’s Balance’ in the National Museum of Scotland was the one used is unknown, but that surviving instrument is clearly a superior scale and not a balance, reputed to be sensitive only to the nearest grain (64.8 mg).70 In Watt’s work on the separate condenser, he too was weighing water, and it was Black’s discovery of latent heat which provided Watt with an explanation of some of his findings. Watt complained of ‘the only scale-beam of a proper size which I had then at my command not being very sensible’, so one assumes it was not made by him. In Watt’s retirement garret workshop there were scales, weights and many sundry boxes of mathematical instruments, but whether any of these were of his own making is not known.71 It is also little-known that Watt was interested in chemistry, for which he would again have needed a balance or a superior scale. For Watt as a maker and repairer of instruments, see reference.72 67 Lound (2015), 5. 68 Gale (1966), 19-20. 69 Lound (2015), 62-63. 70 Black’s balance is currently in the National Museum of Scotland, Accession No. NMS T.1858.275.3, and there is a copy in the Science Museum in London, Accession No. 1915-34. 71 Dickinson (1929), 9, 17. Here he was weighing ingredients for the plaster blocks for his sculpturing machines. 72 Lound (2015), 10-13.
49 James Ferguson (1710―1776), astronomer, artist and clock-maker, gave lectures on Experimental Philosophy, covering a multitude of subjects, including specific gravity, for which he would have needed a good scale. He began his peripatetic lectures after Benjamin Martin had ceased his in 1756. Ferguson was a good friend of Boulton and John Whitehurst and was acquainted with others of the Lunar Society’s circle. Many more interactions are given in Lound (2015), and many others are to be found in Boulton’s letter books and elsewhere. For example, Haas made an exceptionally fine assay balance for Boulton in 1797, sensitive to 1/1000 grain with a pennyweight in each pan. Boulton was so pleased with it that he had no problem in paying £20 for it, which was twice the expected price. He even sent two of his existing fine balances to Haas for renovation and wrote: …and add to them your plan for taking the beam off its edge, & that of the six pins for steadying the scales. [pans] This tells us that Haas was using here an arrestment system, as previously noted in the van Marum balance on pp.33-34. The six pins are presumably of the style used on the Troughton balance 01.02.17, where each pan table bears a triangle of metal points to locate and hold the pan. 73 Earlier, Whitehurst had made an assay balance for Boulton in 1785.74 Wollaston is known to have supplied considerable quantities of his malleable platinum directly to Troughton from 1806 to 1820, and a table of the dates, quantities and prices charged is given in the reference.75 Again, Wollaston supplied a gold/palladium alloy to Troughton and other makers for non-corroding graduated scales on telescopes.76 Finally on this front, another and almost unrecorded interaction was going on, that of skilled workmen transferring from one workshop or factory to another, taking skills with them and in some cases influencing their new employer. 77 In summary, various members of the Lunar Society were weighing, making scientific instruments including balances, making instrument parts for each other, and exchanging know-how, advice and philosophical assistance. Only some of the interactions have been recorded; there must have been many, many more. Thus the Robinson ― Kater partnership was following in an established tradition, and was not as strange as it at first seems to modern eyes. Undoubtedly there were similar groups of philosophers and instrument-makers elsewhere collaborating at a lesser intellectual level, but unrecorded. So, where are we now; how can all of the above be summarised? Discussion. Two hundred years is a long time ago, so it is unsurprising that fine balances made in pre-Robinson times are very thin on the ground; this needs to be kept in mind during the following discussion. The first real step towards the development of the high precision beam balance is, on current evidence, due to the brilliant mind of Henry Cavendish. At first sight this seems improbable, with his research balance featuring a heavy iron box beam, a 73 Bryden (2025), 6, from the Boulton papers, Birmingham Library MS3782: 12/42/22, 23, 26 74 Craven (1996), 124 f/n 55. 75 Chaldecott (1987), 93. 76 Chaldecott (1987), 95. 77 Lound (2015), 65-66, 78-79.
50 crude wooden case and a massive and overly complex geared arrestment mechanism, but conversely he has made all three bearings of the knife and plane type, in steel, and has included the all-important adjustments of the knives to ensure the essential requirement that they be equidistant, co-planar and parallel. Further, he devised an arrestment frame carrying cradles which catch and locate the ends of the beam and the end planes separately, thereby ensuring truly accurate placement of the centre knife on the centre plane, and of the end planes on the end knives, at exactly the same locations every time, thereby ensuring much greater repeatability than any other system of placement known at that time. To this must be added a sensitivity adjustment, a fine poise adjustment, crossed levels and four fully adjustable feet to ensure precise level. The degree of adjustability of this balance is most notable for the period. This clearly is in line with the immense experimental rigour shown in his gravitational research as noted above. The identity of the instrument-maker who built the Cavendish balance is still in some doubt.78 Later makers used various selections of its features, as can be seen in their notes above. Only Robinson adopted the Cavendish cradles, thereby ensuring that his balances were more reproducible than those of his competitors, although the Field of 1826 must have come close. On the other hand, later makers simplified the arrestment mechanism until it became the enduring camshaft, cam and push rod, but sadly via some unfortunate aberrations such as see-saws and Robinson’s own spring-loaded lever-lock. Beams which were held by V-notches of one type or another solely at the centre knife suffered a loss of accurate placement when the notches began to wear with use; a notable case here is Ramsden’s cone beam balance. A similar affliction must have occurred on Fidler’s, Troughton’s and the other copies. Later makers arrived, eventually, at a far superior design for high-precision beams. At the period in question there appears to be, at this distance and in the data above, a rivalry between cone-beam balances and those featuring the flat rhomboidal lattice beam and, likewise, a competition between beam-end pointers moving over lengthy graduated charts, and flat beams with descending central pointers moving in tiny arcs over very short graduated charts located within cut-outs in the stand. As a generalisation, cone beams were favoured for the more expensive and the larger balances, whereas the flat rhomboid was preferred for the simpler and smaller instruments. Whence these two approaches? European influences must not be overlooked, but any definite influence by this maker on that maker seems impossible. Dutch instrument-makers were, in the 18th century, the first makers of fine balances. They were followed by French makers such as Fortin, Gondolfi and Mégnié creating fine balances at the end of the 18th century, most notably for Natural Philosophers such as Antoine Lavoisier (1743 ― 1794). Some were large chemical balances, some were high-capacity reference balances, and many were assay balances. From these different types the later and more practical high precision chemical, analytical and assay balances evolved. From the end of the 18th century onwards many French philosophical instrument- makers became famous by making weights, balances, measures and ancillary apparatus needed for the recently introduced Metric System. 79 78 Curiously, not his preferred maker, Edward Nairne. Instead, a Mr Harrison, but which one; surely not John Harrison of marine chronometer fame? See Buchanan (1982), 120-121, and McConnell (2007), 64. 79 The thoughts in this paragraph have been taken further in Oliver (2020), 5-6.
51 The cone beam. This was an unfortunate aberration in beam design and, in short, an evolutionary dead end. This design was devised by Jesse Ramsden, as recounted above. Cone beams were essentially obsolete by c.1815 but, unfortunately, because of Ramsden’s deserved reputation as a leading instrument-maker, many other makers both here and abroad followed this one unfortunate lapse by him: Fidler, T. Jones, W&S Jones, Mendelssohn, Tralles, Troughton and others. Many followed not only the cone beam but also the suspension pieces with loops over steel end knives, greatly inferior to Cavendish’s end knives and planes. One wonders how much further the influential Troughton might have progressed in the design of fine balances without having the restriction of the cone beam upon his thinking. McConnell quotes a very interesting observation concerning the Royal Society’s Ramsden cone-beam balance:80 In pointing out the balance to me one evening, Dr Wollaston told me it was so delicate, that Mr Pond, the Astronomer Royal [from 1811-1835], when making some observations with it, found its indications affected by his relative position before it, although it was enclosed in a glass case. When he stood opposite the right arm, that end of the beam preponderated, in consequence of it becoming expanded by the radiation of heat from his body; and when he stood opposite the left arm, he made this preponderate in its turn. It is probable that Mr Pond had previously adjusted the centres of gravity and suspension so near to each other as to give the balance its maximum sensibility, consistent with stability. By ‘delicate’ he meant sensitive, and this of course was originally seen as a virtue, and the balance was reported as having a sensitivity equivalent to 0.6 mg. This new observation must have been made no earlier than 1811, some twenty years after the balance was made. Presumably the effect of John Pond’s proximity was a slight warming of the air at that end of the cabinet. This would not have affected a flat rhomboid beam much if at all, but the far greater surface area of the cone and its large volume would have effected a change in the upthrust due to the displacement of air by one end of this large body. Warmed air being of lower density would give a reduced upthrust, hence the beam would have dipped slightly (‘preponderated’) at the warmer end. This seems to be a more likely explanation than actual expansion of the arm length. As nothing was said on this subject by Blagden when reporting the critical weighings for which the balance was built, one wonders to what extent this factor imposed unperceived variations on his weighings; in other words, becoming a significant part of his experimental error. Once made, this observation is yet another reason to add to the more obvious others for abandoning the cone beam: it is difficult to make, therefore expensive, and difficult to arrest and release reproducibly every time. That last reason is because the cone beam has always been held only by the lugs on the centre knife, which allows for a minute wobble on release, and increasingly so as the centre V’s wear with use; for no cone beam has been seen where the beam is held other than solely at the centre. Troughton’s springs seem not to have worked, as no other makers took up the concept. All these limitations apply to all cone beams. Thus the cone beam, already more-or-less obsolete, was given its final death sentence by this observation of Mr Pond sometime after 1811. It is totally unclear why W&S Jones, seemingly alone, persisted in offering cone beam balances for another thirty years or so. 80 Quoting from Andrew Ure’s Dictionary of Arts, Manufacturers and Mines, (3rd ed. 1846).
52 The flat rhomboidal lattice beam. This has totally obscure origins. Jenemann has studied the development of the chemical balance and of the high-precision beam more thoroughly than any other recent scholar, and his views are usually considered to be the most authoritative available.81 Tracing the development of high precision beams backwards in time, Jenemann concluded that it derives from two designs by Tralles.82 The one in Fig. 30 was, he said, clearly a forerunner of the soon-to-be-conventional lattice rhomboid, with fixed end knives. The other was a cone beam, with agate planes and adjustable end knives, incorporated in a balance built in the Mendelssohn workshop in 1808.83 In their two papers of 1808, Mendelssohn and Tralles cite each other; clearly this was a cooperative venture. FIG. 30. THE TRALLES BEAM OF 1808. But the above is all rather too neat; Jenemann is incorrect. The lattice rhomboid was not first conceived by Tralles, but seemingly appeared out of nowhere in Magellan’s paper of 1781 as seen above. A very early surviving example of this type is the Haas & Hurter balance of c.1790―1795, at least ten years pre-Tralles and pre-Mendelssohn. Clearly the concept was ‘in the air’ towards the end of the 18th century; its true origin still remains unknown. The probability is that Magellan picked up the idea from some unspecified source and passed it off as his own. The Tralles beam now appears to be an elegant derivative of the Magellan design, with more sensible vertical reinforcing braces. These braces became the standard style for rhomboidal lattice beams thereafter, so only in that regard was the Tralles beam pioneering. Philosophically, the cone beam and the flat rhomboidal lattice may be seen as not deriving from two fiercely competing origins. Think of it like this; most of the stresses applied to a balance beam are in the vertical plane, so the lateral strength of the cone form is largely wasted. Consider a vertical cross-section along the length of a cone beam; it is in effect a rhomboid, so why not start with a simple flat brass rhomboid form and dispense with the rest of the cone? It would have many advantages; not least it would be much easier to make. Most makers took this view, including Haas, Bate, Robinson and, later, the two Oertling brothers. To date not one cone beam has yet been attributed to any of these makers. The Tralles/Magellan three-knife flat precision beam concept carried the day but using the adjustable end-knives and agate planes from Tralles’ own cone-beam. This evolved form was clearly in use in England before 1820, and adopted by Robinson and Kater. Exactly how it arrived here is still unknown. 81 A wide-ranging review of the evolution of the precision balance in all of Europe, including Britain, is to be found in Jenemann (1997), 27-48. 82 Johann Georg Tralles, 1763―1822. He explains his thinking in a convoluted manner in Tralles (1808). This seems to be almost a throw-away idea, for only this one paper out of about seventy is on the beam, so it is not a topic of great interest to him. Presumably his interest in the beam arose from his specific gravity work. How curious that this one paper, a mere side issue, became so influential with a massive permanent effect. In fact Jenemann has over-rated Tralles’ importance. For brief biographies of Tralles see Poggendorff (1863) and The Annual Register for 1822, Chronicle Nov.19. 83 Mendelssohn (1808). It was obviously influenced by Ramsden’s cone beam.
53 But we need to go further back still: whence the sharpened knife edge? It appears to have been invented by Jacob Leupold, appearing in the form we now call pippin steel, in his sumptuous physics book of 1726.84 This was a pre-formed and hardened steel knife, but Ercker was soldering stout pins into steel beams and filing the protruding stubs to sharp edges, in situ, in the 16th century.85 Leupold has another significance too. He seems to have been the first to devise the basic structure of the true fine balance, with a central plane seated atop a pillar firmly fixed to the base of a closed case, as opposed to a suspended beam hand-held or hanging inside a case. All scales and balances pre-1726 were suspended, and a few such were still being built in the late 18th century, e.g. by Whitehurst as seen in Fig. 11 above. Leupold’s beam was solid and featured three co-planar knife edges. He devised panstops as part of a spring-loaded arrestment system which separated the centre knife and plane when not in use, in a surprisingly modern manner. Most of this derived from his practical instrument-making experience. We can now see that balances capable of achieving one-part-per-million sensitivity had already been achieved by 1820, but only with large, one-off specials, such as those by Troughton in the 1790s. All the essentials for such high precision had already been invented: the flat rhomboidal lattice beam with vertical bracing as required for the load; beams with the three knives equidistant, co-planar and parallel, and the adjustments to make them so; gravity bobs to adjust the sensitivity accurately; flags to adjust accurately the poise; integrated panstops; descending arrestment frames capable of placing knives on planes very reproducibly; agate planes; platinum wires, pans and masses; crossed levels; glazed cases with adjustable feet and sash doorslides, and accumulated experience in making hardened steel knives accurately straight and true. Most of these advances were created on unique balances or models made in small numbers. While education and scientific research expanded greatly as the 19th century progressed, what was needed was a balance-maker able to incorporate all these ideas into a set of practical designs capable of production in quantity, although not mass-production in the modern sense. In England that rôle fell to Robinson or, more accurately, to Captain Kater and Robinson working together to take the final step across the divide between fine scale-making and fine balance-making. Thus, as we approach Robinson’s entry into the English fine balance-making scene we find, based on a very low survival rate, cone beam balances with end pointers seemingly in profusion, especially for larger balances, but smaller balances were mostly with a short vertical pointer moving in a short arc over a very short chart inside a cut-out in the stand. A few balances, mostly anonymous and featuring very short charts within cut-outs in the stand, occur in various museum collections. This construction may now be reasonably called the English style. This is in contra-distinction with the Continental style of a long pointer pitched forward and swinging in a wide arc over a long chart fastened towards the base of the stand, creating an instrument much easier to use and of greater sensitivity as expressed in mg per chart division. But Troughton’s assay balance of no later then 1830 in Fig 16 was already pointing in this direction. This style was probably first used in Europe by Pistor, and adopted by other European makers, including August Oertling and others of the Berlin School of balance design. Pistor may well have been 84 Leupold (1726), Part I, 22-35 and plate VII. 85 Ercker (1574), 37; Ercker (1580), 82.
54 inspired by learning of the Whitehurst and Field balances with wide charts, which he would have heard about via the floods of correspondence between the learned societies of Europe and the UK, and in their journals. In instrument research it seems that nothing is ever novel; almost every advance has some earlier progenitor or inspirational source. Finally, Fig. 31 shows another and anonymous example of the English style. This is in the History of Science Museum, Oxford; accession number HSM 17549 (was MHSO 160). With its 10¼″ (260 mm) lattice rhomboid beam with vertical braces and divided for a rider, steel centre knife on an agate plane, a wide and unique arrestment frame with lever release, it seems very progressive and post-Robinson, yet it carries unrelieved non-adjustable swan necks, and an extremely small chart of three divisions per side within the pillar. Buchanan is of the opinion that the glazed mahogany cabinet, with its front and rear doorslides and bun feet, was not made for this balance. The doorslide rises on three stops, and the glass is held in with faded red putty, akin to domestic window glazing. In the scientific instrument field this is considered to be an early style of cabinet glazing. HSM 17549 is an undateable strange mix of the conservative and the progressive. The front doorslide has been removed for the photograph. [© HISTORY OF SCIENCE MUSEUM, UNIVERSITY OF OXFORD. FIG. 31. HSM 17549.
CHAPTER 5 The Typical Robinson Balance OME TWENTY ROBINSON and Robinson-style balances have survived and have now been studied, so it is currently possible to see exactly what constitutes a typical Robinson balance. It also means we can now see the progression in the designs with the passage of time. Such an opportunity was denied to Stock who knew of only three such survivors. But before we begin to examine the design progression, let us look first at the typical Robinson balance. Figures 32 and 33 overleaf are of a typical production balance by Robinson, with and without its detachable glazed wooden lantern. The balance selected as ‘typical’ is the only known surviving Robinson 8″, despite it being a rare size, because it has all the standard features and just one very tiny and unique feature which will be silently passed over here, but will be considered in detail in Chapter 13. This balance and all the others are descending arrestment, not rising plane, balances. Whilst the most distinctive feature of all Robinson balances is, by far, the Robinson Tulip, there are others and all are listed below; each will then be discussed in turn. The combination of these listed features enables one to spot and identify a Robinson balance at a considerable distance. The ‘Robinson Tulip’ The tiny chart within the ‘Tulip’ The very tall finger bolts holding the centre plane The flat rhomboid lattice beam with Robinson slit ends, pushing and pressing screws The ‘Cavendish’ cradles on the arrestment frame The spring-loaded lever-lock release and the sleeved pillar The distinctive ‘Wollaston pan stabilisers’ The detachable lantern of the cabinet Plus Robinson’s informative and elegantly engraved name and address, in copper- plate script filled black, on (almost?) all of his balances, although this can of course be seen only on close-up inspection: RRoobbiinnssoonn 3388 DDeevvoonnsshhiirree SSttrreeeett PPoorrttllaanndd PPllaaccee,, LLoonnddoonn Sometimes the LLoonnddoonn is missing. S
56 NAMING OF PARTS IN ALL ROBINSON SCALES AND BALANCES This table is intended to assist the reader in identifying easily the smaller and more unusual and distinctive parts of Robinson scales and balances, and to learn the correct terminology. Common items such as the beam, pans, cords and spirit levels are regarded as obvious and not needing identification assistance. As the balances are mainly symmetrical, often only one of a pair has been labelled, for clarity. A ‘Tulip’ (Aa outer and Ab inner on the swan-neck ‘Early’ scales.) A1 Tulip Boss A2 Arrestment Arms A3 Arrestment Frame (triangle, consisting of horizontal top, arms A2 and base A4) B Pointer (with gravity bob B1 and locking ring B2) C Chart D Centre Knife (block) E End Knives (blocks) F Flagpost and Flag G (‘Cavendish’) Cradles H Set Screw I Pressing Screws J Pushing Screw; J1 stiffening screws K Locating Slots (on cradles for end plane stubs) L Trunnions (Stubs) L1 of End Knives; L2 of Suspension Pieces M Suspension Piece with J-hook; often referred to simply as ‘J-hooks’ N Robinson Beam End Slit; N1 reinforcing insert O Height Adjusting Screw for Cradles; O1 fine levelling screws for Cradles P Anchor Bolts for Cradles Q Finger Bolts R1 Centre Plane (Shown on its displaced brass holder after the beam has been removed) R2 End Planes S1 Stand Base S2 Stand Sleeve S3 Stand Sleeve Pin S4 Stand Ball-and-Socket Joint S5 Swan-neck with Vertical Adjustment S6 Swan-neck with Horizontal Adjustment T Release / Arrest Lever U Specific Gravity (SG) Pan V Lantern; V1, V2 Lantern Hooks to Base W Lantern Solid End Doors X Lantern Handle or Knob Y Drawer; Y1, Y2 Drawer Hooks to Base; Y3 Drawer Pull Z ‘Wollaston Pan Stabilisers’; Z1 (L) and Z2 (R) centres of motion.
57 [MOST PHOTOGRAPHS OF BA104 APPEAR BY COURTESY OF ROTHAMSTED RESEARCH / LAWES AGRICULTURAL TRUST. FIG. 32. ROTHAMSTED BA104 WITHOUT ITS LANTERN. [© PENNY BIRD. FIG. 33. ROTHAMSTED BA104 WITH ITS LANTERN IN PLACE. The most distinctive feature of a Robinson’s stand is the ‘Robinson Tulip’, shown as A in Fig. 32, and close-up in the frontispiece and Fig. 34. The pointer B and the chart C are enclosed within the Tulip frame. The steel pointer is screw-threaded for Q A3 A2 G Y Z T Y3 A WW
58 a brass gravity bob B1 and locking ring B2, and the pointer tip hovers above the very short silvered brass chart C. The proportions of the Tulip are not always quite as seen here, but this one is typical. Some are a little more squat or more elongated. There is some debate about the use of such a very short chart and arc swing of the pointer. Most lab staff who had to count swings on early balances wanted a wide arc and a fine scale with which to determine the rest point using the conventional methods as outlined in the Oertling book.1 But balance theory rests on the arc swing of the pointer being very small, about 1°, otherwise the theory does not hold. Therefore, it is argued, the wider charts are incorrect theoretically. But any errors will be the same on each side of a centre zero chart, thereby cancelling each other out when counting swings. All elderly balance engineers and lab staff who had to work with free-swinging balances, and were known to the author, preferred reading charts with wide arcs. The grey marks on the upper front face of this Tulip, with two black spots, appear to be nothing more than an unusual corrosion of two screw heads. FIG. 34. THE ‘ROBINSON TULIP’, CHART AND POINTER. Robinson production beams were nominally flat, brass, rhomboidal lattices, as shown in Fig. 35, and fitted with knife-and-plane bearings throughout. These were usually steel knives and agate planes of good length: nominally 1″centre planes and ½″ end planes. The steel centre knife D and the two steel end knives E, E are clearly seen. In fact the beams were not really flat, but about 3 mm thick at the centre, tapering to about 1½ mm and expanding to 3 mm again for the slit ends. The attachment of the pointer B to the beam is clearly visible; likewise flag F on its post. FIG. 35. THE ROBINSON PRODUCTION BEAM. 1 Oliver (2020), 691-706 (Chapter 40). A3 A B B2 B1 A1 C A2 F D J E N I HB I E A4
59 Now consider Robinson’s beam ends as shown in Figs. 35, 36 and 37. In Fig. 36 the Robinson slit N at each end is clearly seen, but the adjustments are partly hidden by the cradles G. Instead, the drawing Fig. 37 shows clearly Robinson’s sophisticated beam ends. The knife block is attached to the beam end from below by a single set screw, H. The two pressing screws I, I, working on the knife block may clearly be used together to lengthen the arm, or in opposition to achieve parallel with the centre knife. Note the highly original slit adjustment with the pushing screw J to widen or close the slit N.2 The motion of the knife will of course be radial, thereby changing not only the knife height, but also the arm length, so these adjustments are partly interactive. Almost certainly Kater or Robinson was the true inventor, as a search for a slit-end beam prior to 1820 by a European maker has proven fruitless. FIG. 36. THE ROBINSON PRODUCTION BEAM AND ARRESTMENT FRAME. AFTER THE DIAGRAM IN STOCK AND BRYDEN (1972). FIG. 37. THE ROBINSON PRODUCTION BEAM END. 2 The reader is advised to note carefully the distinction between Pressing and Pushing screws. This terminology has caused much confusion, and even some of the later Oertling servicemen erroneously called them all Pushing screws. Q P O1 O1 A3 J E H N I I I L1 E L1 J G N
60 The cradles need special attention; see Figs. 38, 39 and 40. They are of excellent design and execution and they truly work. All have been inspired by those of the Cavendish balance. The cradle G at each end of the arrestment frame has locating slots K, K for the end plane, plus sloping inner faces to catch the sloping ends of the stubs L1, L1 on the end knives in a light wedging action. Fig. 39 shows the beam end being correctly located in the cradle G via the stubs (trunnions) L1, L1 on the steel knife E. Equally clear is the suspension piece, M, with its agate plane R2 and J- hook, located by its own stubs L2, L2 finding the slots K, K on the cradle correctly. Hence the plane has been lifted off the knife, and a small gap is visible. In Fig. 38 the arrestment frame and cradle have dropped, and the end plane now rests on the end knife in its working position. FIG. 38. RELEASED. FIG. 39. ARRESTED. Fig. 40 is an end-on view. Out of focus to the rear are the giant finger bolts, Q, Q with the end-on view of the beam fading away between them. Each cradle is individually adjustable for height and level, as it is essential that the clearances between the knife and plane at the two ends of any beam are equal, or the beam will ‘kick’ badly when released. See Fig. 41 for these adjustments. Screw O raises/lowers the cradle whilst the two screws O1 are used separately to micro-adjust the level to ensure the end plane and end knife are exactly parallel. Each cradle is fixed to the arrestment frame by anchor bolt P. Figs. 39 and 40 show a slight parallel adjustment is needed.3 FIG. 40. END-ON VIEW OF THE ARRESTED BEAM END. 3 It also shows that the delicate trunnions of this suspension piece have been slightly bent at some time, leading to the lack of parallel of the end plane with the end knife. MR2 L2 L1 Q L2 H L1 G M R2 E E M K L1 L2 R2 O O G G
61 FIG. 41. THE CRADLES ON THE ARRESTMENT FRAME, AND THEIR ADJUSTMENTS. FIG. 42. THE FINGER BOLTS AND CLEAR AGATE CENTRE PLANE. This particular typical balance is the Rothamsted Robinson which will be further described in Chapter 13. Figures 41 and 42 were taken during the author’s detailed investigation of this balance. These images, taken from the rear, show the interior faces of the cradles, and the inevitable green corrosion of those un-lacquered brass faces. The clear agate centre plane R1 is visible, glued to its brass carrier with a clear adhesive, probably one of the cheaper types of isinglass.4 Also apparent is the way the plane carrier has been mounted across the front and rear parts of the Tulip with the giant finger bolts Q, Q. In a fully assembled balance the plane and its carrier pass through the centre light of the beam. The way the top part of the triangular arrestment frame curves awkwardly around the rear of the Tulip is very obvious. Concerning the Cavendish cradles it is most regrettable that Robinson’s successors in London, namely Bate, De Grave, Oertling and later others, all ignored his pioneering lead in optimal arrestment of routine production balances. Possibly they thought it was too expensive for making on a routine basis, but this author argues that if it had become routine and the best possible procedures had been found, that ‘too expensive’ argument does not convince. However, this excellent feature was never followed up. 4 A fish-scale gelatine, not the common fish glue. G P O1 O1Q R1A3 P Q R1 A3 A2 O
62 The very tall finger bolts, Q, Q holding the centre plane really catch the eye, and they too do not appear to have been copied by other makers. Robinson clearly envisaged frequent dismantling of the balance for transport, with the beam, pointer and other parts inserted into their specific places in the fully fitted drawer. The reasons for this assumption on Robinson’s part will become apparent later in the book. As the centre plane R1 on its carrier passed through the centre light of the lattice beam, so it would have been necessary to remove the plane before the beam could be lifted out. Robinson’s tall bolts simply made this operation quicker and easier than having to unscrew normal bolts with a screwdriver. Where this style of construction was used, as it often was by many makers, the beam is said to be ‘captive.’ The construction of the stand takes the form of a pillar supporting the Tulip with its chart, surrounded by a sleeve S2 carrying the arrestment frame A3. When arrested the release lever T is slightly elevated as in Fig. 43. When the lever is pressed down it pulls the sleeve and arrestment frame down the pillar; the lever is then slid slightly sideways to the right to latch, leaving the balance in action without the user having to hold the lever down. The lever is now horizontal or slightly lower, as in Fig. 44. Consequently a gap has opened up between the Tulip boss A1 and the base of the arrestment frame A4. The lever passes right through the stand and is pivoted in a ball-and-socket joint S4 behind the stand, as in Fig. 43. This S4 socket is the normal one, but those on the early scales are slightly different. Whatever the particular form the lantern takes, there is always a vertical slot for the release lever, with a small sideways extension to receive the latched lever, as seen in Fig. 33 and many others. The action is spring-loaded to arrest the balance; it is called a lever-lock mechanism. The user has to be careful when arresting, as the spring is strong and likely to take over completely, giving a great jolt to the balance; hold the lever firmly. The lever had an ivory tip to its handle only on some of the Robinson & Barrow balances. FIG. 43. ARRESTMENT LEVER AND BALL & SOCKET. FIG. 44 ARRESTMENT LEVER IN USE. The so-called ‘Wollaston pan stabilisers’ or ‘pan stays’ or ‘Wollaston Apparatus’ Z were also copied from the Cavendish balance.5 These are shown clearly in Figs. 32 and 44, and in Kater’s original drawing in Fig. 47 in the next chapter. Note that there are just two vertical pins on each articulated arm, and they rest alongside the pans, they do not support them. Their function appears to be two-fold; first, simply to stop pan swing whilst loading and unloading, and secondly, as the arms are useable from 5 Just visible in the Cavendish photograph in Stock (1969), 12. See Chapter 4 and Buchanan (1982), 122. T S1 Z2 Z1 S4 S2 A1 A4
63 outside the closed case in both the Cavendish and Robinson designs, they were used to reduce pan swing after release if either pan was asymmetrically loaded. These were wisely ignored by most other makers, as they seem to be more trouble than they are worth. Note that on the Robinson in Figs. 32 and 33 the suspension cords have been shortened to such an extent due to wear that the pans pass right over the pan stay pins, giving rise to the impression that they might have been some type of support; impossible of course with only two pins, for three pins symmetrically placed in a triangle would be needed to do that. This is seen even more clearly in Fig. 110 of The Dancer balance in Chapter 19. The base S1 of the stand is a hollow brass casting, hand-filed and polished to a fine finish; it includes lugs for the rear spirit level. The base is finely engraved in Robinson’s customary three lines of copperplate script, filled black, thus: RRoobbiinnssoonn 3388 DDeevvoonnsshhiirree SSttrreeeett PPoorrttllaanndd PPllaaccee,, LLoonnddoonn by a professional engraver (not Robinson), but the fore-aft spirit level eclipses a little of the inscription and the head of one of the blued steel woodscrews holding the base in place, so looking like an afterthought. One wonders why this level is not on lugs like the other. Some levels eclipse the fixing screw quite badly, as in Figs. 72 and 73 in Chapter 10, and many others. Surely a craftsman as talented as Robinson would not have intended to eclipse a screw head in this way, and instead mounted the level on matching lugs to the side. He presumably then realised that to do so would make that level potentially interfere with using the L lever of the ‘Wollaston Apparatus’, already rather near on the small 5½″ version of his production balance. So, the level had to be mounted on top of the base as we find it, and it stayed there on the larger models, presumably so that one base casting could be used on both larger sizes. It is important to point out that Kater and Robinson created their suspension piece in the form we now know as a ‘J-hook’, which became the standard form of many makers for decades to come. It does look as if they invented this style of suspension piece, especially when one observes the primitive first effort used on ‘The Wollaston Robinson’, to be shown in Chapter 8. This is one of the main reasons why Kater and Robinson must be taken together as key developers of the high-precision balance. The brass pans were suspended on dark green plaited silk cords, the traditional scale-maker’s method. The use of silk is explained at the end of Chapter 7. The balance case consists of a wooden drawer base and a removable glazed wooden lantern, as shown in Figs 32 and 33. Two small hooks are usually fitted to prevent the drawer flying out during transit. In this instance the hooks are missing but the four screw holes remain; hooks and pins were there once. The drawer is fully fitted for the beam, pans, a specific gravity (SG) pan, suspension pieces, forceps, a brass weight set down to 100 grains, and a fractions set from 40 grains down, as shown in Fig. 45 overleaf. The SG pan is often hidden under the other pans, with the hook descending into a hole in the middle of a velvet-covered thick wooden insert in the pan pen. Sometimes it has a pen of its own and of the same pattern. The lantern takes various forms, all detailed in Chapter 22, and the drawer contents including weights are covered in Chapter 23. The drawer is usually equipped with a single ivory pull for the 5½″ size, and two pulls for the 8″ and 10½″ sizes, but sometimes the pulls are of an unidentified wood.
64 FIG. 45. THE DRAWER CONTENTS. There are more photographs of this balance in Chapter 13, and Kater has given a detailed description of a Robinson balance of this size and development.6 In the 20th century serious metrologists and balance-makers began to perceive the limitations of the arrestment of a beam solely by holding a stub on each side of the centre knife, a ‘two-point’ arrestment. Kinematic arrestment was introduced to provide precisely reproducible arrestment, placing the centre knife at exactly the same place on the centre plane every time via a ‘three point’ arrestment action.7 Yet 100 years earlier Robinson was leading the way using a ‘four point’ arrestment, with two points at each end of the beam of all of his balances. One unintended consequence of Robinson’s provision of fine adjustments to his beam and arrestment is that he essentially created the need for service engineers. Whilst some customers would have been capable of using these refinements correctly, others would not, and needed to seek professional help. Exactly how this was done in Robinson’s time seems unrecorded, but Oertling in London soon had a representative visiting his customers for both sales and service, when there were too many clients for Ludwig to attend personally. Those clients living outside London had to crate up their balances and send them back. It is most noticeable that Robinson balances were highly regarded, therefore worth having them serviced, and they often appear for servicing in the Oertling ledgers. 8 Indeed, Dittmar recorded that some Robinson balances were still in use as late as 1890.9 During research for this book, it has been found that a few were in use well into the 20th century. Having considered the structure of the typical Robinson balance, we can now restart at the origins and trace the progression of his designs in the following chapters, up to the typical balances. Thereafter it is possible to follow the development of his style up to and beyond the death of Robinson. We can then examine some anonymous ‘pseudo-Robinson’ designs and attempt to decide which, if any, are actual Robinsons, or precursors or created by copyists. 6 Kater (1831), 322-324. 7 Oliver (2020), 847-851. The highly negative opinions expressed by physicists in 1933/1934 at NPL upon Oertling’s routine two-point arrestment system are very telling (p.850). 8 Oliver (2020), 353. 9 Buchanan (1982), 213.
CHAPTER 6 The Origin of Robinson’s Balance Designs HE FOLLOWING LIST comprises all the known surviving scales and balances by T.C. Robinson, his successors and copyists and possible predecessors. They are in date order as far as can be ascertained, and by the logic discussed later. The anonymous three at the head of the list are provisionally included on the basis of the later discussion. The author argues that a distinct progression of design can be seen across all these examples. MAKER BEAM INSTITUTION ACCESSION NO. NAME / NOTES Children’s balance sketch in Fig. 46, beam length unspecified. ― Has not survived Anonymous missing, c.5″ Hist. Sci. Mus. Ox. 18579 Corless or Robinson? Anonymous missing, c.10″ Whipple Museum Wh.1446 Corless or Robinson? Anonymous 6″ Sci. Mus. ex-GIII 1927-1138 Corless or Robinson? T.C. Robinson 5½″ Private collection ---- The ‘Steel Scale’ T.C. Robinson 5½″ Science Museum 1977-159 The ‘Welcome’ T.C. Robinson 5½″ Whipple Museum Wh.1109 The ‘Wollaston’ T.C. Robinson 5½″ Whipple Museum Wh.1413 The ‘Miller’ T.C. Robinson 5½″ Matt Hass Colln. 1 T.C. Robinson 5½″ Matt Hass Colln. 2 T.C. Robinson 10½″ Whipple Museum Wh.1593 Used by W.H. Miller T.C. Robinson 10½″ Cambridge U. ---- Damaged, incomplete T.C. Robinson 8″ Rothamsted Res. BA104 Possibly ex Kew T.C. Robinson 10½″ Royal Mus. Scotland IL.2010.19.2 ex Edinburgh U. T.C. Robinson 18″ U. Mus. Utrecht 0285-188871 ‘Dr Moll’s Balance’ Robinson & Barrow 15″ Unknown ---- ‘Miller 1843/1845’ Robinson & Barrow 10½″ Whipple Museum Wh.1371 ex Stokes Collection Robinson & Barrow 10½″ Science Museum 1977-340 ex British Museum Henry Barrow 10½″ Mus. Sci. & Ind. Y1969.5.78 Signed J.B. Dancer Adie & Son 10″ Royal Mus. Scotland T.1968.25 The ‘Adie’ Anonymous 9″ Private Collection. ---- Cheap look-alike? T
66 When one considers an instrument bearing a signature such as Robinson, one automatically expects this to be the name of the designer and maker. It has slowly become clear that this is not the case with Robinson balances and with some of his other instruments too. What we are looking at are the products of a partnership between Captain Kater and T.C. Robinson. This seems to have been very successful, very productive and harmonious, over a considerable period of time. Kater was a Natural Philosopher and Robinson an artisan instrument-maker or, to put their collaboration into modern terminology, Kater was the ideas man and Robinson was the maker. The question is: how did this partnership begin, when and where? There is no definite answer, but consider the following observations. There is a lengthy paper by Kater on creating a seconds pendulum.1 Here Kater was attempting to arrive at the precise length of a pendulum that would beat seconds accurately for astronomical time-keeping. He devised a reversible pendulum oscillating on one or other of two knife edges. It is very clear that Kater had a thorough understanding of knives, how to make the edges accurately straight and horizontal, how to measure their form, and the problems of steel knives working on steel planes versus agate planes. This paper was published in 1818, so he was therefore fully equipped to advise Robinson on knives for fine balances in the 1820s. His knife-maker is given as ‘Mr Stodard’ of whom nothing is known. In his paper Kater states that one phase of the practical pendulum work was carried out: …at the home of Henry Browne FRS, in a part of Portland Place not liable to much disturbance from the passing of carriages.’ Portland Place is a street of exceptional width, 110 feet, now running from Langham Place up to Park Crescent, but initially it terminated abruptly at the south end near the connection to Duchess Street. It has not been possible to locate Browne’s house, but presumably it was at the south end where passing carriages would be fewer. Devonshire Street is a side road off Portland Place a little higher up, such that Robinson was probably the nearest instrument-maker to Browne’s house; see the map at the end of Chapter 2. Possibly this was how Robinson was first brought to Kater’s notice. This is of course pure speculation, but it is a distinct possibility. Until recently the question as to how Robinson acquired the concept of the rhomboidal lattice beam, or did he re-invent it, was still unresolved. The answer, buried in two brief journal communications of 1821/1822, is that Captain Kater designed a simple and cheap balance for hydrostatic work, and that a few were made for him by Robinson at that time. This collaboration has become known due solely to the clarification of a misunderstanding in a quarterly London journal, wherein Kater in 1822 had to correct a statement made anonymously in a previous issue in 1821. These two communications are both short, very informative and worth quoting in full, together with the sketch of the first Kater―Robinson scale.2 The 1821 sketch of Kater’s design, as built by Robinson, is shown in Fig. 46. Description of the Balance represented in Plate V. fig. 1. A GOOD pair of scales is an essential implement to the chemist and mineralogist, but it has hitherto been scarcely attainable, except at a price far too considerable for the generality of purchasers. We are indebted to our friend Mr. Children for the drawing of the beam represented in the above-mentioned plate, and which will, we believe, be found to possess considerable accuracy in a small and convenient compass, and at a very moderate expense. 1 Kater (1818). Captain Kater spent many years working on the creation and use of such pendulums beating seconds extremely precisely for astronomical and geophysical work. For more on Kater’s enormously wide spectrum of labours see Chapter 4. 2 The full double reference is Kater (1822).
67 The beam, which is of platinum and made so light as to be extremely sensible, is at the same time rendered sufficiently strong by its form. The adjustments for bringing the points of suspension of the scale pans, equidistant from, and in a right line with the axis, are performed by the screws a and b at the ends of the beam; these screws also serve to strengthen the curved ends of the beam and prevent their bending. The axis of the beam is a piece of very hard steel, of an equilateral triangular form, passing through the beam and bearing on agate planes; the knife edges are ground to an angle of 120°, which has been chosen thus obtuse from the liability of a more acute edge to be broken when suddenly lowered upon the agates. The ends of the axis are chamfered from the top to the knife edge, so that when the beam is lowered upon its bearings, they are clear of the lifting frame. An index or pointer descends from the beam to a divided scale; upon this index is screwed the ball c; the office of which is to adjust the oscillation of the beam, or in other words to give it its required sensibility; should the index not exactly coincide with the middle division of the scale, it may be adjusted by turning the piece of wire d, at the top of the beam. The lifting frame, e f g h, is pressed upwards by an internal spring, its use being to lift the beam from the agates, when not in use, or when weights are put into the scale pans: this frame is lowered by the lever i, and is retained in that position by turning the lever into the notch on one side. At the bottom of the short stringed pan is a hook for the convenience of attaching substances whilst ascertaining their specific gravities. The whole instrument is covered with a glass case, and provided with a level for ascertaining the horizontality of the agates, and with tweezers, and a box of platinum weights from 1/100 of a grain up to 100 grains. The price of this apparatus as manufactured by Mr. Robinson, of Devonshire-street, Portland-place, is £6, with the beam and weights accurately adjusted, or £4 when unadjusted. FIG. 46. CHILDREN’S SKETCH OF THE ORIGINAL KATER-ROBINSON SCALE.
68 Kater responded in the next quarterly issue with a letter to the editor as follows: A Letter respecting the Construction of a Balance, from Capt. HENRY KATER*. London, Aug.8th, 1821. DEAR SIR, I OBSERVE in the last number of the Quarterly Journal, edited at the Royal Institution, the description of a Balance from a drawing by Mr. Children. I should not have troubled you merely to claim this as my own arrangement, but to notice an error in the description where the beam is said to be “of platinum;” it should have been “of bell-metal,” which combines the essential desideratum lightness with the requisite degree of strength. This beam had its origin in a wish to render the hydrostatic balance less expensive without diminishing its accuracy and sensibility, and it has in every respect answered my expectations. Dr. Wollaston has contrived the following additional apparatus, which he finds particularly convenient. The annexed wood-cut represents the bottom of the case supporting the balance, A and B the scale pans. Four slips of wood, or brass, are connected by pins forming centres of motion at a, b, and c, the centre c, being fixed to the case. Four brass pins, e, f, g, h, of a sufficient length, project perpendicularly from the side pieces. These pins are to be brought into lateral contact with the scale pans by means of the ends k, l, when the beam is elevated above its [FIG. 47. KATER’S ORIGINAL SKETCH OF THE ‘WOLLASTON PAN STABILISERS’.] support. The knife edge being then lowered upon the agate planes, and the pins withdrawn, any want of equilibrium becomes instantly perceptible. I am, dear Sir, Yours very faithfully, HENRY KATER. W. T. Brande, Esq. * We need scarcely observe that we were not aware to whom Mr. Robinson was indebted for the plan of the very useful balance of which we have given an account in our last Number. We merely saw the instrument in the possession of Mr. Children, who was kind enough to furnish the sketch from which our engraving was made. ‘Mr Children’ was J.G. Children, FRS. For the most part these are two very clear and informative accounts. A platinum beam would have been too heavy as well as prohibitively expensive, and recall this was intended to be an economy model. It is believed that the unsigned original article in 1821 was written by the editor (Brande), who states clearly that the design in the sketch was being made by Robinson, implying a routine production. Kater makes it clear that the original design was his.
69 This is technically a scale (unrelieved end bearings), and not a balance, but many of the principal characteristics of Robinson’s later true balances are already present: the Robinson Tulip, a flat rhomboid open beam, the short pointer hovering over a tiny chart within the Tulip, the sleeved pillar with the lever-lock release, the beginnings of an arrestment frame, and the detachable lantern on the drawer base of the cabinet. The distinctive slit-end beam, the Cavendish cradles and Wollaston’s ‘pan stabilisers’ have yet to appear. The tiny gravity bob on the pointer persisted for some time, but would eventually be replaced with one of a more sensible and conventional size. Finally in his letter Kater gives an account of the ‘Wollaston pan stabilisers’ making clear their intended lateral contact with the pans, but his final sentence is unfortunately suggestive of a supporting function. It is this plus the fact that many pans hover nowadays above the ‘pan stabilisers’ because of shortening of the cords due to wear, that it is widely supposed that they have some sort of supporting function, which they do not; see for example Figs. 32 and 33 in Chapter 5. It may be reasonably surmised that Kater, having encountered the Magellan beam in the 1781 Observations Sur La Physique and/or the Tralles beam in the 1808 Annalen, used the concept of a flat open rhomboid beam in 1821, but we have no statement by him on this. Kater as a Natural Philosopher would have been routinely scanning the literature and seen these items, but it is very doubtful if the artisan instrument-maker Robinson would ever have seen them unless introduced to them by Kater. In the same way Kater, being at the very centre of Philosophical research at this time, was presumably also aware of the high capacity balance constructed by Miller & Adie in Edinburgh, with a steel open rhomboid beam. Although Kater’s beam is an open rhomboid, it has the traditional scale-maker’s swan-neck ends, but to Kater’s credit, he has made good use of this inherently weak construction with the reinforcing screws a and b. These were two different inter- pretations of swan-neck spring end adjustment, the left one altering height and the right one arm length. The screw-adjustment for arm length was first seen in Benjamin Martin’s superior coin scale (in Chapter 4), but the vertical adjustment appears to be first conceived by Troughton. Kater would have perceived that in this manner not only could arm length be equalised accurately, but the three suspension points could be brought into a straight line, i.e. made co-planar. Thus precise adjustment was clearly being considered from the very start, but this traditional scale-maker’s swan-neck is not very satisfactory in high precision balances. These two screws, as well as enabling adjustment of arm lengths and height, provide a measure of stiffening of the swans and reducing their tendency to bend under load. His opinion of the maximum sustainable load was about 150 grains, say 10 grams. Poise (zero) was adjusted by turning wire d, the predecessor of the flag. Nothing is known as to the identity of the maker of the centre knife here, nor indeed of any of the knives in Robinson’s production balances, but for these early scales Mr Stodard is an obvious possibility. The only known preserved Robinson scales corresponding to the Children sketch are The ‘Wellcome’ and the ‘Steel Scale’ which are the subject of the next chapter. Overleaf is a photograph of one of them, Fig. 48, showing it to be very similar to the sketch plus added ‘Wollaston pan stabilisers’. The original design of the stabilisers had a single centre of motion, but these two early scales and all later Robinson balances have two centres of motion, as seen in Fig. 48 and many subsequent figures. Otherwise it is almost as if the two drawings have been merged; highly satisfactory.
70 Nowhere in the above quoted texts is the length of the beam stated. The ‘Welcome’ and the ‘Steel Scale’ are its earliest known physical realisations. Both have been measured as approximately 14 cm or 5½″, but all swan-neck beam lengths are perforce approximate. Experience has shown that early knife edge balances by many different makers have beam lengths which are noticeably variable around their nominal length, often as much as ¼″ either way. In early times this was irrelevant, the important point being that the two arm lengths were as equal as skill could make them. With the introduction of panstops on production models in the late 19th century, there then had to be absolute uniformity of beam lengths, as the points of contact between panstops and pans had to be precisely vertically under the end knives, or the balance would ‘kick’ horribly upon release. For readers unfamiliar with hydrostatic weighing, see Appendix 3. [J. WATERS. FIG. 48. THE ‘WELCOME’. (The release lever and the ball-and-socket joint are missing.) Reporting of Robinson beam lengths. We have seen in his catalogues that Robinson expressed the beam lengths of the three versions of his standard model to be 5½″, 8″ or 10½″ in length. In this book, to avoid confusion, all Robinson balances will be defined by their nominal beam lengths in inches, not the actual values to the nearest millimetre, although they may also be quoted if available.
CHAPTER 7 The ‘Swan-Neck’ / ‘Early’ Robinson Scales HE STARTING POINTS for an assessment of Kater and Robinson’s design progress are currently two in number. Each is a swan-neck scale design, quite different from the well-known Robinson production balances with the Robinson slits and Cavendish cradles. As these two surviving scales are almost identical with the original Robinson sketch in Fig. 46 above, it follows that these are clearly the earliest surviving Robinson weighing instruments known to date. The single significant difference from Children’s sketch is the addition of Wollaston’s pan stabilisers, in brass. The two instruments are: 1) The ‘Steel Scale’ Robinson, with a steel beam, in a private collection; 2) ‘The Wellcome’ Robinson, with a brass beam, previously Wellcome Trust no.219, now Science Museum no.1977-159. As all Robinson balances or scales, early or late, of whatever size, have brass beams, so the scale with the steel beam stands out, and one wonders if Kater was trying out several ideas for his ‘affordable’ hydrostatic balance. Currently only these two scales are known to have survived, and he obviously decided that the conventional brass was the correct metal for the production beams. These two early offerings must be of very similar date, in the few years c.1820, certainly by 1821 or 1822. It did not take long for the deficiencies in this design to become apparent, and the true Robinson balances soon emerged. The earliest known date for his production balance in three sizes is 1828, but the cross-over point between 1822 and 1828 is currently unknown. THE ‘STEEL SCALE’ ROBINSON. See the figures overleaf. All the brass parts are lacquered, as would be expected. The total weight of the instrument with lantern is 1,372g. Beam and pointer. The open steel beam (no braces) measures 5½″, 141 mm, between the swan neck ends S5 and S6.1 It is 1/32″ thick (~1 mm), not tapered, with the ends thickened to 1/16″ (~2 mm) to support the swan necks; the maximum depth is 16 mm. The steel knife D is 16 mm long fixed in a short steel cylinder which is a perfect fit to a circular hole in the central feature of the beam. This assembly is held in the beam between two brass discs 12 mm dia, sandwiched together with two tiny brass countersunk screws passing through the beam from the front disc and screwed firmly into the rear disc; see Fig. 49. This assembly is very rigid; no adjustment is possible. 1 All items labelled as ‘steel’ have been identified as such by their being strongly attracted to a magnet. T
72 [T. ALLGEIER. FIG. 49. THE ‘STEEL SCALE’ BEAM AND POINTER. [T. ALLGEIER. FIG. 50. THE ‘STEEL SCALE’ COMPLETE. F X A2 F S5 A2 A A1 Y3 B S6 B1 S3 B1 B S6 S5 D V
73 The steel pointer B is 59 mm long, (but clearly is now 5 or 6 mm too short) and nominally 1/32″ (~1 mm) diameter, threaded over most of its length for the tiny brass gravity bob B1, 2½ mm diameter, weighing only 0.127 g. The complete beam assembly with pointer and bob weighs only 6.216 g. [T. ALLGEIER. FIG. 51. THE ‘STEEL SCALE’ MINUS LANTERN, FRONT VIEW. [T. ALLGEIER. FIG. 52. THE ‘STEEL SCALE’ MINUS LANTERN, OVERHEAD VIEW. Y3 Y V1 Z T Z Z B B1 A1 S3 Z
74 Stand. The stand comprises the base, the sleeved pillar topped by two Robinson Tulips, an inner and an outer, plus the arrestment mechanism including the two sloping arms. The brass stand base S1 is 57 mm diameter, 6 mm deep and neatly engraved with the copperplate inscription: RRoobbiinnssoonn 3388 DDeevvoonnsshhiirree SStt.. PPoorrttllaanndd PPllaaccee Unfortunately this has been partially obscured later with the crude addition of the crossed spirit levels; seen in Fig. 53 (as received, before conservation work). The levels are 36 mm long, 8 mm diameter, 9 mm at the ends. Both are intact. They are clearly early, with the screw adjustments for level clearly not thought out properly. They are neither Robinson’s type nor later Oertling-made additions. Also, it is hard to believe that Robinson would ever have defaced his earlier work in this way. Their maker is not known for certain, but the probability is that they are by R.B. Bate, judging by the design of his levels on the Bate/ Oertling balance at Rothamsted.2 FIG. 53. THE ‘STEEL SCALE’ STAND BASE. Whoever committed this outrage has clearly demonstrated that the real need for levels was soon perceived. Robinson later allowed for this with the rectangular base on his classic production design, to be covered in the next chapter. The pillar is held in the brass base with a single screw. At the top it expands to form the inner, fixed, Tulips Ab in brass tinted to a strong blue colour.3 The tops of the Tulips, front and rear, each hold an agate plane 6 x 3.5 mm held in place by red sealing wax. A horizontal circular ring 10 mm diameter in the same ‘blued’ brass connects the two uprights at the top, providing rigidity; the pointer descends through the wide hole in the middle. Aa and Ab show poorly here; instead see Figs. 58/59. The base of the inner Tulip carries a tiny lacquered brass chart C, a mere 9 x 7 mm, sparsely engraved as shown in Fig. 54 at approximately six times actual size. There are only two divisions each side of centre zero. All the lines are lightly engraved and unfilled. It is very difficult to read, not only by being so small and faint but is also effectively poorly lit inside the two Tulips. The pillar is enclosed in a brass sleeve S2, 8 mm in diameter. At the top of the sleeve is the brass Tulip boss A1 above which rises the moveable brass outer Tulips Aa mimicking the outline of the inner, fixed, Tulips Ab. It not only rises to lift the centre knife off the two agates, but the two sloping arms simultaneously lift the FIG. 54. THE CHART. beam; seen more clearly in the ‘Wellcome’ figures below. 2 The Rothamsted Bate; see Chapter 4, p.46, Fig. 28. 3 This is chemical tinting, whereby the brass piece is dipped in a specific chemical solution, for which see Graham (1887), 67. Z1 Z2 T S1 S2 74 Stand. The stand comprises the base, the sleeved pillar topped by two Robinson Tulips, an inner and an outer, plus the arrestment mechanism including the two sloping arms. The brass stand base S1 is 57 mm diameter, 6 mm deep and neatly engraved with the copperplate inscription: RRoobbiinnssoonn 3388 DDeevvoonnsshhiirree SStt.. PPoorrttllaanndd PPllaaccee Unfortunately this has been partially obscured later with the crude addition of the crossed spirit levels; seen in Fig. 53 (as received, before conservation work). The levels are 36 mm long, 8 mm diameter, 9 mm at the ends. Both are intact. They are clearly early, with the screw adjustments for level clearly not thought out properly. They are neither Robinson’s type nor later Oertling-made additions. Also, it is hard to believe that Robinson would ever have defaced his earlier work in this way. Their maker is not known for certain, but the probability is that they are by R.B. Bate, judging by the design of his levels on the Bate/ Oertling balance at Rothamsted.2 FIG. 53. THE ‘STEEL SCALE’ STAND BASE. Whoever committed this outrage has clearly demonstrated that the real need for levels was soon perceived. Robinson later allowed for this with the rectangular base on his classic production design, to be covered in the next chapter. The pillar is held in the brass base with a single screw. At the top it expands to form the inner, fixed, Tulips Ab in brass tinted to a strong blue colour.3 The tops of the Tulips, front and rear, each hold an agate plane 6 x 3.5 mm held in place by red sealing wax. A horizontal circular ring 10 mm diameter in the same ‘blued’ brass connects the two uprights at the top, providing rigidity; the pointer descends through the wide hole in the middle. Aa and Ab show poorly here; instead see Figs. 58/59. The base of the inner Tulip carries a tiny lacquered brass chart C, a mere 9 x 7 mm, sparsely engraved as shown in Fig. 54 at approximately six times actual size. There are only two divisions each side of centre zero. All the lines are lightly engraved and unfilled. It is very difficult to read, not only by being so small and faint but is also effectively poorly lit inside the two Tu
75 The sleeve is lifted by an internal spring located somewhere at the top of the pillar, such that ‘arrested’ is the relaxed position. To release the balance, the arrestment lever T is depressed and moved slightly sideways to the R where it latches in a slot, now with the spring under tension.4 The beam descends and the centre knife rests on the two agates (a ‘divided’ centre plane) and begins weighing. This steel lever is 67 mm long, 2 mm thick and 3 mm wide, and passes right through the pillar and into a crude approximation to a ball-and-socket joint S4 mounted just behind the pillar as seen in Children’s sketch in Fig. 46. S4 was modified and placed a little further back behind the pillar in the production balances, as was shown in Fig. 43. The sleeve bears a short slot towards the top of each side, which allows it to rise and fall over the small pin S3 which passes right through the pillar and sleeve, protruding about 1 mm each side. This prevents the sleeve from rotating when the release lever is moved sideways. In Fig. 46 the artist has drawn the pin as almost falling out, and it is the same in reality on the Welcome balance in Fig. 57, presumably his model. Perversely, the pin in this ‘Steel Scale’ is impossible to move, so that the exact manner of the springing is still unknown. The artist has also drawn the lever too short. It has to protrude a little to be useable when the case is closed. The overall height from wood base to the top of the brass Tulip is 172 mm; from the six mm circular base to the underside of the Tulip boss 77 mm, and the brass Tulip including boss is 89 mm. [6 + 77 + 89 = 172] The arms, sloping at about 60°, are 96 mm long, tapering from 4 mm to 2 mm at the ends, with the cross pieces 7 mm wide. Pans. There are three lacquered brass pans as would be expected in a hydrostatic balance. The specific gravity (SG) pan U with its hook is as shown in Figs. 46 and 55; the short cords are missing, but the holes for them are clear in Fig. 55. The diameter is 44 mm, 1¾″. The two normal pans are 42 mm in diameter, on fine cotton white cords but very well fitted up. At the top the three cords are not only tied together in the classic manner, but then FIG. 55. THE SPECIFIC GRAVITY PAN. sealed with a tiny dab of red sealing wax. Where the cords pass through the pans, there is a dab of clear lacquer applied to the underside of the pan, sealing each cord in place, thereby preventing the fraying of the cords against the edges of the holes and creating cord breakages. This is a very rare refinement. The drop from the S hooks is approximately 6″. ‘Wollaston Pan Stabiliser’ alias ‘Wollaston Apparatus.’ This is exactly as seen in Fig. 47, from the Kater (1822) double reference, with one exception; all the surviving Robinson Wollaston apparatuses have two centres of motion, Z1 and Z2 36 mm apart instead of the single central one; see Figs 47 and 53. The motion of the arms is rather restricted with respect to the slots in the lantern front, which could have been cut a little wider. Further, the R arm is too near the 4 The slot is barely visible in Fig. 43, but shows clearly in Figs. 59 and 65 in Chapter 10.
76 liquid jar when in hydrostatic weighing mode. This is less of a problem on the 8″ and 10½″ versions of the production balances. The front arms measure about 107 mm (both outer ends are damaged; see photos); the rear arms are 39 mm long. Both front arms were cut short, but the L arm has a piece riveted on, binding on its case slot while the R arm has been left slightly short but perfectly useable. The vertical brass pins are 20-22 mm tall and at a separation of 35 mm. The Case (lantern + base + drawer). All parts of the case are fitted together with some very fine tongue-and-groove work, and glued. There are no signs of screws, pins or dowels. The only exceptions are on the drawer, of which the front piece is held with hidden dovetails, and the rear board with simple dovetails. The detachable lantern V is nominally 9″ x 8″ x 4″, created in 5/16″ (8 mm) gauge dark mahogany, rigidly glazed front and rear. The glass is of the period, slightly ‘wobbly’. Each pane is held in by four strips of matching mahogany with the front edges rounded, and mitred at the corners. More precisely, the lantern including the upper slight overhangs is 232 mm wide, 206 mm high and 107 mm deep. Each end panel has a solid hinged door W, nominally 5½″ x 2¾″ (140 x 70 mm), with the leading vertical edge chamfered by 2 mm to create a close fit. It is greatly to the credit of Robinson’s cabinet-maker that both doors are still a perfect fit 200 years later. All the mahogany has been sanded to an exceptionally fine smooth surface, both inside and out. All surfaces have been French-polished just to a silk finish. There is an amateur non-French-polished repair to the lantern at the front, top L corner, and a poor Araldite resealing of a parting at the rear. The top of the lantern has a large ivory knob X with which to lift it off the base; nominally 1″ diameter, actual 28 mm. This impressive three-part knob is probably not the original, but it is all of real ivory and therefore almost certainly an early replacement no later than WWI. The lantern has three pegs on the underside, locating it accurately in three holes in the base; two L and one R, which show clearly in Figs. 51 and 52. The lantern is firmly clipped to the base with a single brass hook on each side, V1 and V2, engaging tightly on a steel screw. The L and R hooks work in opposite directions, possibly as an attempt at a safety measure. The two end doors each have a single retaining brass hook engaging a steel pin, set loosely. None of the hooks and pins is original, as the originals were of a much smaller size as shown by the location and diameter of the nearby tiny holes. The significance of these points will become clear in Chapter 22 on Robinson’s balance cases. The lower front member of the lantern has a shallow slot for each of the Wollaston levers to protrude for outside operation, but they do not line up well with the levers. There is a thick L-shaped front cut-out as seen in the photographs, to allow the release lever T to latch sideways. It is 20 mm high, 5 mm wide at the top and 11 mm wide at the bottom. The base is nominally 9″ x 4″ x 1½″, actually 230 x 106 x 38 mm, in the same dark mahogany. The original adjustable brass feet, as per the sketch, are now missing, but the three pairs of screw holes remain underneath. The drawer Y is 212 x 98 mm, laid out as shown in Fig. 56. It is lined with deep navy cotton velvet, almost black. The SG pan goes at the bottom of the pan pen with the hook in the central hole; the other two pans then fit easily on top of it. Steel forceps of a strange type are present, but seem unsuitable for this application. The original Robinson screwdriver is present in its slot. The large rectangular pen carries a latchable weights box with four recesses for fractional weights, now missing. The tiny weights pen is only 11 x 13 mm, and seems to be for something like lead shot; see
77 Fig. 60 for a similar item in the ‘Welcome’ drawer. It is curious that there is no pen allocated to the principal weights, but note the wear on the velvet to the left of the pointer, suggesting where they might have been. It is too soon to guess the weights used; this will be pursued later in Chapter 23. Originally the drawer was fitted with retaining hooks but these and the matching pins in the base are now missing. The drawer is nominally 1″ deep, actually 27 mm. The drawer pull Y3, of 12 mm dia., is almost certainly of ivory but is not the original; it is too modern. [T. ALLGEIER. FIG. 56. THE INTERIOR OF THE ‘STEEL SCALE’ DRAWER. Work Done. This balance was in a very distressed state when received. At some date after its retirement, the beam had its fittings removed. At a later date someone, with no understanding of balance beams, reassembled the beam fittings upside-down and back-to-front, rendering it unusable. The vertical adjustment screw at the L end and the flag post were both missing. It subsequently became clear that the pointer had lost 5 or 6 mm at the top end; this is something about which nothing could be done. The hole for the pointer had been filled in with a tiny scrap of brass, which has had to be removed with a ½ mm slitting saw, creating a slot. The upper end of the pointer had been filed by someone to a flat which was fortuitously a weak interference fit in the slot, but a smear of Loctite has provided it with a permanent grip. The balance was dismantled as far as possible and much was learned about its construction, but pin S3 refused to budge. For the most part the conservation was perfectly straightforward, in accordance with the advice in Oliver (2020), Chapter 41. The centre knife D was found to be in a very good condition; it was removed and replaced the correct way up. A replacement flag post with a flag of platinum wire was fitted. The last problem was the missing fine screw for the vertical adjustment of the L swan neck S5. Eventually it was replaced with a cut-down 11BA bolt. In this condition the balance was found to be remarkably sensitive. With the lantern in position 1 mg sends the pointer gently off-scale, and 2 mg does so rapidly, which is highly satisfactory after 200 years. The lever-lock arrestment system needed no adjustment and works perfectly, provided one keeps a tight hold of the lever and does not allow the spring to take over; this point needs to be emphasised.
78 THE ‘WELLCOME’ ROBINSON. Science Museum accession no. 1977-159; was Wellcome no. 219. Very little needs to be said about this scale, for it is not only a very close match to the two sketches in Kater (1822), but is also an almost perfect twin of the ‘Steel Scale’ with most dimensions agreeing to within ± 1 mm, except for the case dimensions which are well within ± 5 mm, and S1 being only 53 mm diameter instead of 57.5 The principal differences are: 1) The beam is of brass, the conventional metal for small scales and balances; 2) It is as built as originally intended without spirit levels; 3) The divided centre plane is of steel; 4) It retains its original brass adjustable feet; 5) The release lever, ball-and-socket joint, flag and lantern are all missing. [J. WATERS. FIG. 57. THE ‘WELLCOME’. 5 Despite this the engraver has managed to give the address fully as Robinson / 38 Devonshire Street / Portland Place S5 A2 A1 S1 Z S6 V1 A2 Aa S3 Ab Y 3 3 3 Z Y2Y1 Y3
79 [J. WATERS. FIG. 58. THE ‘WELLCOME’ BEAM AND TULIPS. The inner Tulips Ab are of a slightly paler blue-grey colour than in the ‘Steel Scale’, and the ring connecting the two is more clearly seen in Fig. 59. The two arms A2 are currently loose and in need of tightening up. The stand sleeve pin S3 can be seen almost falling out. The pointer B is 7 cm long and its tiny gravity bob B1 is present as before. The upper part of the pointer has a piece of wire wrapped around it, seemingly an amateurish way of correcting the very undersized gravity bob. Note the small holes to receive the locating pegs on the FIG. 59. THE ‘WELLCOME’ BEAM AND THE PAIRS OF TULIPS. underside of the missing lantern. See too the side hook V1 for the clip-on lantern. The three brass feet are of the traditional offset design, with the knurled disc component above the offset fixed part, in the reverse of 20th century practice. The switch from four feet to three is immaterial; in fact levelling a balance with three feet is easier than with four. The drawer Y is nominally 9″ x 4″ x 1½″, actually 225 x 105 x 40 mm, and fitted out for the beam and pans, with each compartment lined with navy blue velvet, much as on the ‘Steel Scale.’ Two hooks Y1 and Y2, not in the Children sketch, have appeared to prevent the drawer from sliding out during careless transport. The ivory drawer pull Y3 has lost its button. S5 S6 A2 A1 F B1 S3 B A2 Aa AbAa Ab C
80 [J. WATERS. FIG. 60. THE INTERIOR OF THE ‘WELLCOME’ DRAWER. As can be seen in from the layout in Fig. 60, the beam and pointer are stored in one piece, exactly as before. Two brass pans and two brass Specific Gravity (SG) pans are currently stored in the drawer. One of the SG pans is an intruder. The weights pen contains a curious selection of masses, none of which would appear to be the originals. The pile of small, rough balls, possibly lead shot, is to be discussed in Chapter 23. The drawer also holds what appears to be a small weight lifter 80 mm long, which fits neither any of the weights present nor the ‘lead shot.’ A note on pan suspension cords. Both of these scales use the traditional suspension of the pans by three cords set symmetrically around the pan at 120° intervals. This was the traditional scale- maker’s method reaching far back in time, used on money scales, tea scales and the like. However, the cords seen here on both scales are cotton replacements for the silk originals. Note that normal twisted silk is unsuitable for this application; it has to be of the ‘plaited’ variety otherwise it will tend to form knots when stored in a drawer, however much care is taken. Silk was used because it is much less hygroscopic than cotton, and is therefore unlikely to promote rusting of any steel part from which it is suspended. The most usual colour for silk cords was a dark shade of green. No such corrosion occurs where the cords enter the tiny holes on the brass pans, but there is a problem here too. Where the cord passes through the hole in the brass pan, it tends to fray on the sharp edges of the hole and eventually break. Where the cotton cords on the ‘Steel Scale’ pass through the pans, there is a dab of clear lacquer applied to the knot on the underside of the pan, sealing each cord in place, thereby preventing motion and the fraying of the cords against the edges of the holes and creating cord breaks. This is probably a customer modification as it has not been seen before. A better way is to tie each cord to a tiny brass ring or steel S-hook passing through the hole in the pan, so the cord is contacting only a curved smooth surface. This was the standard method for all pans suspended on platinum chains. Robinson seems never to have adopted this method for silk cords, or used platinum chains. Replacement of broken cords is difficult but possible; Diana Crawforth has given full instructions in EQM, but the author has never attempted this operation.6 6 EQM (No.3 1986); 928.
CHAPTER 8 The ‘Wollaston’ Robinson HIS BALANCE belonged to the pioneering chemist William Hyde Wollaston (1766―1828), according to R.T. Gunther. Once a part of the Wollaston Collection, it was transferred in 1951 from the Cavendish Laboratory to its current location in the Whipple Museum, now with accession number Wh.1109. 1 [WWW.JENEMANN.ORG. FIG. 61. THE ‘WOLLASTON’ MINUS ITS LANTERN. [ See Fig. 125 in Chapter 22 for the lantern.] 1 Gunther (1923), 30, item 36. Wollaston Collection number G249. T N G A4 T VF A3 B1 C S2 Z B Y3 A S3
82 Whilst being obviously closely related to ‘The Wellcome,’ the design has evolved into a true balance, with relief of the suspension points on the beam. The 5½″ (139 mm) brass open beam has a rectangular centre section carrying the 10 mm steel centre knife which rests upon a divided centre plane of agate, each piece 6 x 3 mm, held by red sealing wax. The beam carries a screw-threaded 65 mm steel pointer B with a tiny gravity bob B1, hanging above a short, silvered brass chart C within the Tulip. The flagpost F carries a tiny platinum wire flag, invisible in Fig. 61. The beam now has the classic Robinson slit N at each end, although it must be said that they are very wide and poorly made here; clearly an early attempt, and executed much better in all the following examples. Two 9 mm steel end knives have now appeared, with pressing screws, as was shown in Chapter 5. The beam is more slender than the typical Robinson, tapering from 2 mm at the centre to 1 mm and back for the slit ends. That really is very thin, potentially easily bent sideways by careless handling. The two arrestment arms A2 of the ‘Wellcome’ have now been linked horizontally to create the rigid triangular arrestment frame A3 common to all later Robinsons and, in one form or another, to most other makers’ balances for the next century. The horizontal element has perforce to divide and pass around the upper part of the Tulip; again pioneering, and later amended. The fixed Tulip is in lacquered brass, not a blued structure as on ‘The Welcome,’ and the outer moving Tulip has vanished. Kater conceived an improvement to the very short knife and plane of the 18th century Cavendish end bearings: two 9 mm steel knives with trunnions (stubs), with each knife carrying a first attempt at an end plane; see Figs. 62 and 63. This is, with its own stubs, a single U-shaped piece of steel, the underside highly polished to provide the bearing surface, but cut away as shown in Fig. 62 to keep the contact area to a minimum. That was soon seen as a mistake, corrected in the next balance. The underside was fitted with a first attempt at a J-hook, excessively curled, again pioneering and soon corrected. [DRAWINGS BY THOMAS ALLGEIER. FIG. 62. ‘WOLLASTON’ END PLANE, UNDERSIDE. FIG. 63. DITTO, SIDE VIEW. (All dims. in mm.) When the release lever T is depressed against the arresting spring it pulls down the sleeve S2 carrying the arrestment frame A3 with it, thus putting the beam into operation; the lever is then slid sideways to the right and latched, leaving the operator free to use both hands. The lever pivots behind the stand in a curious structure which is a crude ball and socket joint; later developed. All this is carried
83 over unchanged from the preceding swan-neck scales in Chapter 7. The Sleeve Pin S3 part way up the sleeved pillar of the ‘Steel Scale’ and the ‘Welcome’ is again present, although it barely shows in Figs. 61 and 64. The diameter of the sleeve is only 7 mm, cf. 8 mm on the scales. The brass pans are each 45 mm in diameter. They are suspended on dark green plaited silk cords, as are all the later Robinson production balances. There is no ivory tip/grip/handle to the lever, and it is unclear whether there was one at this early stage in Robinson’s designs. As can be seen in Fig. 64, the lever protrudes only 9 mm from the cabinet, and it looks very short without such a handle, especially when compared with later versions. The circular brass base S1 of the stand is 58 mm in diameter, 2¼″, and engraved as before without the ‘, London’ and is held in place with two blued steel woodscrews. The use of the fully adjustable Cavendish cradles in the arrestment for the first time ensured that the three knives and three planes all contacted at exactly the same place at every release, ensuring a higher reproducibility. The use of the Robinson slit at the beam ends together with pressing screws on the end knives pioneered the routine fine adjustment of knife-edge beams. The cradles are each attached to the arrestment frame by four screws of which three are adjustable, clearly allowing yet more precision adjustment, to get the two cradles exactly horizontal and level with each other and horizontal with the centre knife. [WWW.JENEMANN.ORG. FIG. 64. THE WOLLASTON WITH ITS LANTERN. X W Y2Y1V1 Y
84 The cabinet has a detachable lantern as before, held on with two hooks V1 and V2, and located with three pegs fitting into three small holes (2L, 1R) in the top board of the drawer base. It has only solid end doors W. There is a hole in the top of the lantern for a lifting knob which was present in 1969 but not in 1976. The brass (?) knob X is visible in Fig. 64; it is now still missing but marks on the lantern top show that it was about one inch in diameter. The Wollaston Collection serial number of the balance, G249 in white paint, is still present just behind the knob position. The lower rail of the front of the lantern has two shallow cut outs from which the two Wollaston stabiliser handles protrude. It also has a central vertical L-shaped slot for the release lever, with the short horizontal extension to the right to allow the release lever into its latched position. All these features are visible in Fig. 64, although the handles are tucked inside. There is an excellent photograph of the separated lantern and balance in Chapter 22, Cabinet Styling. This balance has retained its brass feet. The drawer Y is held closed as before by two hooks, Y1 and Y2. It contains an SG pan, a spirit level, copper (?) 110 mm forceps with double-cut (file-like) marks on the outer surfaces (not inner as expected), the original screwdriver in its slot and two loose 100 grain knobbed brass weights.2 The fractions pen is empty. Most unusually it has managed to retain its original 15 mm diameter ivory drawer pull Y3; cf. the original sketch in Fig. 46. This was a very popular 19th century style of small knob for drawer pulls, for all makers. The ‘Wollaston’ balance incorporates five important structures ― slits, pressing screws, end planes, cradles and a triangle ― to optimise performance, almost but not yet quite the full Robinson design. It follows the ‘Steel Scale’ in using agate for the divided centre plane. All three of these pioneering instruments, the ‘Steel Scale’, the ‘Wellcome’ and the ‘Wollaston’, have a descending arrestment ahead of their time, when most scales and balances were of the rising plane type. It is essential to emphasise that Kater and Robinson have made three great leaps forward here: First, they have created the triangular arrestment frame, later often known simply as the Triangle. Second, they have then fitted the triangle with Cavendish cradles, thereby arresting the beam and the arrestment frame separately in the theoretically correct manner. Third, they have created an early version of a J-hook suspension piece, which was subsequently improved rapidly into its conventional form. Almost all other commercial makers took up the triangle and J-hook concepts, but none took up the Cavendish cradles.3 This latter is greatly to be regretted, for it means that many high-precision balances, of both this period and later, have not been able to reach their full potential due to less than perfectly precise placement of their three knives upon their planes. To summarise, the ‘Wollaston’ contains several original attempts at optimising performance, not yet perfected but clearly pointing the way forward. As Wollaston died in 1828, this balance cannot be of a later date, and was probably made a few years earlier, but later than 1821/2; let us say c.1825. It would be interesting to perform some loaded sensitivity trials on those two early scales, the Wollaston balance and the later surviving Robinsons, but the author feels that the silk cords are now too fragile for this to be attempted safely. 2 The masses were each confirmed as exactly 100 grains on a portable electronic scale. 3 None in Britain or Europe to the best of the author’s knowledge, only Duffey in Philadelphia; see Chapter 22, pp.129-130.
CHAPTER 9 The ‘Miller’ Robinson HIS BALANCE once belonged to Professor William Hallowes Miller, FRS, (1801―1880), and is currently in the Whipple Museum, accession number Wh.1413. Miller was a fellow of St John’s College at Cambridge University and Professor of Mineralogy there 1832―1880. He wrote extensively on crystallography and mineralogy, and to a lesser extent on mathematics, hydrostatics and hydro- dynamics. He was notable for his great ability in constructing highly accurate apparatus from simple materials.1 His career took a new direction in 1843 when he was appointed to the parliamentary committee for the preparation of new standards of length and mass, following the destruction of the old standards in the 1834 fire of the Houses of Parliament. Miller was largely responsible for the creation and adjustment of the new standards of mass using several Robinson and Robinson-style balances, as will be discussed in Chapter 16. This balance was transferred to the museum in December 1961 from the Mineralogy and Petrology department of the University, where it was reputedly used by Miller for work on the lesser masses involved in the 1843 work. The ‘Miller’ builds very firmly on the advances made with the ‘Wollaston’; see Figs. 66 and 67 overleaf. Most importantly, the Robinson slits N are now properly made although more open than on later balances; their formation is still in progress. The R one has been opened very wide. The pressing and pushing screws are as before, and the cradles too. The suspension pieces M have now developed as shown in Fig. 65, having achieved the simplified form which became the standard ‘J-hook’ for many makers for decades to come, until the introduction of the stirrup suspension piece circa WWI. The clear agate planes R2 are 10 x 4 mm, held in their brass holders with a clear adhesive.2 The width across the trunnions L2 to L2 is 18 mm. [DRAWING BY THOMAS ALLGEIER. FIG. 65. THE ‘MILLER’ J-HOOK. 1 The biographical data is largely from McKie’s DSB biography, and Miller’s obituary in Nature 22 (1880), 247-249. 2 At this date the clear adhesive is probably, as before, a cheaper form of isinglass. T L2 R2 L2
86 The ‘hook’ is now a loop of a single turn, which is all that is required. The J-hook suspension pieces M, M are hidden behind the cradles G, G in Fig. 67. The spring-loaded lever-lock release is as before, pivoting in that strange structure behind the pillar which is slightly changed with some type of internal guide, just visible in Fig. 43.3 Significantly, pin S3 has vanished. The lever protrudes 25 mm, much further than on the ‘Wollaston’. Also present and unchanged are the ‘Wollaston pan stabilisers.’ The brass pans are slightly larger than before, now 47 mm in diameter, one of which is an SG pan with an underside hook, suspended on short cords, as seen in Fig. 67. All the cords are of pale green plaited silk. [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 66. THE ‘MILLER’. The obvious visual difference from the ‘Wollaston’ is that the same open beam has now become captive, with a continuous centre plane R1 passing through a light cut in the rectangular central feature of the beam. The centre plane is of clear agate 18 x 5 mm, again held by a clear adhesive. This plane has to be removed before the beam can be extracted, and is held in place by the two enormous finger bolts Q, Q which really catch the eye and are one of the most distinctive features of fully developed Robinson balances. The steel centre knife D of the beam is 13 mm long. The beam length, k-k, is ‘obviously’ the same 5½″ as before, but in reality is 149 mm, which is almost 6″. Other later Robinsons are the same, and it will be recalled that in the 1842 posthumous Robinson catalogue, the 5½″ had been changed to 6″, which we can see now was simply a coming to terms with reality. The beam is still of the same delicate tapered form and dimensions as in the ‘Wollaston.’ Again it is fitted with a screw-threaded pointer B, 55 mm, carrying a very small gravity bob B1 hanging just above a silvered brass chart C. 3 There was insufficient time to strip the balance down on the day and examine this, so its working is still unclear.
87 Also as before there is a wire flag F for poise adjustment, although not visible in these photographs. Recall that the horizontal element of the triangular arrestment frame A3 in the ‘Wollaston’ divided and passed clumsily around the upper part of the Tulip. In the ‘Miller’ the horizontal element curves around only the rear of the Tulip, seen clearly in Figs. 66 and 67. This too looks awkward but is sufficiently rigid to work perfectly. [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 67. THE ‘MILLER’, WITH AN SG PAN AND A NORMAL PAN SUSPENDED. The cabinet is as before, with a hole in the top board for a lifting handle, now lost. It again has two end doors in solid wood, likewise the fixed front window with the same slots for the ‘stabiliser’ handles and the release lever. Curiously, a hole has been bored through the base and the drawer just beneath the L knife. The ‘obvious’ interpretation is as an adaptation for under-the-bench weighing, but this seems unlikely as the balance capacity is only 26 grams. Specific Gravity work under the bench also seems unlikely as an SG pan has been provided. One possibility is that it is for the determination of the SG of a liquid housed in some special container too large to put into the balance. Hooks Y1 and Y2 were provided to keep the drawer in position as before, but the drawer seems to have been only partially lined in navy-blue velvet. It contains two single-piece knobbed brass masses, marked 100 and 200 grains, each in its own well.4 Also present is a spare J-hook and another 110 mm forceps of the strange type seen in the ‘Wollaston’. 4 These were found to be 100 and 199 grains on a portable electronic scale. Q, Q N N G SG PAN A3 G B B1 C Y1 Y2 T Y3
88 There is no ‘fractions’ box and no ‘fractions’, and there is no longer a slot for a screwdriver. The ivory drawer pull Y3 has lost its button. The stand bases in the previous models were all circular and engraved with Robinson’s name and address by a professional engraver, as was shown earlier. In the ‘Miller’ this base S1 has now become rectangular, 53 x 43 mm, plus two rear lugs 12 mm deep and 8 mm wide. The base bears two spirit levels at right angles, ‘crossed levels.’ This format for the stand base, but with slightly different dimensions, is retained in all later Robinson balances. It is odd that whilst the rear level rests on the two lugs, the L one rests on the stand base. This makes it look like an afterthought, yet the rear lugs are a part of the base casting, clearly designed in. Unlike all later balances, here the L level does not eclipse the L blued steel woodscrew holding the stand to the wooden baseboard. Robinson’s name and address continues to be engraved professionally as before, but now includes ‘London’ on this and all later balances. Unusually, Robinson’s name is engraved in a slightly wide upper case serif typeface exactly as in Robinson Hass 2, seen in Fig. 73. ROBINSON 3388 DDeevvoonnsshhiirree SStt PPoorrttllaanndd PPllaaccee,, LLoonnddoonn Any features of the balance not noted here may be taken to be as in the ‘Wollaston.’ In conclusion The Miller has further advanced the attainment of high precision by several innovations, as follows. First, the use of a continuous centre knife working on a continuous centre plane. This avoids the difficulties of aligning the two halves of a divided centre plane, and the longer contact line ensures lower pressures in use, which should extend the working life of both knife and plane. This also enhances the value of the precision adjustment of the two cradles. Second, the clumsy first attempt at an end plane has been replaced by the true J- hook suspension piece as used by Robinson’s immediate successors R.B. Bate and Ludwig Oertling, and almost every other balance-maker for many decades thereafter. Third, all three planes are now of agate, the standard format for Robinson and all other makers of high-precision balances for decades to come. Finally, the Robinson slit is now close to its final form.
CHAPTER 10 The ‘German’ Robinsons HE AUTHOR believes there are several Robinson balances in Germany, but has been able to find only the two in the private collection of Matthias Hass. Nothing is known of their provenance. For convenience here these two will be called Hass 1 and Hass 2. Hass 1 appears to have come from a museum or other collection, and had an accession number PH SC 33, so far untraced. All photographs in this chapter appear by courtesy of Matthias Hass. FIG. 68. HASS 1, FRONT VIEW. T Y2Y Y3Y1
90 FIG. 69. HASS 1, OVERHEAD VIEW. FIG. 70. HASS 1 MINUS ITS LANTERN. W V1
91 There is another overhead photograph of Hass 1 as Fig. 126 in Chapter 22, showing the cabinet even more clearly. The two instruments differ only in their cabinets, and the way the ‘Robinson’ name has been engraved on their stand bases. The actual balances appear to be identical 5½″ Robinsons, exactly equivalent to the ‘Miller.’ FIG. 71. HASS 1, DRAWER. FIG. 72. HASS 1, STAND BASE. FIG. 73. HASS 2, STAND BASE. As can be seen from the photographs, Hass 1 and Hass 2 both have the inconvenient library front door, hinged R, but whereas Hass 1 has the wooden end doors W, Hass 2 has only solid ends; front access only. The R hook and pin V2 to hold the lantern onto the base can be seen clearly in Fig. 74; it also shows the front door hinges and the R hook and pin Y2 to secure the drawer. Figs. 70 and 75 show that only one pin and hole were given each side for locating the lantern of these two balances to their bases, instead of the usual one plus two. Another difference between them is that Hass 1 has the front feet set conventionally L―R whereas Hass 2 has them set unusually at 45°. The drawers are much as before, but as usual each drawer pull has lost its button. The two Continental masses are replacements for the original 100 and 200 grain masses. The ‘fractions’ box contains five recesses and would have supplied masses from 40 grains down to thousandths of a grain, with each recess containing four geometric masses (4, 3, 2, 1) at each order of magnitude; see the Weights chapter. These and the two principal masses are all that were needed for a balance of only 400 grains capacity. Currently the ‘fractions’ box contains only Metric masses.
92 The instruments are as alike as it is possible to be with hand-made items, but there is one notable unexpected difference in the engraving on the stand base, as shown in Figs. 72 and 73. Hass 1 has the normal inscription, but Hass 2 has the word ‘ROBINSON’ engraved in serif capitals, as in the ‘Miller’. The rest of the address is slightly rearranged; presumably the work of a different outside engraver. Fig. 75 shows Hass 2 without its lantern. It is clearly identical to Hass 1, but the beam seems to be a little displaced and is slightly off- centre in the photo. The gravity bob is missing from the pointer, but the flagpost is clearly present at the apex of the beam in Fig. 74. FIG. 74. HASS 2. FIG. 75. HASS 2, WITH LANTERN REMOVED. V2Y2
CHAPTER 11 The ‘Used by Miller’ Robinson HIS BALANCE appears to be the earliest 10½″ version discovered to date, and is simply an expanded version of the 5½″ ‘Miller.’ It was transferred to the Whipple Museum in August 1971 from the Mineralogy and Petrology department of Cambridge University. It now has Whipple Museum accession number Wh.1593. All the 5½″ balances noted above were built to carry only 400 grains (26 grams), but the capacity of the 10½″ model was increased five-fold, to 2000 grains (130 grams). As a result, the entire balance has been strengthened, but it is still very much an enlarged ‘Miller’ and everything not mentioned below may be taken to be much or exactly as in the ‘Miller.’ [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 76. ‘USED BY MILLER’ 10½″ ROBINSON. Professor Miller reputedly used it for adjustment of a standard 1 lb mass. This was impossible as the capacity of Robinson 10½″ balances was rated at 2000 grains, but a 1 lb mass weighs 7000 grains. This balance could of course have been used in the T
94 adjustment of the masses of some of the smaller subdivisions of the pound. Presumably someone has confused this balance with the Robinson & Barrow 15″ balance which Miller did use for weighing a standard 1 lb; see Chapter 16. The most obvious difference is that with the lengthened beam, 265 mm, a vertical brace has been added to each arm of the previously open beam to ensure perfect rigidity. It now therefore classifies as a lattice beam. The thickness of the beam has been only slightly increased, being 2½ mm at the centre, but still tapering to almost 1 mm before returning to 2½ mm for the slit ends. The Robinson slits are here very well made and have now achieved their final true form. Alas, the R slit has been over-opened and bent, effectively bending the beam and making it unusable and uncorrectable. What remains much as before are the silvered brass chart and the revised ball-and- socket joint as in the ‘Miller’. Also remaining are the wire precursor of the flag, and a screw-threaded pointer, now 90 mm with a larger gravity bob 10 mm in diameter. A less obvious change is that the clear agate centre plane has been enlarged to 26 x 10 mm to accommodate a longer steel centre knife of 26 mm. The centre plane is again held by a clear adhesive. The end knives are also extended, to 13 mm, being 20 mm across the stubs. Therefore the knives are now nominally 1″ and ½″ respectively, becoming the standard sizes for all the 10½″ Robinsons from here on. Also enlarged are the brass pans, now 75 mm in diameter, together with a 68 mm brass SG pan. A notable find on this balance concerns the design of the cradles. Whilst the L and R front, and R rear cradles are as before, the inner face of the L rear cradle has an extra V-slot lower down, to engage the knife stub L1, giving even more adjustability. Alas, no photograph of this is available. This feature has not been noted on other, later, Robinsons, but it just may have been missed. The J-hooks currently fitted are later Oertling replacements, one of their standard 19th century styles. This style is essentially just a copy of Robinson’s J-hook suspension piece shown in Fig. 65 in Chapter 9.1 Each J-hook carries a clear 18 x 6 mm agate set in a brass holder with red sealing wax, and the holders are engraved L and R in the usual Oertling manner. This does show that once Kater and Robinson had created their final J-hook, it remained the usual style, unchanged until supplanted by the stirrup suspension piece circa WWI. The rectangular base is 70 x 58 mm excluding lugs, and carries what appears to be an engraved serial number, 517, in addition to the usual signature and address, including London as on the ‘Miller.’ No serial number has been found on any other Robinson or Robinson & Barrow balance, so there must be a question mark over the significance of ‘517.’ It could be an owner’s inventory number, but these are usually crudely punched on balances generally, not engraved. As with the smaller rectangular base of the ‘Miller’, the rear spirit level is on lugs and the L level is mounted on top and partly eclipses the L screw head, but no part of the engraved inscription is covered. An obvious difference from both the earlier and later Robinsons is that there are no ‘Wollaston pan stabilisers’ present, and no screw holes suggesting that a set had ever been fitted, and there are no slots in the door for the handles. 1 For Oertling suspension pieces generally see Oliver (2020), 123, 126, 133. The other 19th century style noted above is not shown in Oliver (2020), but is intermediate between the two shown there, having a J-hook but located by vertical pins, rather than horizontal stubs.
95 As before there is a removable lantern, but it is fitted with a full-width front door closing under the top board, very much as the one on the ‘typical’ Robinson balance in Chapter 5 and the Hass balances in Chapter 10. Again this one is hinged on the R, but such doors are most inconvenient to use. The lantern still carries the two very nicely made solid side doors which are also somewhat unfriendly for the user. The drawer was fully fitted out but some parts are now missing, most notably the velvet pad to carry the pointer, when the beam is in its nicely laid out section with blue velvet lining, and held in place with ivory turnbuckles. The pen for the masses has clearly been removed, and no masses remain. The two brass pans with their green silk plaited cords and their J-hooks are sitting in their pens as intended. The drawer has its own hooks Y1, Y2 against accidents en route. Although bigger than before, this balance is still light enough to carry easily. Although much wider, the drawer carries only a single ivory pull Y3, rectified to two in later examples. Once again, only the stem is present, the button having been lost. As can be seen in Fig. 76 there is damage to the case just under the drawer, but the drawer itself is undamaged. A strange double-ended fork-like brass object in the drawer is a weight lifter 98 mm long overall, 15 mm wide. This is for handling knobbed masses to avoid any contamination from bare fingers or the clumsiness imparted by gloves. The forks are of two different spacings, 5 and 8 mm, for smaller and larger masses. There is more on the manipulation of masses in Chapter 23 (Robinson’s Weight Sets). There are some interesting observations to be made from Fig. 77.2 FIG. 77. AN ENGRAVING OF UNKNOWN DATE OF A ROBINSON 10½″ BALANCE. 2 Allgemeine Encyklopädie Der Physik, Leopold Voss, Leipzig, 1869; volume I, p.580, Fig 55.
96 The date is unclear as is the origin. The book was published in Germany in 1869, long after Robinson’s death, but the original engraving must be much earlier, and is almost certainly English. The author suspects that this is the earliest representation of a 10½″ Robinson balance. The gravity bob on the pointer is tiny, as is found only on the earliest Robinson balances. The drawer pull is of the type shown in Children’s sketch in Chapter 6 and Fig. 61 (the ‘Wollaston’) in Chapter 8, again indicating a date very early in the 1820s. The engraving shows the fore-aft spirit level on the R of the stand base, whereas all other known Robinsons in reality or photographs or drawings, have it on the L. The full width library door is hinged on the L, something never seen in reality. Perhaps this engraving was created before the design had fully settled or, more probably, the engraving was reversed before the lettering was added. The principal features are all much as the ‘Miller’ but stretched, becoming very close to Wh.1593. The artist has cunningly drawn the Cavendish cradles on the R beam end, but omitted them on the L to show the J-hook and end knife, although not very clearly. The close-ups 55b and 55c are much more accurate. Note too that there is no ivory tip to the release lever, and it begins to look as if this feature belongs only to the Robinson & Barrow era. What its appearance in this encyclopaedia does show is that Robinson balances were valued for many years not only in the UK but also in Europe.
CHAPTER 12 The ‘Cambridge University’ Robinson HIS WAS SAID BY STOCK to be a 10½″ Robinson similar to the one above, but incomplete and damaged. This one has not been examined by the author or Buchanan, but Stock did see it and reported it, though only very briefly as follows, when talking about Professor Miller:1 He was Professor of Mineralogy in the University of Cambridge and is much better-known to present-day scientists as the originator of the “Miller indices”. Closer inquiries in Cambridge showed that the Department of Mineralogy still has a damaged and incomplete 10½-in. Robinson balance. Fortunately, Robinson was in the habit of engraving his name and address upon his instruments. Stock says no more about it, so presumably he did not consider it was worth investigating alongside the complete ones he had found. Thus it is, or was, in the University Department of Mineralogy, but is now missing. No photograph or drawing of it is known to exist. This is not the missing ‘Robinson & Barrow’ in Chapter 16, for that one was a 15″, a much bigger instrument. 1 Stock (1968), 254. T
98
CHAPTER 13 The ‘Rothamsted’ Robinson HE ONLY KNOWN 8″ surviving Robinson balance currently resides in the internal museum at Rothamsted Research near Harpenden, Herts. It has inventory number BA104. This is the intermediate size built to weigh 1000 grains (65 grams) and resolve to 0.001 grain, as claimed by Robinson in his early catalogues. This is also very much a ‘stretched’ version of the 5½″ Miller, and has therefore the added braces on the beam as also appear in the 2000 grain 10½″ model. This is the balance selected as a typical Robinson for Chapter 5, and everything in that chapter may now also be regarded as the first part of this chapter. FIG. 78. ROTHAMSTED ROBINSON 8″ PARTIALLY DISMANTLED; GENERAL VIEW FROM REAR. All photographs in this chapter appear courtesy of Rothamsted Research / Lawes Agricultural Trust. T
100 In this second part we will look mainly at how the design has progressed from that of the ‘Miller’, and finally at the small, unique, distinctive feature of the beam which was quietly passed over earlier. The author was fortunate in being allowed to take this balance apart for critical examination, and Fig. 78 shows BA104 in a dismantled state, with most of the parts arranged for examination and photography. This photograph has been taken from the rear, and the ‘Wollaston stabilisers’ have been pulled backwards. The stand is almost exactly as earlier versions, and the release lever is again pivoted in an approximation to a ball and socket joint behind the pillar. This release is not spring- loaded like the other Robinsons, unless the spring has failed, but it does make for a far smoother release/arrest action. The lever is 130 mm overall. The stand is shown loose with the three steel woodscrews, blued but now rusting, arranged in their true order on the paper in Fig. 79; F = front. The pillar is fastened to the stand base by a single bolt which is so big that the wood baseboard has a recess to receive it. The two spirit levels are mounted on the stand base as before, but this time the L level totally eclipses one of the blued steel wood screws, and even eclipses a small part of the finely engraved Robinson address. The three steel knives are: centre 19 mm, ends 11 mm. The agate centre plane is 20 x 8 mm, and the agate planes on the suspension pieces M, M (‘J-hooks’) are 12 x 5 mm; all three planes held by a clear adhesive. The beam now has a proper 17 mm brass flag F. The brass gravity bob B1 on the 72 mm threaded steel pointer B is now of a more sensible size and has acquired a locking ring B2. The tiny chart C is again of silvered brass. The dished brass pans are 55 mm diameter and suspended on short cords, so short that the pins of the ‘Wollaston Stabiliser’ pass under the pans in a useless manner. There are two more brass ‘pans’ 51 mm diameter without holes for cords. These are not pans at all, but ‘skiffs’ for weighing chemicals on, analogous to the later clock glasses but much lighter. They are rather corroded, unlike the true pans, showing that they have done their job. The dishes pass between the cords with only slight tilting, confirming the interpretation. An SG pan was also provided, as shown. Much of the non-lacquered brass is showing corrosion; note particularly the inner faces of the cradles, and the rear foot. Likewise, the steel knives, the pressing and pushing screws, the J-hooks and the pointer are much rusted. This shows up badly in Figs. 80 and 81. The handles of the ‘stabilisers’ are badly corroded where they have been left poking out of the case during lengthy storage. Undoubtedly this is due to the lacquer having been worn away by frequent use, leaving unprotected brass. That BA104 was heavily used is not in doubt, as the tip of the release lever has gouged a deep groove in the front edge of the wooden base; this is very visible in Figs. 32, 43 and 45 (in Chapter 5) just above the key. The recent history of BA104 is unknown, although there is a possibility that it is ‘ex-Kew’, but Kew records were unable to find any mention of this instrument. The suspension pieces M,M are in the form of ‘J-hooks’ bearing agate end planes on brass flats fitted with stubs for the cradles to catch, and a steel J-shaped hook. The brass pans are suspended from the J-hook by dark green plaited silk cords. At first glance the cords look to be originals by virtue of the lack of wear as they pass through the holes in the pans, and that they have been professionally tied at the top end. So, the cords must have been shortened, but it was done professionally.
101 The 8″ beam is shown clearly in Fig. 79. It measures 196 or 197 mm k―k, which is 7¾″ rather than 8″. By far the most interesting feature of this beam, and so far unique, is the additional reinforcing insert N1 behind the Robinson slit; see the close-up Figs. 80 and 81. Clearly a piece of metal shim has been inserted from the end, but why? A metrological engineer’s opinion is that this is a standard engineering practice where a soft and malleable metal is used in a rather weak construction, as here, but this is an unusual application of the principle. A slit is cut and a piece of stronger and more ‘springy’ metal inserted and thoroughly silver- soldered in. It is important that the two metals are completely soldered together face-to-face and not just with spot contacts. The insert would usually be steel or bronze. It is difficult to tell in this case, but bronze seems the more likely. The insert provides resistance and elasticity under stress where the softer metal would merely deform and stay deformed. There is no doubt that the brass of the Robinson beam end is highly stressed around the slit. The twin pressing screws I are the main adjustment for knife parallel and arm length, and the effect of the pushing screw J is to move the end knife upwards and outwards by only a few microns, with the insert providing resistance for the screw to work against. FIG. 79. ROTHAMSTED ROBINSON 8″ BEAM AND POINTER, PANS, CORDS AND J-HOOKS. FIG. 80. THE ROTHAMSTED ROBINSON 8″ BEAM END, TOP VIEW. J P I N N1 L1 E O1
102 FIG. 81. THE ROTHAMSTED ROBINSON 8″ BEAM END, END VIEW. Finally on this topic, it needs to be considered whether this rare inserted shim method of reinforcing the Robinson slit was a one-off experiment or a very brief production method. With only one example seen so far, it seems safer to assume it is an experimental version, until such time as another one turns up. A distinctly better reinforcement method ― stiffening screws ― appears in one of the Robinson and Barrow balances (Chapter 17) and would have been easier to make. The drawer contains the best preserved set of Robinson weights observed so far, which will be considered in Chapter 23 with all the other Robinson weights. N1 M
CHAPTER 14 The ‘ex-Edinburgh University’ Robinson HIS IS ANOTHER 10½″ Robinson balance, which appears to be absolutely identical to the 10½″ Robinson ‘Used by Miller’, above, except for the added ‘Wollaston pan stabilisers.’ This balance is on loan from Edinburgh University to the National Museum of Scotland, where it now has the accession number IL.2010.19.2; it was previously 1970L2. [© NATIONAL MUSEUMS SCOTLAND. FIG. 82. THE EX-EDINBURGH UNIVERSITY ROBINSON. Superficially little needs to be said on this balance itself, being so similar to the ‘Used by Miller’, but there is one extra item which must be an owner’s later addition; there is a simple Oertling-style riderslide over the right arm of the beam. The engraved divisions for a rider are only on the R arm, and are engraved 1 to 9, but unfilled. T
104 These are the first and only (so far) rider graduations to appear on a Robinson beam. Stock, who has examined this balance, thought the engraving ‘appears amateurish when compared with the engraving on the base.’1 Indeed they are very amateurish, but this is the case with some very early Oertling graduations, because they were made on the beam in-house, not by an outside professional engraver as for both Robinson and Oertling stands. The probability is that at some later date a riderslide was bought from Oertling and installed by the University. The beam graduations were probably engraved by a local man, for if Oertling had done it the graduations would have all been rather better, and filled black. This balance briefly: the stand has the usual Robinson signature with ‘London’; beam 261 mm, which is 10¼″; 86 mm pointer with a good-sized gravity bob, steel knives; agate planes; J-hooks, and platinum pans 55 mm diameter hung on green silk cords. The cabinet is superficially identical to that of the ‘Used by Miller’ with a removable lantern. Again it has a user-unfriendly full-width front door hinged on the R with a pair of side doors. As before, the full-width drawer has hooks to prevent disasters in transit, and the ivory drawer pull has lost its button. There is another photograph of this cabinet in Chapter 22. [© NATIONAL MUSEUMS SCOTLAND. FIG. 83. THE EX-EDINBURGH UNIVERSITY ROBINSON DRAWER. The platinum of the pans has been shown by semi-quantitative XRF analysis to be of a very high degree of purity.2 Whether these are of plated base metal or stamped out of thin platinum sheet is an unresolved question for now. If the latter, these would be a first, for the few surviving platinum pans of the period are thick, heavy castings, such as those noticed in the tale of Professor Story-Maskelyne in Chapters 18 and 21. The indicated purity of the platinum of the four sides of the two pans varied from 98.0 to 99.4, with the remainder mostly copper plus traces of iridium; see the Weights chapter for the details. The original Robinson geometric wire masses are likewise composed of platinum of a high purity, from 97.5 to 99.3%. These too are discussed in more detail in the Weights chapter. 1 Stock (1968), 255. 2 Troalen (2025). XRF stands for X-ray Fluorescence Spectroscopy, for which see the Weights chapter.
105 The incomplete set of weights in the 4-3-2-1 grains system is rather odd, with the expected knobbed brass masses and platinum wire masses visible in Fig. 83, and added flat brass masses tucked into the velvet lined grooves meant for the missing forceps. Briefly, the masses are, in descending order: knobbed brass masses 1000/400/300/200/100; the square flat brass masses 100/90/80/70 (plus W&T Avery 60 or 50?); platinum wire geometric masses 40/30/20/4/3 (10 is missing); playing card brass masses 6/5/3/2/1; and the tenths, hundredths and thousandths of a grain are all missing. [© NATIONAL MUSEUMS SCOTLAND. FIG. 84. TWO PARTS OF THE WEIGHT SET. (The upper right ‘intruder’ is presumably 60 or 50 gr.) John E. Mackenzie D.Sc., Reader in Chemistry in that Department of Edinburgh University, writing to Johnson Matthey in November 1935 and January 1936, notes a 10½″ Robinson at the University, carrying solid platinum pans. The identification as platinum rested on the determination of the density of one of the pans at 21.536. He mentions another ‘in our Natural Philosophy Department’ with conventional brass pans, implying but not stating directly that it is also a 10½″, and saying ‘it is not such a fine instrument.’ Johnson Matthey were unable to date the pans for him. Which of these two instruments is this one, and where is the other? Pans can of course be interchanged, and solid platinum pans are likely to be greatly valued and re-used over and over again, transferring from balance to balance as the years pass, before finally disappearing. See for example chapters 18 and 21 recounting the tale of Professor Nevil Story-Maskelyne and his platinum pans with their probable travel from Oxford to London. Edinburgh University says it has no knowledge of the other Robinson balance, which would appear to have absconded with the solid platinum pans.
106
CHAPTER 15 ‘Dr Moll’s Balance’ alias The ‘Robinson-Olland’ HE PREVIOUS CHAPTER concluded the research on the small Robinson scales and routine production balances, but Robinson is known to have made a few much larger bespoke balances, to be considered in this and the following chapter. Until 2024 not one of the large Robinsons was known to have survived, but Dr Moll’s balance was rediscovered on a Dutch website by a Dutchman, Ritzo Holtman, a serious collector and researcher. As the balance stand was engraved ‘Robinson’, therefore of English origin, and had clearly been modified, Holtman did not buy it. Instead he managed to trace and interview the owner, inspect and photograph it, identify it, find that it had come from the University of Utrecht many years earlier, that the cabinet was not the original, establish connections with the University museum curator and ultimately enabled the balance to be returned to the museum to the curator’s delight. All this he has reported in an extensive paper.1 Professor Gerard ‘Gerrit’ Moll, (1785―1838) is best described as a Dutch polymath. Born in Amsterdam, his family background was in commerce, and his early education was directed towards him joining the family trading business. To that end he studied English, French and German, but he became drawn to mathematics, science and the classical languages instead. He took up astronomy seriously in 1801, and assisted his tutor J.F. Keijser in making observations. Keijser taught him geometry and algebra as well as astronomy. Still suspended in his education between trade and science, on a trading trip to London in 1804 he bought a 10″ sextant from Edward Troughton’s workshop. This was probably his first scientific instrument. His father enrolled him into the Athenaeum Illustre of Amsterdam studying under, inter alia, J.H. van Swinden (1746―1823). It was van Swinden who convinced father Moll that his extremely talented son should devote himself entirely to his scientific interests and not commerce. Yet at the Athenaeum he was also taught Latin, Greek, law, history and antiquities. A key step in his development came in 1806 when he was asked by van Beek, the Professor of Mathematics and Physics, to assist in the determination of the distance in longitude between Amsterdam and Utrecht. In 1809 at Leiden University he gained a bachelor’s degree in philosophy, which at that date included the sciences as we have seen earlier. In 1810 he went to Paris to further his knowledge of mathematics and astronomy under Delambre. Here then was an exceptionally talented and very widely educated young man in his mid-twenties, already experienced in astronomical and geodetic observations and, presumably, the related data reduction. Where would he go next? 1 Holtman (2024). T
108 He became director of the Utrecht Observatory in 1812, and in 1815 was appointed Professor of Mathematics and Natural Philosophy at Utrecht University, holding these two positions until his death. He was also awarded an honorary doctorate in 1815, which is why he is often known as ‘Dr Moll’ even publishing under that title although already a professor. As he is most often referred to as ‘Dr Moll’ that is how he will be referred to hereinafter. His early language education directed towards a commercial career actually served him well, enabling him to publish technical papers in those three foreign languages as well as his native Dutch. He published widely on astronomy, the speed of sound, electricity, magnetism and hydrology, but the standard reference works seem largely to ignore his work on various aspects of metrology, that side of his work most relevant to this study. Dr Moll’s Original Robinson Balance. Exactly how and why Dr Moll became involved in the inter-comparisons of various copies of the International Prototype Kilogram and other mass standards is not made entirely clear, but van Swinden is the link. Needing a balance large enough to take a kilogram, Dr Moll ordered a high capacity balance from T.C. Robinson of London in 1826 or 1827. He gives a very detailed description of the Robinson balance and the results he achieved with it in a lengthy Dutch paper. Here the portions relevant to the balance are translated, with some explanatory notes in [brackets].2 The balance I have been using, was made according to the ideas of Mr Kater by the artisan Robinson in London. However it differs somewhat from the balance described by Kater in the Philosophical Transactions of 1826. Mine is closer to the one recently described by Kater (*). The beam is made of a single piece of bell metal and is, including the pans, lightweight, but strong and rigid, and it has the appearance of the beam of a steam engine, it has a length of 18", and everything that was not needed to keep it strong has been filed away. The pans hang on steel prisms which rest on flat agate planes, each with a length of 1". The centre axis also consists of a steel prism, of which the downwards pointing sides form an angle of 120 degrees. This prism has a length of 2½" and rests with its total length on a flat agate plane. A pillar carries the agate plane, on which the knife alias prismatic axis of the balance rests; the balance is levelled by means of two spirit levels. The whole instrument is enclosed in a glazed case, and can also be used for hydrostatic experiments. On the outside of the case is a crank which can arrest the balance and the pans, lifting the knives off the agate planes. A very ingenious apparatus, invented by Dr Wollaston, will reduce the oscillations of the pans and quickly bring the balance to rest. At each end of the beam is a sharp steel pointer, running over an ivory chart, each having 20 divisions. These charts are attached to the case, and each [chart] contains 20 divisions. The chord of the arc is 9/8", so that each division measures .0562". Protruding above the top of the lever [beam] in the middle is a thin steel post with a finely cut screw thread; two brass balls of different sizes can be screwed up and down, and serve to heighten or lower the centre of gravity of the balance at will. There are still other features present, to change the distance between the mounting points of the pans [the end knives] and the axis [centre knife], but a description of all this, without a picture, is difficult to comprehend. When this balance is loaded in each pan with 1 kilogramme or approximately 15432 [troy] grains, it is clearly moving with the smallest available [Robinson-made] weight, that is .005 or 5/1000 grain, so with less than a three millionth of the total load. On the [balance] by Fortin, used to determine the [International Prototype] kilogramme, you could only weigh up to one millionth part of the kilogramme, and not better. Moll also gave a brief description of his chosen balance in his own English words, here showing how the van Swinden connection gave rise to this kind of work: 3 The late Professor Van Swinden was one of the Commissioners sent by this country to the general meeting at Paris, and in consequence he became possessed of a set of authentic copies of the metre and kilogramme, on which he always set the highest value, and which were constantly 2 Moll (1831a), 119-155. Parts translated by Ritzo Holtman. The item marked (*) refers to the balance in reference 5 below. 3 Moll (1831b), 64-65.
109 preserved with the greatest care. After Mr Van Swinden’s demise they came into my possession [by way of a public auction]. Being furnished with one of the well-known balances by Mr Robinson of Devonshire Street, and with very accurate British troy weights by the same artist, I was anxious to compare Mr. Van Swinden’s authentic copy of the kilogram with the British troy weight… The beam of the balance which I employed is about eighteen inches in length; the knife-edge of the axis is about two inches and a half long; the knife-edges on which the scale pans are suspended are about one inch. When loaded with a kilogramme in each scale, 5/1000 of a grain are sufficient to show a derangement in the equilibrium of the beam. 5/1000 of a grain is 0.3 mg, implying a sensitivity of one part in three million, three times more sensitive than Robinson’s routine production offerings. Moll further describes how he used his Robinson balance six times between 1827 and 1830 to weigh van Swinden’s personal brass copy of the standard kilogramme and compared it with the Robinson English Troy weights, to find out that an English Troy pound (= 5760 grains) weighed 373.244 grams. He also used it to make numerous comparisons of various standard masses, generating vast amounts of data. But why would Dr Moll buy from an English maker when there were noted makers nearby in Europe, such as Becker, Fortin, August Oertling, Pistor and Repsold? He was probably influenced by Captain Kater’s great paper of 1826 on the construction and adjustment of the new Imperial Standards, wherein he describes work on smaller balances for the smaller masses, praising two Robinson balances and one by Bate.4 All three must each have been a ‘larger balance’, but nothing further was said. Kater later praised a Robinson balance in a major text of the time.5 So Dr Moll ordered a large balance from Robinson in 1826 or 1827 and was already using that instrument at Utrecht University in 1827 for weight comparisons. The subsequent history of the balance has been determined by Holtman as far as is now possible, and is given succinctly below. After the death of Moll in 1838, the kilogram comparisons were performed on a better instrument created by C. Becker of Groningen. Unlike the Robinson with its beam-end pointers, the Becker had a vertical pointer running over a chart set low on the stand. But the Becker balance was quickly abandoned and was replaced in 1844 by an even better balance, one built by the Amsterdam instrument maker Eduard Wenckebach. Holtman managed to track down the curator of the Utrecht University Museum, Paul Lambers, and through him to examine the records of their Robinson balance. It was mentioned in the University Year Book of 1832. By a convoluted set of events it made its way to the Physics lab of Utrecht University, next to but not a part of their scientific instruments collection. It then seemingly vanished until reappearing in a report by J.A.C. Oudemans in 1887. It still resided in the Physics lab [Natuurkundig Kabinet] of the university. It had clearly been neglected, so Oudemans had the instrument serviced by the Dutch instrument-making firm Olland, removing the rust and sharpening and re-aligning the knives. He then determined the sensitivity of the balance per chart division, going from 2.55 mg unloaded to 23.34 mg at a load of 1kg. The sensitivity of approximately 0.3 mg at a load of 1kg, as noted 60 years earlier by Moll, was not possible anymore. Even so, the Robinson balance was considered suitable for hydrostatic weighing and was used to weigh several standard kilograms in distilled water. It is difficult to see the point of this in view of the greatly inadequate and highly inconsistent sensitivity. The balance was also used for demonstration purposes.6 4 Kater (1826). 5 Kater (1831), Chapter 21 only, ‘On Balances and Pendulums’. 6 There is the possibility that the balance sensitivity was deliberately lowered for demonstration purposes.
110 This instrument had been bought in 1838 and given inventory number 633 in the Physics lab catalogue. It appears again in the 1905 catalogue and is described as a ‘balance in mahogany case max. 500 gram’.7 Two remarks were made: 'In 1893 reassembled by Olland.’ 'Case of this balance in use at the Ph. practical Jan. 1953.’ What is clear at this point is that Dr Moll’s balance went twice to Olland for serious conservation work, first in 1887 and again in 1893. Lambers produced the index card for this Robinson balance, found in the 'Ornstein' card index box of written-off equipment of the Physics lab of the University. It now had the number 3455 but with a cross reference to 633. The index card makes it clear that in January 1953 the glazed wooden case and the balance were in separate places, confirming the note in the catalogue. The case was in use at the Physics lab, while the travels of the balance could be followed between October 1953 and January 1960. The entries then stop, with a reference to a new index card which has not yet been found. The balance case was not saved and the balance itself was seemingly lost or written off. Holtman was then able to establish the later history of the balance up to the present day by some clear thinking and by tracking down the recent owner. But first, what can be deduced of Robinson’s balance as originally delivered to Dr Moll, when there is no engraving or later photograph? To put it another way, what are we looking for? Fortunately Moll’s descriptions in the quotations above give much essential detail. From an understanding of this it should be possible to deduce how the new ‘Robinson’ differs from Dr Moll’s original. Are they the same? His tedious description can be summarised succinctly in modern terms as follows, and a few useful points from Oudemans’ report are included. The beam is a single piece of brass or bronze, filed away to form a lightweight lattice 18" long, but just strong enough to weigh a kilogram. At each end of the beam is a sharp steel pointer, swinging over an ivory chart. These charts are attached to the case, and each chart contains 20 divisions.8 At the apex of the beam is a thin screw-threaded steel post carrying a two-part gravity bob. [There is no mention of a flag.] The centre knife is a steel prism 2½" long, facetted to an angle of 120°, resting for its total length on a flat agate plane on top of a pillar. There are two spirit levels, locations unspecified, for levelling the instrument. The end knives are steel prisms each 1" long, with four pressing screws around each end knife to reduce or extend the arm length [and adjust parallel]. The pans are lightweight, yet strong and rigid, hanging from flat agate planes. The whole instrument is enclosed in a glazed cabinet. It can be levelled with the aid of two spirit levels [suggesting adjustable feet, but no feet are mentioned]. The cabinet is suitable for hydrostatic weighing by virtue of having a 'lower floor’, presumably in the manner of the Ramsden cone beam balance as noted in Chapter 4. A crank on the outside of the case gives full arrestment of all knives and planes. The case is fitted with the ‘Wollaston Apparatus’. 7 The ‘500g’ is probably an error made by a clerk unfamiliar with balances, who made a guess based on its size; ignore this. 8 In the quotation the curious phrase ‘has the appearance of the beam of a steam engine’ refers to the massive lattice beams then in vogue in steam engines for pumping water out of mines.
111 The New ‘Robinson―Olland’ Balance. Holtman was able to make some essential deductions from examination of the rather poor original photographs, long before contacting the owner and the museum curator. By virtue of possessing an extensive collection of European balances and balance-makers’ catalogues, he was able to spot that the arrestment handle, the pans, pan hangers, the gravity bob, pointer and the chart were all typical of the style used by the Dutch instrument-maker H. Olland, who worked in Utrecht between 1853 and 1896. Olland received his training from the instrument- and balance-maker C. Becker of Arnhem. The vertical pointer was very much in the style of Olland, and also of the pointer described on the C. Becker balance which first replaced the Robinson at Utrecht University; see p.109 above. As can be seen later in Figs. 96 and 97, the chart has been insensitively screwed to the original Robinson stand base, largely obscuring his address, and is in a style typical of both Olland and C. Becker. The flag post now has a flag as well as a huge gravity bob in three parts. The stand has a Troughton- style ‘railway arch’ supporting the centre plane. FIG. 85. DR MOLL’S REBUILT ROBINSON BALANCE IN ITS NEW CABINET. [All photographs in this chapter are by Ritzo Holtman.] Holtman was allowed to visit, inspect and photograph the newly found ‘Robinson’ balance in a village north of Apeldoorn. The owner inherited the balance some 45 years ago from an almost unknown uncle, and it was surmised that this instrument had been used in the Physics Lab of Utrecht University. When the balance reached him it stood on an oak block, from which it easily toppled over. While cleaning the instrument one of the intermediate links in the length extenders between the suspension piece and the pan hanger was lost; the owner made a new link in a different colour but with the right dimensions, as can be seen in Figs. 92 and 93.
112 Unfortunately he had not seen the original balance case, but to avoid more problems he decided to build a mahogany case for the instrument, and this is how we find the balance today, as in Fig. 85. It has two library case doors meeting in the middle, obscuring the pointer and chart, but this was a fault found in many balance cases of the 19th century balance-makers. Holtman determined the length of the beam as 41cm, 16¼". He has also been able to say that this is definitely not an Olland beam. There is a tiny screw-threaded hole at each end, where a short pointer may have been, and this would explain the discrepancy between 16" observed and 18" originally if this really is Robinson’s beam. By critical comparison of styles in his balance collection and catalogues, Holtman now believes that the Olland modifications are probably not by H. Olland but more likely to be by his son and successor W.C. Olland. This he has deduced from three characteristic details.9 This is a good point at which to stop and see just where we are. The instrument shows a mix of styles and instrument-makers. Is it Dr Moll’s original Robinson rebuilt by Olland? Everything points in this direction: the owner believed it came from the University of Utrecht, where Dr Moll worked; the university records state clearly that Dr Moll’s balance went twice to Olland for serious conservation work; the stand base with the Robinson signature is definitely genuine; the beam is not an Olland type, but although totally unlike any Robinson beam seen before, it has the Robinson slit at each beam end; the arrestment frame carries Robinson/Cavendish cradles. Nothing seems to contradict the hypothesis. It is time to look more closely. 1) Capacity. The capacity of Robinson’s largest routine model, the 10½″, was 2000 grains. From Dr Moll’s figures and from Kater & Lardner, the copies of the Inter- national Prototype Kilogram weighed 15,432.3 grains. That is more than seven times the maximum for which Robinson was building before. Thus this new balance is far bigger and will therefore have to be of a significantly different construction, as seen in Fig. 86. So, the strangeness of the beam does not rule it out as a Robinson. FIG. 86. THE ROBINSON BEAM. 2) Beam Length. Now that it is concluded that this beam is actually the original, no matter how un-Robinson it looks from 1) above, it follows that the discrepancy in beam length is simply the current length of circa 16″ plus a 1″ pointer at each end, making it circa 18″. The threaded holes for the pointers are still visible; see Fig. 87. The charts for the pointers were fixed to the original cabinet, now lost. On really old balances like this, the beam length was in those days not considered to be of much importance, and was frequently expressed only roughly or not at all. The convention of defining beam length as end knife to end knife seems to be rather more recent. 9 The 'modern' numerals used on the chart, which are different from those used by H. Olland; the style of the gravity bob and flag; the pointer has no eye at 1/4 from the top whereas H. Olland pointers had an 'eye' at that spot.
113 3) Robinson Slits. The beam ends are in Robinson’s usual style with the slit ends for adjustment of the knife height and the arm length, plus the matching ‘Cavendish’ cradles as shown in Figs. 87―89. The two adjustments are, alas, partly interactive.10 Fig. 87 shows that the hole at the beam end is screw-threaded for the end pointer, now removed. It also shows that the trunnions (stubs) on the knife are chamfered to create a slight wedging action as they engage with the sloping inner faces of the cradle on arrestment. Fig. 88 shows one of the two V-cuts on the inner faces of the cradle to engage the lugs on the two tips of the end plane when arrested. The lugs and end knife are one piece of steel, and there are four pressing screws to adjust arm length and knife parallel instead of the two on Robinson’s production balances. Note the very three-dimensional strong form of the lattice beam in these images. FIG. 87. THE ROBINSON BEAM END. FIG. 88. FIG. 89. THE ROBINSON BEAM END, SIDE VIEW. Close examination shows one significant difference here from Robinson’s usual slits; the pushing screw to open up the slit is missing, but the stiffening screw is still present. This can be rationalised; at the massive loading of over 1kg per end, the suspended weight is trying strongly to close the slit, and the stiffening screw opposes it. Effectively the stiffening screw performs both functions here, although in these photographs it is clearly not engaged. 10 See the Oertling book, Oliver (2020), pp. 144-147.
114 4) Knives. Now we know that the original knives were 1″ and 2½″ and that the current ones are the same length, it seems reasonable to assume that all three knives are the originals. We have a clear statement from Dr Moll that his centre knife rested for its full length on an agate plane. What we see before us today is a centre knife ‘waisted’ which was presumably done by Olland in one of their two refurbishments of the balance. See Figs. 90 and 91. Their ‘waisting’ of the centre knife is most regrettable, as the waist puts the full load on only two short end stretches of the knife, enhancing the blunting of the knife edge under heavy loads. There were two philosophies about the loading of knives in the nineteenth century. For those readers interested, see Appendix 4. FIGS. 90 AND 91. THE ‘WAISTED’ CENTRE KNIFE. 5) Pans and Hangers. FIG. 92. THE OLLAND PAN, HANGER AND SUSPENSION PIECE.
115 The original pans and hangers were said by Dr Moll to be of bronze, but they must have been very stout and rather heavy, despite him describing them as lightweight. They must have been much heavier than the very lightweight Olland pans and hangers we see today in Fig. 92, looking very light for the application. Fig. 93 shows the suspension piece with its agate plane, chamfered lugs and extension piece to prevent the Olland pan hanger from touching the Robinson arrestment frame. FIG. 93. THE SUSPENSION PIECE PLUS EXTENSION. 6) Pointer and Gravity Bob 3 FIG. 94. OLLAND POINTER PLUS MOUNTING PLATE. 5 FIG. 95. THE BEAM AND NEW POINTER IN POSITION. The large three-part gravity bob is there to compensate for the extra weight of the new pointer and mounting plate, in order to keep the CG in the correct position. The flag post is threaded into the Olland-style fitting applied to the apex of the Robinson beam. The ways the new Olland descending pointer and its mounting plate are fixed to the Robinson beam are both very hard to understand; see Figs. 94 and 95. What exactly is going on here? It is so obviously a hasty modification which has not been thought out properly; all sorts of apparent errors
116 are present. ‘Cobbled together’ is probably a suitable description. In Fig. 95 the gleam of light slipping under the waisted part of the centre knife can just be seen as the knife rests on the centre plane. Note the finger bolts to hold the centre plane, and that one bolt head has been shaved to prevent it fouling the pointer. 7) Chart. The Olland brass chart was almost certainly originally silvered, the usual 19th century practice. Holtman has shown that the chart, see Fig. 96, is exactly in Olland’s normal style. It is to be regretted that Olland had so little respect for Robinson that they placed their chart so crudely over Robinson’s address, as seen in Fig. 97, but consider that some sixty years had passed since it was engraved. Robinson’s name had by then probably been forgotten in continental Europe, and he was passing out of the trade collective memory here too. FIG. 96. THE OLLAND CHART. 8) Engraved and Extra Baseplate. FIG. 97. THE ROBINSON ENGRAVED BASEPLATE.
117 FIG. 98. THE ROBINSON AND EXTRA BASEPLATE. FIG. 99. THE RACK AND PINION. The stand baseplate is 89 mm square, and engraved with Robinson’s name and address in his usual style; it is definitely original. But there is another plate below, as seen in Fig. 98, more than trebling the thickness to 22 mm. It provides room for the rack and pinion of the arrestment mechanism which in the original case must have been operating partly in the open space below. Because the stand is original, we may reasonably assume that the rack and pinion and the crankshaft are too, for no Robinson reference material, physical or literary, exists. This mechanism can be seen from the underside of the new baseboard, as in Fig. 99. There is evidence that the mechanism has had to be adjusted; observe the bright steel modern nuts, bolts and packing piece. This may have been necessary due to wear or, more likely, a slight mis-alignment after fitting into the new case. Holtman is of the firm opinion that the actual crank handle is an Olland replacement on the original shaft. 9) Stand and Arrestment. FIG. 100. ARRESTMENT FRAME AND PUSH ROD, ARRESTED. FIG. 101. DITTO, RELEASED. The first thing to note about the stand is the Troughton ‘railway arches’ supporting a centre plane passing through the centre light of the beam. The plane fits into the two recesses at the top of the arches, secured by two finger bolts from below, as in Fig. 95. In this heavy-duty balance Robinson has abandoned his usual sleeved construction, and introduced what later became the normal stand structure for most makers, with
118 a moving push rod inside a stationary hollow pillar. In this case a rack and pinion motion connects the crankshaft and crank handle to the push rod and arrestment frame. In Fig. 100 the frame is up, arrested, and in Fig. 101 is down with the balance on release. Is this how Robinson built his other large balances? It is another point to look for when identifying a suspected Robinson balance. 10)Spirit Levels. These are not the ones used in Robinson’s production balances, which are smaller and mounted differently. The levels here are much longer at 3″ and are fixed to the extra baseplate by a hefty lug at each end of a level; see Figs. 97 and 98. This is Robinson’s style on his optical and magnetic instruments, so is not a problem. These levels are probably the Robinson originals reused, but it is just possible they are Olland copies. Levels were not fitted on the Robinson baseplate, as there are no holes, so they must have been fixed originally to the cabinet baseboard. 11)The Cabinet. The recent owner has chosen his cabinet style well, using the ‘library bookcase’ style which Robinson is known to have used and was common in his period. Perhaps this was the exact style of the original cabinet, but we will probably never know. Robinson used a full-width single opening door on his production balances, but for this much larger balance two doors were required. Much more is said on cabinets and their doors in Chapter 22. 12)Recent ‘Conservation’. The recent owner has much to apologise for, as follows. First, for emery-papering all the surface corrosion and lacquer off the bronze and brass parts, leaving all totally covered in micro- scratches, as seen most clearly in Fig. 102. This image has been given excessive contrast to show the scratches more clearly on the printed page. Second, his use of four extra screws to hold the stand to the new case, when the four original screw holes were available, FIG. 102. THE MICROSCRATCHES. needing only four longer countersink screws. Why use domed screws, totally unsuitable for fine instruments? Third, the way he has chewed up so many of the screw heads. Fourth, and most importantly, in order to get complete coverage in de-rusting the knives and rubbing the nearby bronze areas, he must have removed all three knives, emery-papered them, and refitted them, not understanding that the alignments and spacing are ultra-critical, and that knives should never be touched by amateurs. The result is that this balance will never work properly again, unless someone can be found who can set knives accurately by the very demanding ‘Setting the Corners’ operation, as described in the Oertling book.11 The author has never dared to do this, although he has seen it done, twice, by an Oertling senior service engineer. 11 Oliver (2020), Appendix 11, pp.873-874.
119 Conclusions. How much is original Robinson? The cabinet and the feet are modern, but the balance parts can now be summarised thus: Olland: gravity bob, post & flag Robinson: beam pointer and mounting knives and planes pans and hangers J-hooks second baseplate stand chart arrestment mechanism & frame crank handle original signed baseplate levels (probably) As the surviving balance is about 90% Robinson it can now fairly take its place in the canon of surviving Robinson balances, now numbering ten. It adds a new aspect to our knowledge of Robinson balances, being the first large one to be identified. In particular, it gives researchers a view of what a large Robinson beam looks like. Recalling that Robinson did not sign any of his known beams, this could be very useful in identifying a Robinson large beam separated from its stand. There is the possibility that his other large beams and stands may have taken a different form, but it has to be hoped that Kater and Robinson were as consistent in their larger designs as they were with their small beams and stands. The Robinson-Olland balance now carries the Utrecht University Museum inventory number 0285-188871. Finally, the first line of Dr Moll’s lengthy quotation above is a very valuable incidental find, stating that this large Robinson was designed by Captain Kater and built by Robinson. This is how the early scales were reported in Chapter 7, and it has long been assumed that in Robinson production balances the Robinson Slit was Kater’s design, alongside his adoption of the Cavendish cradles. Thus it is now definite that for ‘Robinson’ we really should always be thinking of a Kater-Robinson partnership throughout Robinson’s career, with Kater as designer and Robinson as artisan craftsman. It is now clear that this collaboration was known in the trade at the time, but has long been forgotten because only Robinson’s name appears on their surviving instruments, whether optical, magnetic or metrological. In further confirmation of the above, the 1829 Robinson catalogue contains the following paragraph in the balances section: ‘The above balances have steel knife edges acting with agate plane surfaces, with crystal knife edges instead of steel each size 1 Guinea extra. Larger sizes are also made to order their arrangement embraces several important improvements by Captain Kater, Dr Wollaston and the Maker, all the size are furnished with platinum weights to 100 Grains, and upwards of brass. They will indicate .001 of a Grain with 1000 Grain.’ There appear to be punctuation errors which rather change the meaning. The author believes that a full stop should occur after ‘to order’, and another after ‘Maker’ instead of the comma. ‘Size’ should obviously be ‘sizes’. It then says that the production balances have improvements made by Kater, Wollaston and Robinson himself. These remarks may well also apply to the larger, bespoke balances made to order, but it is not made clear.
120 Other Robinson Large Balances. To date Dr Moll’s balance is the only Robinson large balance known to have survived, but there are others mentioned in the 19th century literature. a) We already have the two balances Robinson/Kater A and Robinson/Kater B dating from 1824, as reported on pages 43 and 44 above, and about which nothing further is known. b) There is a massive published archive of correspondence between the German researchers Schumacher and Gauss during the years 1808―1850.12, 13 Holtman has found a few nuggets relevant to Robinson balances within this enormous mass of research material, but most of the brief mentions of balances are of those made by Repsold and, to a lesser extent, other European makers. Schumacher owned a Robinson large balance, which is initially reported in the Schumacher―Gauss correspondence, where he replies to a request from Gauss for information on his balance as follows, in letter 353 on December 16th 1828, from Altona: 14 The balance is by Robinson. It costs, if I remember correctly (the papers are in Copenhagen) £26 or 26 guineas. [There is then a section on weights; see the weights chapter.] You can load each pan with 1Kg, so the total load is 2Kg. With one pound in each pan, and uniform temperature, the greatest (but very rare) deviations from the mean are 0.005 Grains i.e. 1/1152000 of the whole. (The average error of my weighings you have calculated yourself to be 0.00111 Grains at this load.) I don’t know how long you have to wait on average. I have waited more than one year. I am very happy to order this for you and [Captain] Nehus, who will travel to London in January, can chase it up. This is the first time we have any data on the performance and price of a Robinson large balance. The same balance was later reported in an English paper.15 As noted earlier, and again here, very little is said of the balance itself. All Schumacher says is: I BEING desirous of obtaining an accurate copy of the English Imperial Standard Troy Pound, for an intended comparison with our weights therewith, I applied to the late Captain KATER, and he had the goodness to procure for me not only a copy made by Mr BATE, exactly similar to those described in his paper, but also a balance of Mr. ROBINSON, of the same dimension and construction as that used by himself in comparing the legal standard in the custody of the Clerk of the House of Commons. So it would seem that Schumacher’s balance was a copy of either Robinson/Kater A or Robinson/Kater B, and it also appears that Schumacher did not use his Robinson balance after 1839. From the later silence on the topic in their correspondence it would seem that Gauss did not follow up on Schumacher’s offer of help, and never bought a Robinson balance. The present location of Schumacher’s Robinson balance, if it still exists, is unknown despite searches. c) Another missing large balance is the Robinson & Barrow noted in the next chapter. All attempts at finding more surviving Robinson or Robinson & Barrow balances, large or small, have proved fruitless. This included two world-wide calls for help on Rete, a message board for anyone interested in the history of scientific instruments. 16 12 Heinrich Christian Schumacher, 1780-1850; Doctor of Laws, Göttingen, 1806; Professor of Astronomy at University. of Copenhagen; Foreign Member of the Royal Society of London. 13 Carl Friedrich Gauss, 1777–1855. A great mathematician who was, amongst many other things, scientific advisor to the Danish and Hanoverian governments. He was especially interested in dealing with extensive quantities of numerical data, such as those arising from the then ongoing measurement of an arc of the Meridian, and those arising from comparisons of standard masses, both topics of interest to Dr Moll. 14 Peters (1860), 195. Excerpt translated by Thomas Allgeier. Altona was a borough of the city state of Hamburg, situated on the R bank of the river Elbe. From 1640 to 1864 it was under the personal jurisdiction of the Danish king. 15 Schumacher (1836), 457. 16 https://hsm.ox.ac.uk/mailing-list. There is no charge for posting questions or answers on this site.
CHAPTER 16 The Robinson & Barrow ‘Miller 1843―1845’ EADERS WILL RECALL from Chapter 3 that Henry Barrow bought the Robinson business in 1842. Barrow prepared a larger Robinson-style balance to be used by Professor Miller from 1845 onwards in the preparation of replacements for the Imperial Standard Troy Pound, and for other standards lost in the 1834 fire at the Houses of Parliament. This is the earliest reference to a Robinson & Barrow balance known to date. Despite intensive searching by Stock, Buchanan and the author, no trace of this important special balance has yet been discovered. No engraving of it is known to survive, but Miller gives a brief description of it in his massive paper of 1856, as does Chisholm.1,2 He quotes the bronze beam as being made sufficiently strong as to carry a kilogram in each pan. The steel centre knife is 1.93 inches long, resting entirely on a ‘quartz’ (read: agate) plane. The steel end knives are 1.05 inches long, carrying quartz planes. The beam length between end knives is given as ‘approximately 15 06 inches’. The beam was fitted with two end pointers, the R one carrying a tiny engraved ivory chart viewed through a compound microscope with a single horizontal wire in the eye piece. This proved so superior to noting the position of the L pointer against a fixed ivory chart by eye that usage of the L end was abandoned. It was then ‘discovered’ (his word!) that arms expanded with temperature changes, all recounted in detail. [Recall John Pond’s experience with the Ramsden balance somewhere between 1811 and 1835, on pp.50-51 above. Was Miller unaware of this?] The balance was enclosed in a glazed mahogany case, but few details are given. The microscope was fitted into the case opposite the R chart, and a glass screen was erected between it and the observer to reduce the thermal effects. As hydrostatic weighing was intended, there was a sliding brass plate under the R pan, and a corresponding hole in the table beneath enabling ‘below the bench’ weighing in a vessel of water, with a mass suspended on a wire hanging from a hook on the underside of a R pan. This is unlike the Ramsden and Troughton method with the balance on tall legs fitting into sockets on a board, as shown in Chapter 4. This instrument was usually referred to as the ‘Barrow’ balance, but the exact date of it is unclear. It was engraved ‘Robinson & Barrow’ so it must have been made after the take-over in 1842, and as was made clear by the Standards Commission, Miller was using it in 1845.3 The committee to oversee the construction of the new 1 Miller (1856), 762. 2 Chisholm (1877), 147-148. 3 Standards Commission Fifth Report (1871), Appendix V, 38. R
122 Standards was set up in 1843, so it presumably dates from 1843―1845. During this work two surviving old Robinson balances, a 10½″ and a 5½″, were used for the smaller masses.4 Nothing further is known of these. The Barrow balance was highly regarded, and was greatly praised by Miller:5 That part of the weighing which depended on the performance of the balances was most satisfactory. When the large balance, constructed by Mr. BARROW, was loaded with a pound in each pan, the probable error of a single comparison, by GAUSS’s method, was 0.00056 grain, or less than one-12 millionth of the weight in either pan; with a kilogram in each pan, the probable error of a single comparison, by BORDA’s method, was 0.00162 grain, or less than one-9 millionth part of the weight in either pan; by GAUSS’s method it was 0.00112 grain, or one-14 millionth of the weight in either pan. Thus it seems clear that Barrow was equally as good a balance-maker as Robinson. For the Gaussian weighing, various apparatuses for the interchanging of the pans were tried, some of which were described by Miller. He makes no mention of who devised the interchange apparatus, but only some details of the pans which were cleverly used.6 It all sounds rather home-made and amateurish, yet at least one clearly worked. The Barrow balance was later overhauled by Ludwig Oertling in 1868 and fitted with ‘improved arrangements’ devised by Miller meaning, amongst others, the facility to interchange weights without opening the case during further work on the Standards.7,8 The new pan-changer apparatus was therefore built by Oertling, but no description was given. There is a note on the revised Barrow balance in the Standards Commission 1871 report, in an account of their Reference balances. It was upon these Reference balances that the BoT Inspectors’ own Working Standards (‘outdoor’ weights) were checked against their own ‘indoor’ Standards; likewise any Local ‘indoor’ standards which were sent in after adjustment. Having given the lists of their many De Grave and Oertling Reference balances, other Reference balances in their possession were also listed, including the Miller Vacuum balance and these two: 9,10 (32) [The Robinson and Barrow]: A balance to weigh a standard pound or kilogram in each pan, made by Barrow, to turn with 0.01grain when loaded, fitted with two microscopes for reading the Index scales, and used by Professor Miller in his numerous observations for the restoration of the Imperial pound and its copies; with a mahogany case in which it can be packed for travelling. (33) ‘Captain Kater’s large balance with mahogany beam, for weighing a standard bushel, and the larger measures of capacity, when filled with water; to turn with a grain with 300lbs in each pan. Made by Mr. Bate, 1825, and reconstructed by Messrs. Ladd and Oertling in 1867.’ Except for the Kater and the Miller Vacuum, all these old balances seem to have disappeared. Alas, item (32) tells us little, but by quoting two microscopes, and the phrase ‘with a mahogany case in which it can be packed for travelling’, raises all sorts of queries. Chisholm says it was lent to the International Metric Commission in Paris for inter-comparisons of the new Standard kilograms, so it did travel abroad. It is interesting to see how, with the passage of time from the 1820s to the 1850s, Miller has replaced Kater as the intellectual driving force in British metrology. 4 Miller (1856), 765. 5 Miller (1856), 943. 6 Miller (1856), 764-765. 7 Standards Commission Fifth Report (1871), Minutes of Proceedings, l; Appendix V, 38, 50-51; Appendix XIX, 1-2. 8 The repairs are captured in the Oertling 1865―1876 ledger in an 1868 entry: Oertling (1865), f.367. 9 For the full list see Standards Commission Fifth Report (1871), appendix XIX, 1-3 or 226-228. Here the 1855 Oertling models are wrongly attributed to Ladd & Oertling, from whom Standards had recently bought masses and a few balances. 10 Perhaps more easily found in Oliver (2020), 257-258 and 519.
CHAPTER 17 The ‘ex-Stokes Collection’ Robinson & Barrow HIS BALANCE, now resident in the Whipple Museum, accession number Wh.1371, has come from the Stokes Collection via Cambridge University’s Cavendish Laboratory. It looks every inch a routine T.C. Robinson balance until one reads the engraved inscription on the rectangular stand base: RRoobbiinnssoonn && BBaarrrrooww 2266 OOxxeennddoonn SSttrreeeett LLOONNDDOONN.. Having the Oxendon Street address, this balance must date from 1843 (the date of the move there) to 1850, after which Barrow signed as ‘Henry Barrow.’ or ‘Henry Barrow & Co.’ [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 103. THE ‘EX-STOKES’ ROBINSON & BARROW 10½″ WITH LANTERN IN PLACE. T
124 [P.D. BUCHANAN. FIG. 104. THE ‘EX-STOKES’ IN DEMONSTRATION CIRCA 1976. At first glance it is exactly the same as the ‘Used by Miller’ in Chapter 11 but with ‘Wollaston pan stabilisers’ added. The 10½″ beam is actually 268 mm and the dished brass pans are 70 mm in diameter, suspended on dark green plaited silk cords. The Robinson slits are a perfect fine pair. Fig. 104 taken in 1976 shows the balance without the lantern, with Dr Peta Buchanan demonstrating how the pan ‘stabilisers’ were used in a correctly set-up balance. Little need be said on the technical details as they are mostly as on the late 10½″ Robinson ‘Used by Miller’ balance, such as three agate planes, but there are some detail differences introduced by Barrow. Thus the (very rusty) threaded steel pointer now screws into the beam, a simple practical improvement, together with a larger gravity bob at 9 mm diameter, plus a locking ring.1 The short ivory chart has six divisions engraved to both L and R of centre. The J-hooks are very close to Robinsons; the brass body is 15 x 5 mm, and the hook is a little longer at 19 mm. The stand diameter is 14 mm, a little more than Robinson’s. However, the ‘ex-Stokes’ does contain three further significant developments in fine balance design. First, it is important to note that there is no reinforcement of the Robinson slit by the inserted shim of BA104; instead a pair of stiffening screws J1, J1 have been added beneath the slit N, working in opposition to the pushing screw; see the Fig. 105. Experience must have shown that the basic slit could deform under heavy loads. Currently it is unclear whether the stiffening screw development was by Henry Barrow or by Robinson in his final years, as no such survivor by the latter has yet been found, but Barrow seems the more likely. In Berlin, August Oertling was using the Robinson slit from 1829 on, and he too stiffened the slit but with only one screw working in opposition.2 It is assumed he copied the basic slit concept directly from Robinson, as a search for a slit-end beam prior to 1820 has proved to be fruitless. Therefore it is now almost certain that Robinson and Kater were the true inventors of 1 The readily removable screwed-in pointer was later routinely adopted by Oertling. 2 Those Robinson slits and stiffening screws used by August Oertling in Berlin appear in Oliver (2020), 493-495.
125 the Robinson slit. Few of the later balance-makers used the Robinson slit, but J.J. Griffin was selling balances with slit-end beams by an unspecified maker as late as 1866.3 When Ludwig Oertling was setting up his workshop in London in 1846 he must have realised the deficiencies of the slit concept which had been in use for some twenty years; consequently he avoided it entirely. [ L. OERTLING LTD. FIG. 105. THE STIFFENING SCREWS FROM BELOW. Second, the flag has been replaced by a nut travelling along a short horizontal threaded rod tucked in one of the end lights of the beam. This can just be seen in Fig. 104 and in those of later balances. This is a far more precise way of adjusting the poise (zero), and is comparable with the technique used by later makers, e.g. by Oertling from 1914 onwards. Once again Barrow, or possibly Kater, was way ahead of the field. The third is an extremely simple but practical development. Although the idea of a cabinet with the very awkward hinged library front door has been retained, there are now two doors of unequal width, so that when closed the pointer and chart can be clearly seen head-on instead of being eclipsed by door frames which meet in the middle. The L door is 10½″ wide and the R 6″, both closing under the top board, clearly shown in Fig. 103. There are no side doors. In addition to the external hooks V1 and V2 holding the lantern to the base, this balance also has two internal large brass latches; all four latches are now broken. The top rear strip of the lantern is now totally loose; there is no rear doorslide. The lockable wide drawer was fitted with two ivory pulls, but only a single shaft remains. The lever protrudes from the slot in the larger door by 15 mm. Barrow not only continued Robinson’s tradition of supplying boxed sets of weights in the drawer, but still used the ‘Old’ 4/3/2/1 grains system for the principal masses. The brass knobbed masses of 1000/400/300/200/100 grains are still present, housed in a beautifully French-polished mahogany box 131 x 36 mm with a deep matching latched lid.4 It sits in the beginning of a pen. A brass lifter with a 5 mm slot is present, and is the correct one for those principal masses. The original 3 Griffin (1866), 38-39. Almost certainly these were European imports, with August Oertling as the most likely source. 4 The mass values were all confirmed on a portable electronic scale. E L1 H J1 N J1
126 fractions box is missing. Barrow has continued Robinson’s practice of supplying two pairs of forceps, but both are missing; one slot is now occupied by the lifter. Any points not mentioned here may be taken as being exactly as in the ‘Used by Miller’ balance. As this is the first surviving balance by Barrow, it is worth pausing here to sum up his changes to Robinson’s customary design features. Summary. The pair of stiffening screws J1, J1 have been added beneath the slit N, working in opposition to the pushing screw, to prevent possible deformation with heavy loads. This is of course an admission of a deficiency in the Robinson slit concept, and can be seen as hastening the end of using such slits. It can be argued that the concept is fine with light loads, as in assay balances, but slits in chemical balance beams are clearly potentially problematic. The flag has been replaced by a nut travelling along a short horizontal threaded rod tucked in one of the end lights of the beam. This is a far superior method of adjusting the poise. There are now two doors of unequal width, so that when they are closed the pointer and chart can be clearly seen head-on instead of being eclipsed by door frames which meet in the middle. The pointer is much improved by screwing directly into the lower rail of the beam, by the fitting of a much more practical (larger) gravity bob, and the bob being fitted with a locking ring. Of Barrow’s changes, the revised pointer construction and the poise adjustment became standard for later makers, so he has made significant contributions to fine balance design, even though he missed the doorslide.
CHAPTER 18 The ‘ex-British Museum’ Robinson & Barrow HE BRITISH MUSEUM has lent this fine example of a 10½″ Robinson & Barrow balance to the Science Museum, where it has accession number 1977- 340. It belonged to Professor Story-Maskelyne when working at the British Museum, after he had left Oxford University in 1857. It was bought from him by the British Museum (Natural History) in 1881, and was still in non-critical use there as late as 1970.1 It may have served for a few years more before its transfer in 1977. Superficially this balance appears to be exactly like the ‘ex-Stokes’ 10½″ by Robinson & Barrow, but closer inspection shows several differences and major improvements. It does however declare its identity in exactly the same terms as the ‘ex-Stokes’: RRoobbiinnssoonn && BBaarrrrooww 2266 OOxxeennddoonn SSttrreeeett LLOONNDDOONN.. Again, having the Oxendon Street address, this balance dates from 1843―1850. It is in beautiful condition. From here on it will be referred to as the ‘ex-BM’ for brevity. The author was engaged by the Science Museum to perform some conservation work on this balance (and others) in November 2000. Beside the conservation work the author was able to ‘play’ with this balance for a while, and experience the feel of using the lever-lock release and the ‘Wollaston stabilisers.’ Frankly the latter seemed to be more of a hindrance than a help, an opinion seemingly shared by all later makers and buyers. The opportunity was taken to examine many features of this balance and record them; some extra measurements and observations were made, many of which are listed succinctly below. Because Robinson and Robinson & Barrow have been conservative in many respects (as well as being simultaneously progressive in others), it is hoped that many of these observations will be relevant to several of the earlier 10½″ instruments, and possibly others too. Cabinet: The most notable feature of this balance is its great weight, for the stand has been bolted onto a thick slab of slate, 467 x 182 mm, which is recessed into the usual style wooden base having the normal full-width drawer. The stand is fixed to the slate by three brass bolts and captive brass nuts recessed into the underside of the slate. The slate has been polished on its top surface only, and might easily be mistaken for a 20th century black plate glass insert; see the images overleaf. The slate is affixed to the cabinet by only two blued steel woodscrews. 1 Stock (1971), 386. T
128 Despite the massiveness of the balance the lantern is still of the same clip-on variety as before. This is not a good situation when it comes to moving this very heavy instrument. A rigid cabinet would be expected, which is what it looks like until one spots the hooks, and there are only two of them. Even worse, neither hook is currently working, so the lantern is dangerously loose. [J. WATERS. FIG. 106. THE ROBINSON & BARROW 10½″ GENERAL FRONT VIEW, COMPLETE. [J. WATERS. FIG. 107. THE ROBINSON & BARROW 10½″ MINUS BEAM AND PANS.
129 Another surprising feature of the cabinet is the fact that although the lantern is detachable, it has a rear doorslide rather in the manner of later 19th century rigid balance cabinets. Yet another surprise is that the doorslide is counterpoised by two lead weights suspended in black-painted cylindrical steel tubes in the rear corners of the lantern. This is very much in a Continental 20th century style. The strings are modern, and the thought now arises that the counterpoises may perhaps be modern retrofits. As can be seen from the figures, the two front library doors are separated by a fixed centre pane, giving a clear view of the pointer and chart. There are also two glazed end doors very much not in the Robinson tradition of solid end doors. The drawer is beautifully fitted out for the beam and pointer, the pans and the SG pan. Unlike all previous Robinsons, the drawer has no hooks as the balance is now too heavy for frequent moving. The three original brass feet have knurled heads in Robinson’s usual early 19th century style; the head diameters are 26 mm. [J. WATERS. FIG. 108. THE ROBINSON & BARROW 10½″ MINUS BEAM AND PANS, TO SHOW FINGER BOLTS HOLDING THE CENTRE PLANE, THE CRADLES AND THE ROBINSON TULIP MORE CLEARLY. Beam: brass rhomboid with single brace on each arm; 3 mm thick at centre, tapering to 1½ mm just before the slit, expanding to 3 mm to give sufficient metal to take the slit and the pushing screw. Length 264 mm K―K (10½″), bright steel knives: centre 25 mm, end 13 mm, i.e. 1″ and ½″. Poise adjustment is by a nut on a horizontal threaded rod set in the L light of the beam. Steel tapered pointer, fixed, 88 mm, threaded for a brass CG bob and locking ring. It is notable that the beam has no reinforcement of the Robinson slits, neither by shim insert nor by stiffening screws, which seems odd as it is of the largest size, intended to carry 2000 grains, 130 grams. For now this remains unexplained.
130 This beam has a feature never before seen by the author or reported by any other. Whilst the balance appears to behave correctly, meaning that the three knives are co- parallel, coplanar and equidistant, the knives here are visibly not at 90° to the beam! They must be parallel but all three are visibly slightly askew. When the beam is out of the instrument the effect is very visible; it really catches the eye and one does not have to look for it. J-hooks: agate plane 19 x 6 x 1 mm in brass holder; steel (?) hook. Stand: brass base 70 mm wide, 57 mm deep plus two lugs at rear for rear spirit level. Pillar 15 mm diameter, height 245 mm to centre plane. The Robinson Tulip holds an ivory chart 20 x 13 mm, engraved with only three divisions each side of the central index, as shown. All the Robinson balances had a silvered brass chart of a FIG.109.THE CHART similar format and engraving. The centre plane agate 26 x 9 mm in a brass holder is held on top of the brass Tulip by two large brass milled-head finger bolts. Pans: A further improvement is that the pans are now each suspended on three ten- link platinum chains instead of the troublesome silk cords. The two slightly dished 67 mm diameter pans, and the 68 mm SG pan, are made of solid platinum. There is a long note in the drawer giving a condensed and rather muddled history of the balance, including data on these platinum pans. It also states that the original owner was Professor Story-Maskelyne and it was used by him in his chemical researches, ‘probably at Oxford.’2 A complete transcription of the long note was made by Buchanan, and may be found at the end of this chapter.3 What is made clear is that the surprisingly heavy pans were proved to be of solid cast platinum by spectroscopy and by density determinations of 21.2. During conservation work by the author another note was found making the same statements but in a different order, hidden between the slate slab and the wooden base. It was replaced there upon re-assembly. Why Story-Maskelyne needed platinum pans is not known, but from a perusal of his extensive bibliography it is clear that he must have had a good chemical balance for his chemical, crystallographic and mineralogical researches across the years. It may be that these are not the original pans, but have been transferred from another balance which he used, a Robinson precursor. This hypothesis was triggered by the long note mentioned above, and is laid out in Chapter 21, together with a supporting brief biography of Story-Maskelyne. Arrestment: Descending arrestment frame, brass, with locating ‘Cavendish’ cradles. Each is individually adjustable for the beam and the J-hooks, exactly as reported previously in Chapter 5. The release lever is of bright steel of rectangular cross-section, passing right through the pillar into a ball-and-socket joint mounted on a special support at the rear; again see Fig 108. The lever now has an ivory tip, 20 x 30 mm, nicely sculpted to receive and hold a finger, seen clearly in Figs. 106-108. The pillar is actually two concentric cylinders with the outer one sliding up and down and carrying the arrestment frame; the inner one is suitably slotted to form the lock when the balance is on release. Somewhere inside is a strong spring. Press the lever down slowly against the spring to set the balance in operation, and then slide to the right to latch it. So far, this is perfectly straightforward, but great care must be taken when unlocking or the spring will take over giving a very abrupt arrestment, shaking the entire instrument. The Rothamsted BA104 above, without a working spring, was 2 Professor Mervyn Herbert Nevil Story-Maskelyne, FRS (1823-1911). But not at Oxford; see Chapter 21. 3 Buchanan (1982), 293.
131 so much easier to operate. Note the vertical slot for the release lever in the wood just under the fixed centre window of the lantern. As before, there is a notch to the right to receive the latched lever. The author proved the utility of the adjustments provided on the cradles during the conservation work. Upon reassembly, the beam still kicked badly as it had before conservation. Close examination showed that the gap between centre knife and centre plane was tapered. The centre plane was then reversed; it had at some earlier date been installed backwards according to its own hidden marks. The action was much improved but still kicked to some extent on release. The height of the L cradle was raised such that it now caught the beam a little more firmly, giving a perfect release free of kicking. The end knife-to-plane clearance was set at six thou.4 ‘Wollaston Stabilisers’: thin brass strips, short arm 95 mm, long arm 180 mm to handle tip. Pegs, brass, 2 mm diameter, 40 mm tall, poorly threaded into the arms. Weight set: disappointingly no longer present, but there was a drawer space for it. Repairs: this balance was recorded by Buchanan in 1977 as having ‘8929’ on the underside of the drawer.5 Recent photographs indeed show an ‘8929’ pencilled across one corner of the drawer, very much in the style of an Oertling Works Number (W/No.). If underlined it would definitely have been an Oertling number, but a few authentic ones have been seen without the characteristic underlining. But it cannot possibly be a Robinson & Barrow manufacturing mark, as by 1850 Barrow had ceased trading under that name. The author now has no doubt that it is an Oertling W/No. as a Repair Number and, as such, it dates to August or September 1882, i.e. the year following its ‘repurchase’ by the British Museum. This would seem to be a very sensible time to have it undergo a full works overhaul’ but not with Henry Barrow & Co., no longer extant, but by the current leader in the field, L. Oertling. All makers serviced other makers’ balances, and in Oertling’s case they treated a newly arrived ‘foreigner’ for repair as if it were one of their own, meaning that when it went into the actual workshops for repair, it was given the latest W/No. The W/Nos constituted, in the civil service-speak of the period, a ‘progressive numero.’ A full account of Oertling Works Numbers as Repair Numbers may be found in the reference.6 The date has been achieved by interpolation between W/Nos. accurately dated on some of the 1,200 surviving halfpenny postcards with instructions to Oertling for orders, repairs, catalogues and price lists, missing invoices, queries and complaints etc etc.7 Alas the actual postcard for this repair is not present in the surviving set. The Story-Maskelyne Note. “Balance No.4. Originally belonged to Prof. M.H.N. Story Maskelyne and used by him in his chemical researches (probably at Oxford). On his retirement from the Keepership it was purchased in 1881 through J. R. Gregory_ (Thos. Chas. Rob. 1792―1841 was the first maker of prec. bal). See letter from J.T. Stock 681/20 of 15.6.70. Photo in Mus Studio. This balance is intermediate in design between Barrow’s balance in the Science Museum and the later type recently found in Cambridge. S. Maskelyne was the last user of the chem. lab. in the Old Ashmolean at Oxford in 1857: he had residence there from 1851. See Gunther. Dr. Moss 3.5.72 Composition of pans and chains in Pt of s.g. 21.2 and 21.0 respectively confirmed by spec. plate 135.2 Sensitive 2mg/div. As displayed the RH pan is 14.5 mg. lighter than the LH pan which is within 0.5mg of the s.g. pan in the drawer.” 4 Meaning six thousandths of an inch. 5 Buchanan (1982), 293. 6 Oliver (2020), 40, 354, 600, 604, 606, 648-649. 7 Currently in the author’s collection.
132 A question now arises: what is the meaning of ‘This balance is intermediate in design between Barrow’s balance in the Science Museum and the later type recently found in Cambridge’? This balance is the Barrow in the Science Museum, their 1977-340, and one of the more advanced Robinson types known. What therefore is ‘the later type recently found in Cambridge’? Presumably this refers to the ‘ex-Stokes’ balance in Chapter 17. ‘Recently’ would seem to be in the early 1970s. The only ‘improvement’ present in the Cambridge ‘ex-Stokes’ balance and not in this is the addition of the pair of stiffening screws under the Robinson slits at the beam ends. It might quite reasonably be thought that the Story-Maskelyne note has been put in the drawer of this balance in error, but the fact that a similar note was found hidden between the slate and wood base suggests otherwise; we take the note to refer genuinely to this balance, so the uncertainties remain for now.
CHAPTER 19 The ‘J.B. Dancer’ Robinson, Query UPERFICIALLY, the instrument shown in Fig. 110 appears to be an accurate copy of a 10½″ Robinson & Barrow balance, and is signed ‘Dancer Manchester’. He was John Benjamin Dancer (1812―1887), a Manchester maker and retailer of scientific instruments, especially optical ones, and is rated as the father of microphotography. He was also noted as an inventor. He sold this instrument to the well-known physicist James Prescott Joule (1818―1889) in February 1846.1 It is now in the Joule Collection at the Science and Industry Museum, Manchester, accession number Y1969.5.78. This museum is now a part of the Science Museum Group. [P.D. BUCHANAN. FIG. 110. THE ‘DANCER’ ROBINSON. 1 Stock (1973), 979, quoting brief extracts from Joule’s letters. S
134 The ‘Dancer’ was once taken to London for side-by-side comparison with the Robinson & Barrow ‘ex-BM’ at the Science Museum, 1977-340, above.2 The designs were found to be identical, and also in almost all dimensions. Comment was made that the beam length was closer to 10¼″ than 10½″, but this difference is irrelevant because such a variation was quite normal at this date, as has been previously explained in Chapter 6. The release is again spring-loaded; the lever has no ivory tip and has severely damaged the front edge of the drawer base. The drawer has hooks as before to prevent disasters in transit, and the single ivory drawer pull has lost its button, as is so often the case with 19th century balances. The two spirit levels are both working. The L level again partially eclipses one of the fixing screws. The ‘Dancer Manchester’ inscription is nicely engraved in elegant italic lower case lettering but, most curiously, is off-centre and is filled white, not centred and filled in with the conventional black wax. Stock took a photograph of the ‘Dancer’ in 1973, and reported that the lantern was missing. Buchanan took a few photographs of it in 1980, and confirmed that the lantern was missing. Both researchers showed a beam exactly as that on the Robinson & Barrow ‘ex-Stokes’ Wh.1371 and the ‘ex-BM’ 1977-340, all three with no flagpost but with the horizontal poise adjustment in one of the end lights. Buchanan reported that this balance was then in poor condition and badly corroded yet was working, just.3 The only remaining half-decent photograph, above, shows the balance incorrectly set up and minus the beam and pointer, which one hopes are in the drawer. Each pan is seemingly balanced absurdly on the two pins of each ‘Wollaston Stabiliser’. As to the missing lantern, the Science Museum Group website now reports that the lantern was removed for photographic purposes in 1931, and was subsequently lost. Note that the museum’s date of c.1855 is wrong. No source has made any comment about the possibility of reinforcement of the Robinson slit. The real question is: did Dancer make this balance? A careful search has failed to find any evidence that Dancer ever made a balance of any sort. Further, it seems highly improbable that any competent instrument-maker such as Dancer would make a slavish copy of another man’s work, especially in an area with which he was unfamiliar. Note too that he not only made but also retailed scientific instruments.4 The probability, amounting almost to a certainty, is that this was made by Henry Barrow, but ‘blind’ so that Dancer could engrave it with his own name.5 The unusual mode of inscription noted above would seem to confirm this point. The weight set in the drawer is: 1000/400/300/200/100 grains, in the ‘Old’ grains system normally used by Robinson, and as used in the Robinson & Barrow ‘ex-Stokes Collection’ in Chapter 17. Many years later, c.1870, Dancer signed a microscope made not by him but almost certainly by Smith, Beck & Beck of London.6 This adds even more weight to the above, despite the fact that Dancer did make many microscopes himself. In summary, this is just another routine production Robinson & Barrow 10½″ sailing under another flag. 2 Stock (1973), 979-980. 3 Buchanan (1982), 431. 4 For an excellent account of Dancer and his many activities and achievements see Wetton (1991). 5 See Stock (1973) for more on this aspect, on J.B. Dancer and J.P. Joule. 6 Holbrook (1992), 58.
CHAPTER 20 The ‘Adie’ Robinson Copy NITIALLY, this instrument was thought to be a Robinson balance made ‘blind’ so that a retailer could have their own name engraved upon it, as in the case of the ‘Dancer’ in the previous chapter. It is now quite clearly a 10″ copy of a Robinson 10½″ and made by Adie & Son of Edinburgh at some unknown date. It is currently resident at the National Museum of Scotland, accession number T.1968.25; see Figs. 111 to 113 overleaf. It is signed Adie and Son / EDINBURGH on the chart. The company’s trade card is pasted inside the drawer with the address 50 Princes Street, from which a rather wide date range of 1843―1877 may be deduced. It was therefore built sometime after Robinson’s death in 1841. Alexander James Adie (1775―1858) had been trained in instrument-making by his maternal uncle John Miller, and became a partner in the business in 1804. On his uncle’s death in 1825 Adie continued the business in his own name, and took his son John into partnership in 1834. Adie & Son were the leading members of a group of about a dozen Edinburgh instrument-makers, and were favoured by the local Natural Philosophers and engineers of that time. The father and son were more than just instrument-makers, they were Natural Philosophers in their own right, publishing papers in the Edinburgh learned journals, and each became a Fellow of the Royal Society of Edinburgh, a unique accolade for instrument-makers. For more on the background of Adie and Son see reference.1 The first of the principal ways in which the Adie differs from a true Robinson balance is that although the pressing screws are present, they are laid out in a reverse manner yet having the same effect; see Fig. 113 and compare with Fig. 37 in Chapter 5. The pressing screws have been withdrawn in the photo. The Robinson slit and pushing screw is present, but note the damage to the head of the pushing screw. The knives are in excellent condition, still mirror-polished and with only mere traces of rust as can be seen in Fig. 113. The beam is only 10″ (255 mm) long, not 10½″. The centre knife works on an agate plane, and the suspension pieces are agate-in-brass J-hooks. The polished brass pans are 60 mm diameter on plaited red silk cords, and there is also a single SG pan on short cords. The circular pillar base carries crossed levels. The arrestment frame is very similar to Robinson’s, both in the way it curves behind the Tulip, and it carries Cavendish/Robinson-style adjustable cradles. The Tulip is more open than Robinson’s but still very recognisable. The silvered brass chart is graduated 5―0―5 in unit divisions, slightly wider than those of Robinson. The steel pointer has lost its gravity bob, and the flag is missing from the flagpost. 1 Stock and Bryden (1972). See also Bull. SIS 10 (1986), 5. I
136 [© NATIONAL MUSEUMS SCOTLAND. FIG. 111. THE ADIE. [© NATIONAL MUSEUMS SCOTLAND. FIG. 112. THE ADIE DRAWER CONTENTS AND TRADE CARD. The arrestment is on the same spring-loaded principle as before, but the Robinson- style lever ends in a simple pivot on a bracket well behind the pillar which is stouter than a Robinson pillar. The L-shaped slot for the lever is reinforced with a small brass plate, seen clearly in Fig. 111. The ‘Wollaston Apparatus’ is again present, but arranged slightly differently.
137 [© NATIONAL MUSEUMS SCOTLAND. FIG. 113. AN ADIE BEAM END. The cabinet is very different from those of Robinson, except for the lantern still being completely removable and hooked in position. It has not only a front doorslide but also a rear one. Curiously, the solid wood side panels slide upwards and out too; Adie did not learn this from Robinson, and it seems to be a unique feature. The front doorslide, and presumably the end slides too, were held aloft by a compressed spring strip, clearly visible on the R in Fig. 111. The front doorslide carries a most unusual feature, a magnifying lens set into the glass window; aligning with the chart. The non-counterbalanced front and end doorslides would have been difficult to move and set during actual weighing; the Robinson end doors were rather more practical. There is a Robinson-style fitted drawer, with the expected beam compartment, a pan pen, forceps and two sets of masses. The masses themselves follow Robinson’s values, being of the brass knobbed type in the old 4-3-2-1 Grains system, from 1000 gr down to 10 gr, plus an incomplete set of ‘fractional’ masses in platinum, 4, 3, 2, 1 grains plus 1/10, 1/100 and 1/1000 gr. Both sets live in open boxes within the drawer. There is more on these in the Weights chapter. In the opinion of this balance user, the Adie would have been most unpleasant to use with the lantern on. Somewhere in the date range 1804 to 1825, therefore long before the above balance was created, John Miller and Alexander Adie had already built a high performance and high capacity balance. It had a steel open rhomboid beam carrying a heavy steel knife working upon a steel plane, as reported in the Edinburgh Encyclopaedia of 1830.2 The image of it is taken from the Parker & Delaplaine American edition of 1832, Vol. III, 213-215 and plate LII, figs 1 and 2, as the 1830 plates are not available on-line. In the caption to the American image, see Fig. 114, a date range 1813―1819 is given, but does not make it clear as to what this refers; is it the invention of the balance or the publication date of that volume of the Encyclopaedia? Either way, it is obvious that this balance existed by 1819 at the very latest. 2 Brewster III (1830), 199-201 and Plate LII. In the first American edition, pages 199-201, but 213-215 in the second, the Parker & Delaplaine edition of 1832.
138 The length of the beam was not stated, but it was forged from a single piece of steel. The suspension points at the beam ends were very complex and are not entirely clear (fig.2). The balance was said to be sensitive to 1/100 grain with a load of 10,000 grains, i.e. 0.65 mg in 650 g. This is not only one part-per-million, but is so at a substantial load. It should be noted that although the centre bearing is relieved, the end bearings are only partly relieved, the components remaining in contact although the weight of the pans has been largely removed by a crude form of panstop. This large balance was made for James Jardine (1776―1858), a Scottish civil engineer and friend of Alexander Adie. It was used by the pair of them in 1825 for comparing the old Edinburgh standard masses against the new Imperial standards. FIG. 114. THE MILLER AND ADIE HIGH CAPACITY BALANCE WITH A STEEL RHOMBOID BEAM. These all seem to have been Miller and Adie’s own ideas on the construction of a high precision balance. Having seen a Robinson production balance, or more likely several, Adie later abandoned all these original thoughts and clearly followed Robinson’s lead instead. As he was familiar with Robinson’s work, using Robinson’s standard design as a starting point seems superficially quite understandable, especially as this was after Robinson’s death. But Henry Barrow was still making the very same designs, so why did Adie not simply order one ‘blind’ from Robinson & Barrow and engrave his own name on it, in much the same way as Dancer had done? This would have been much more economical than going to the expense of creating the necessary dies, jigs and tools, unless he envisaged making quantities of these. How many did he make? Currently, this seems to be the only one known.
CHAPTER 21 The Robinson Precursors, Query HE THREE ANONYMOUS balances at the head of the Robinson list in Chapter 6 now enter the fray. All have a very Kater―Robinson feel about them, sharing several features with the original Kater design. They have been tentatively attributed to John Corless the scale maker, 1 Equally, one might see them as more probably very early efforts by Robinson, pre-Kater, and offered anonymously. Corless has now become almost irrelevant in the author’s opinion, but cannot be eliminated entirely, not yet. Let’s begin with the simpler instrument, as seen in Figs. 115 and 116 overleaf. This is in the History of Science Museum, Oxford, HSM for short, previously known as the Museum of the History of Science, Oxford, MHSO for short. MHSO No.157. (Current HSM Inventory No.18579). The earliest reference to this instrument is in the 1823 listing of chemical apparatus belonging to Professor C.G.B. Daubeny (1795―1867), as reported by Hill.2 Daubeny was Aldrichian Professor of Chemistry 1822―1854, and received a £200 grant from the University to purchase new equipment in 1822. This balance fits well with an 1822 date as a possible anonymous very early Robinson. It could perhaps have belonged to Daubeny’s predecessor John Kidd (1775―1851), but this seems less likely. This balance has not been examined by the author, but was examined by both Hill and Buchanan.3 Hill’s in-house report of the balance is as follows: 157 Balance beam and pans missing in wood and glass case: 26 x 24.5 x 10 (cm) The beam arrestment device slides on a square brass pillar; the beam rested on steel bearings. This is probably one of the balances mentioned in Daubeny’s 1823 inventory as being made by Corless, and appeared to have platinum pans. The current HSM website adds the following useful data: Copper alloy pillar with two rounded grooves for holding a weighing beam. The Daubeny inventory makes two clear statements in two parallel columns thus: Two Balances by Corless with Platina Pans One balance by One Pair of Scales Corless (cost) 20..0..0 One Weighing Table One Common weighing Machine To which we add the anonymous handwritten label appearing in the balance case: 1 Crawforth (2010).o data on his other products. 2 Hill (1971), 1-4, 33, 74. Daubeny’s second surviving balance, HSM 18850 (MHSO 158) is probably irrelevant, but is noted. 3 Buchanan (1982), 334. T
140 Small balance of platina by Corless c.1820. [Beam melted down during Great War]. A drawing of the Beam by Children in Quart. Journ. Science 1821, in which it was stated that such balances were made by Robinson, Devonshire St Portland Place at £6. with a platina beam. This is erroneous, based on a misreading of the inventory. Those were platina pans. Plus the formal printed museum label: CHEMICAL BALANCE. BY CORLESS. PLATINUM BEAM SCRAPPED DURING THE WAR 1916. Clearly Hill is more cautious than the label writers, saying ‘This is probably one of the balances mentioned in Daubeny’s 1823 inventory as being made by Corless.’ [Author’s italics.] The pillar has an English style cut-out which is closer to circular, and small, only vaguely suggestive of Robinson, but can easily be imagined as his starting point. This fits well with its most probable date 1822 or 1823. There is no sign of a short chart within the cut-out. The arrestment is crude but interesting. The sloping arms, 87 mm long, are clearly to catch and lift the beam, and at the same time there are two little wires which are clearly intended to lift the beam by trunnions (stubs) on the ends of a centre knife. To be absolutely clear, only the two arms and two wires are moveable; the rest of the stand is stationary. The knife would appear to operate in and on the curved recesses at the top of the stand, very much a scale-maker’s self-centering approach. Hill helpfully adds the beam rested on steel bearings, from which we deduce the brass curves were lined with steel, and this is confirmed by Buchanan. There is no lever with which to elevate the arrestment, rather it was lifted by hand, holding the bronze fitment where the two arms and the two wires meet. This slides up and down the square cross-section pillar, held in place by a captive flat spring. The pillar is 5 x 5 mm, and 88 mm tall. The overall height of the stand is 150 mm. The unsigned baseplate, 156 x 153 mm, is held in the case by two large finger bolts. Beams were never made of platinum, not merely because of the cost, but simply because it is too dense, creating a heavy beam, whereas a beam should be as light as possible consistent with rigidity. We know that the platinum beam statement was nonsense and can be dismissed. As the handwritten label says, the ‘platinum beam’ was simply a careless reading of the platinum pans, and we also have Captain Kater’s correction.4 The absence of the beam is not too great a problem, as Buchanan has deduced it to be ~5″ (~13 cm) from its pen in the fitted drawer, and the pointer likewise to be 5-6 cm long. The five sections at the front, each lined with navy-blue velvet, are presumably pens for pans and weights. But what happened to the platinum pans? Was it they which were melted down in the Great War, WWI? Perhaps, but there is a more interesting possibility. First, we need to look briefly at the career of Professor Story-Maskelyne of Oxford University. 4 Kater (1822).
141 [© HISTORY OF SCIENCE MUSEUM, UNIVERSITY OF OXFORD. FIG. 115. MHSO 157 (HSM 18579) WITH DOORSLIDE IN PLACE. [© HISTORY OF SCIENCE MUSEUM, UNIVERSITY OF OXFORD. FIG. 116. MHSO 157 (HSM 18579) WITH DOORSLIDE REMOVED.
142 Mervyn Herbert Nevil Story-Maskelyne, FRS (1823―1911) was a grandson of Revd Dr Nevil Story-Maskelyne (1732―1811), one time Astronomer Royal.5 MHN Story-Maskelyne was educated at Wadham College, Oxford; BA 1845, MA 1849. He taught chemistry and mineralogy there from 1851 onwards, becoming Professor of Mineralogy 1856―1895. He was assigned a lab in the basement of the Ashmolean Museum in Oxford, for mineralogical analysis, but chemistry generally is also mentioned. For this work he would have needed a reasonably decent balance. Because the chair was so poorly endowed he accepted the post of Keeper of Minerals at the British Museum (BM) in 1857. For this he moved to London, but retained his Oxford professorship until 1895. At BM he performed a massive amount of curating and collecting of minerals, putting their previously disorganised collection on a sound footing. For this work he would have needed once again a good chemical balance. He resigned the Keepership in 1880, and in 1881 BM bought his 10½″ Robinson & Barrow balance, the one detailed in Chapter 18, which is fitted with solid platinum pans. It seems reasonable to assume that BM had originally bought this expensive and advanced balance for him in 1857 or shortly afterwards. But Henry Barrow engraved his products ‘Robinson & Barrow’ only from 1842 to 1850, so it must have been there already, but was assigned to Story-Maskelyne upon his arrival. If so, then why BM’s later repurchase of it from him, and was it fitted with platinum pans? We need to bear in mind that solid cast platinum pans were and are very expensive, consequently very rare. Now reconsider the balances available to him at Oxford. The 1823 Daubeny inventory makes several clear statements: Two Balances by Corless with Platina Pans One Weighing Table One balance by Corless (cost) 20..0..0 We have already seen that the assignment of these two balances to Corless is dubious, but the two balances had platinum pans, and one of them cost £20, a considerably larger sum than for a Robinson balance at that date, when he charged 8gns for his 5½″ model. The platinum pans would account for the huge price difference. At least one of these balances was considered to be so important as to require a dedicated weighing table. It seems a reasonable proposition that Story- Maskelyne was using No.157 with platinum pans for mineralogical and chemical analysis beneath the Ashmolean Museum from 1851 to 1857. The author suggests that upon his appointment at BM, he found that the Robinson & Barrow balance there had conventional brass pans, suspended on either silk cords or platinum chains. Having experienced at Oxford the beneficial resistance of platinum pans to chemical corrosion, he took the No.157 pans with him and had them fitted on platinum chains to the 10½″ Robinson & Barrow. It might be objected that someone at Oxford would have noticed the loss of these expensive pans in the lengthy period 1858 to 1916, but it seems that the Daubeny collection was very poorly managed and maintained, with many moves accompanied by loss and breakage.6 So the author proposes that BM had to buy the Robinson & Barrow back because Story-Maskelyne had brought the platinum pans with him and provided them free of charge. The accountants presumably recorded this as a ‘repurchase’ but in reality they were buying only the platinum pans. There does not appear to be any record of how much was paid by BM for the complete balance as we see it today. The later history of the Robinson & Barrow balance is covered in Chapter 18, its own chapter. 5 Based largely on his biography in DNB 2004; obituary in Mineralogical Magazine 16 (1911), 149-156. 6 Hill (1971); 3.
143 Unlike all Robinson scales and balances, the lantern of MHSO 157 is an integral part of the cabinet, which rather suggests a non-Robinson origin. The glazing is held in by putty, pigmented red, and held by pins underneath the putty; this is just a miniature version of domestic house glazing, and in an instrumental context suggests ‘early’. The brass ring on the top board is clearly for carrying the whole balance; did this suggest that Robinson originally conceived his ring or knob on the top board as a carrying device? In view of his detachable lantern being held onto the base by only two or three hooks, this seems dangerous to the author, and the ring/knob was instead intended only for removing and replacing the lantern. So the question still remains: is No.157 an anonymous early Robinson pre-Kater, or Corless or another scale-maker’s early or even first attempt at a chemical balance? Sci. Mus. 1927-1138 (ex GIII Collection item L14). In reality this was never a part of the George III collection, but was placed there in the 19th century ‘for convenience’, and is now a part of the Science Museum collection. This is a strange instrument, but does clearly relate to MHSO 157 above, and has a somewhat more ‘Robinson’ air about it; Fig. 117. Fortunately this one has retained its beam; see Figs. 118 and 119. The steel beam is of an open rhomboid form with steel swan necks; these are crudely soldered on and are non-adjustable. The distance between the suspension points is 6″ (148 mm). It has a centre knife in a solid metal holder which is adjustable laterally for equalising arm lengths instead of bending the swans. Also attached to this holder is an excellent flag on a flag post, and a finely threaded pointer carrying a small cylindrical gravity bob; see Fig. 119. [J. WATERS. FIG. 117. SCI. MUS. 1927-1138, GENERAL VIEW.
144 [J. WATERS. FIG. 118. SCI. MUS. 1927-1138, MINUS DOORSLIDE. [J. WATERS. FIG. 119. SCI. MUS. 1927-1138, BEAM AND POINTER.
145 The pillar has an opening which is clearly intermediate between that of MHSO 157 and the true Robinson Tulip. The arrestment is performed by hand as in MHSO 157, again held on a square cross-section shaft by a captive flat spring strip. The self- centering main bearing at the top of the pillar is very much as in MHSO 157, but without the two vertical wires to lift the centre knife via end stubs. Here only the two arms raise the beam, thereby also raising the centre knife. There is what appears to be a broken lever mechanism in front of the pillar and under the floor, which has proven impossible to resolve. The pillar rises from a rectangular brass plate which is held on to the case by the two large finger bolts. There are four other bolts which superficially would seem to be the level adjustment working in conjunction with the two excellent crossed levels, but in reality the four bolts do nothing. Clearly this instrument has been modified, probably more than once, and broken. The dark mahogany cabinet has an integrated, not detachable, lantern, and has lost all its glass. The front doorslide slides up and out, assisted by an ivory knob at the top which matches the ivory drawer knob. This style of ivory knob was common in this era, and matches exactly that drawn in Children’s sketch in Fig. 46 in Chapter 6. The case has four adjustable brass feet set unusually at 45°. The drawer is laid out very roughly in the reverse of normal with the beam at the front, but nothing seems to fit properly. There is a hole in the top board where a carrying handle was once fitted, rather as the handle in MHSO 157 and some of the Robinson survivors. The brass pans are only 35 mm (1½″) diameter; their small size suggests that this was intended primarily as an assay balance. The pans were suspended on plaited green silk cords, now broken. The eighteen ‘playing card’ weights have numbers corresponding to the grains in the normal pennyweights Troy. This scale looks very much as if a scale-maker is trying his hand at creating a fine balance, with some sophistication and some crudities, so just missing the target. The important point is that this and MHSO 157 have so many similarities that they are almost certainly by the same maker, whether Robinson pre-Kater, or Corless or otherwise. Could this be the second ‘Corless’ noted in the 1823 Daubeny inventory? Whipple Museum Wh.1446 was originally in the Wollaston collection (G248), but transferred later to the Whipple Museum, who have dated it to c.1820. It is anonymous and does not appear at first sight to be relevant to Robinson, being a very fine beamscale in the traditional style, all in bright steel and suspended from a tall brass stand in a glazed wooden case; see Figs. 120 and 121. The beam is a steel rod with simple swan-neck ends, overall length 262 mm, 10″. The central bearing is only a knife self-locating in suspended rings; see Fig. 122, best examined on the page under a lens. It has a fine, long, up-standing pointer visible in the highly decorative steel shears (gallows). The pans sit on the baseboard, and the beam is hoisted aloft for the weighing by a see-saw. But looking more closely at the lacquered brass stand, it is sleeved, very plain, 12 mm in diameter and very tall at 34 cm. The Robinson-style arrestment lever T is pivoted in a rough ball-and-socket joint just in front of the stand, making it a see-saw to raise the beam. Unfortunately the lever has been forced and has split the sleeve S2 to both L and R, as can be seen clearly in Fig. 123. It is all highly suggestive of a Robinson- style lever-lock although the lever does not now latch, but may have done so in the past. There is an earlier photograph, not shown here, in which the lever is clearly displaced to the R, suggesting that it really did once lock.
146 [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 120. WH.1446 WITHOUT DOORSLIDE. [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 121. WH.1446 WITH DOORSLIDE. S3 T Y1 Y2 X Shears W W
147 [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 122. THE CENTRE SELF-LOCATING KNIFE AND RING, AND PIN S3. The dark mahogany case is of a Robinson-style construction, with the glazed front doorslide made oddly short allowing the lever T to protrude by a few mm; solid end doors W with big hooks and with the leading vertical edge chamfered; a drawer held in place with two hooks Y1 and Y2, and the upper part as a removable lantern with a brass knob X, and held by one large brass hook each side. The doorslide rests just short of the lever T, leaving a circa one-inch gap, but with the top perfectly level with the rest of the lantern; compare Figs 121 and 124. The doorslide glass is puttied in, very much as MHSO 157 and unlike all known Robinsons, suggesting ‘early.’ The feet are missing. The one ivory drawer pull has snapped off. [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 123. THE LEVER IN A BALL JOINT IN FRONT OF THE STAND. Even more remarkable is the presence of the stand sleeve pin S3, preventing rotation of the sleeve as the arrestment is actuated, but as the working beamscale is suspended on a cord and free to rotate, why? Perhaps S3 was an original fixture for the spring? Observe that the beam has rotated horizontally in the photographs, as the R pan is nearer the front than the L one by over an inch. S3 T S2
148 [© WHIPPLE MUSEUM, UNIVERSITY OF CAMBRIDGE. FIG. 124. THE GAP BENEATH THE DOORSLIDE WHICH ALMOST TOUCHES THE LEVER. The drawer is empty and featureless except for two large pan pens, suitable for pans of about 70 mm diameter, not the current 50 mm dished brass pans suspended on dark green plaited silk cords. To summarise, this is clearly a marriage of an all-steel traditional swan-neck scale suspended from a tall lacquered brass Robinson-style stand, and all inside a case made by Robinson’s cabinet-maker. For our purposes the steel scale can now be dismissed as an intruder. Several currently unanswerable questions now arise: 1) Are the stand, arrestment and cabinet examples of Robinson’s work before he was commissioned by Kater to construct his economy hydrostatic balance design? 2) Alternatively, was this design by some unknown maker, the origin of the Robinson lever-lock and the Robinson case styling, or in other words, inspiring Robinson? 3) If not an original Robinson design, then by whom? We have already had the name Corless vaguely attached to the Robinson-style MHSO 157 at the start of this chapter. Now we have a Robinson-style item searching for a maker. It looks as if Corless, whose style is completely unknown, still cannot be dismissed entirely as a possibility. 4) What was the nature of the previous scale or balance in this case? The only known fact is that it had larger pans. 5) This short doorslide would seem to be the precursor of the true Robinson lantern doorslide with a slot for the release lever worked into a deep lower transom. The short doorslide is unique in the author’s experience. One would imagine it to be of a very transient existence before ‘the penny dropped’ as to how to achieve an accessible working release lever with a case fully closed against draughts. Undoubtedly more proto-Robinson or pseudo-Robinson scales will turn up in years to come, as well as more genuine Robinsons. It is to be hoped that such finds will resolve these questions. One known pseudo-Robinson scale has been relegated to Appendix 7, in order to avoid confusion with the presumed precursors presented in this chapter.
CHAPTER 22 The Styling of Robinson’s Balance Cabinets OMETHING needs to be added on Robinson’s balance cabinets and on the development of balance case design in general. Although Robinson’s were neither unique nor in a highly distinctive style of his own, they do appears to have had a considerable influence upon later balance-makers. But first, do we call them ‘cases’ or ‘cabinets’? ‘Case’ is shorter but ‘cabinet’ seems more correct, as by Robinson’s time the cases had already become of such high quality that no carpenters or joiners were capable of creating them, and they had become the preserve of true cabinet-makers and French-polishers. Robinson used the same cabinet-maker throughout, but his identity is unknown. He used a dark mahogany on all the preserved Robinson scales and balances, with a very smooth and silky — not shiny — finish which was common generally on philosophical instrument cases in those days. He was clearly very good at his craft. The ‘Swan-neck’ / ‘Early’ scales. Beginning with these small early products, their mahogany cabinets were of a simple design following one of the scale-making traditions, consisting of a drawer base upon which a glazed lantern was placed, accurately located by three small pegs on the lower edges mating with three small holes in the drawer base. The lantern was then held in place by two hooks, as shown in Chapter 7. In the original announcement of the Kater―Robinson economy hydrostatic scale, nothing is actually said about any cover, but in Children’s sketch the three holes for the lantern’s pegs are clearly delineated; see Fig. 46 in Chapter 6. The principal function of a lantern generally was to keep dust off the instrument, but also served to prevent draughts interfering as scales were evolving into balances and becoming more sensitive. Robinson made provision for using his scale with the lantern remaining in place, as was shown in Chapter 7. A brass handle or an ivory or brass knob was fitted to the solid top board. This was simply for lifting the lantern off the base, but could easily be misused as a carrying handle for his smaller scales and balances. Soon after Robinson’s time balances were so sensitive that dust- and draught-proofing were essential, and a lantern became an integral part of the cabinet, leading to a basically standard rigid design for makers for the next 100 years. The glazed detachable lantern on the ‘Steel Scale’ is a close match to those of many of the later balances, and may be considered representative. This lantern and most of the late others were glazed front and rear only. The solid ends usually contained neat hinged wooden doors, beautifully made with the leading vertical edge chamfered by two mm to create a close fit. It is greatly to the credit of Robinson’s cabinet-maker that both doors on the ‘Steel Scale’ are still a perfect fit 200 years later. S
150 Robinson’s drawers were fitted out for a beam, pointer, pans, a Specific Gravity pan, fractional weights, forceps and even a screwdriver. Unlike many scales of the period, the stand was not packed into the drawer for ease of transport, but was permanently fixed to the drawer base. To prevent the drawer from slipping out during transport, it was held in place by two brass hooks very much as the hooks holding the lantern in place. It is most noticeable that all the hooks observed to date in all surviving Robinsons are replacements, as there are tiny screw holes close to the existing hooks and pins, but clearly smaller and closer. Obviously those lightweight hooks and pins were found wanting, so larger and stronger ones were fitted instead. The two or more hooks holding the lantern in place were a tight fit to their pins, with L and R operating in opposite directions, all to minimise the chances of accidental separation. In contrast the hooks holding the end doors in place are set very loosely so that the user can lift them with a single finger whilst holding a weight or sample in forceps. The reader will find more highly detailed descriptions of the cabinets of the two early scales in Chapter 7. Drawers generally are considered again in Chapter 23, Weights. The routine production balances in three sizes. Moving now to Robinson’s best- known balances, here several types of cabinet are found. All have a basic clipped–on mahogany lantern glazed at front and rear only, as on the ‘Steel Scale’ above, with the ends and top in solid wood, unless otherwise stated. All have the cut-outs in the case front for the ‘Wollaston Apparatus’ handles and the release lever, unless otherwise stated. All have a fitted full-width drawer, with only detail differences in their internal layouts, as shown in the figures in the various preceding chapters. Taking all the instruments in their original order in Chapter 6, we have: MAKER SIZE NAME OBSERVATIONS Children’s balance sketch Lantern not shown, has not survived. TCR 5½″ The ‘Steel Scale’ Ivory lifting knob on top board. TCR 5½″ The ‘Wellcome’ Lantern has not survived. TCR 5½″ The ‘Wollaston’ Ends have solid wood doors; brass (?) knob on top board. TCR 5½″ The ‘Miller’ Ends have solid wood doors; hole only for knob on top board. TCR 5½″ ‘Hass 1’ Ends with solid wood doors; front door hinged R. TCR 5½″ ‘Hass 2’ No end doors, front door hinged R. TCR 10½″ ‘Used by Miller’ Ends with solid wood doors; front door hinged R. No ‘Wollaston’ apparatus or slots for the handles. TCR 10½″ ‘Cambridge. U.’ Damaged, incomplete, not seen. TCR 8″ ‘Rothamsted’ Ends with solid wood doors, front door hinged R. TCR 10½″ ‘ex Edinburgh U.’ Ends with solid wood doors, front door hinged R. TCR 18″ ‘Dr Moll’s’ Unknown; current cabinet not original. R&B 15″ ‘Miller 1843/5’ Unknown; has not survived. R&B 10½″ ‘ex Stokes’ Ends with solid wood doors; two asymmetric hinged front doors; cabinet fitted with additional internal latches. R&B 10½″ ‘ex Brit. Mus.’ Heavy slate base insert; two symmetric hinged front doors with fixed glazed panel between; two large glazed end doors; counterpoised rear doorslide with small lifting knob. HB 10½″ ‘Dancer’ Lantern now missing. Adie 10″ ‘Adie’ Front and rear doorslides. Vertically sliding end doors.
151 [WWW.JENEMANN.ORG. FIG. 125. THE ‘WOLLASTON’ AND ITS LANTERN. The ‘Wollaston’ as seen in Fig. 125 has the earliest surviving cabinet of a Robinson production balance. It is essentially identical to that of the ‘Steel Scale’ cabinet, but with the drawer hooks Y1, Y2 now in place. The ivory drawer pull Y3 is clear to see as is the L-shaped slot for the release lever. Some other points worth noting are: 1) The hole for the screw holding the ring or knob X. 2) The two end door hinges. 3) The end door hook and pin. 4) The pin on the lantern to engage with the hook V2 analogous to V1. 5) The slots for the ‘Wollaston apparatus’ levers are visible, just. [MATTHIAS HASS. FIG. 126. ‘HASS 1’ WITH ALL DOORS OPEN TO SHOW THEIR DETAILS. Y1 V1 4 2 3 1 2 5 Y2 Y3
152 Fig. 126 is an extra photograph of ‘Hass 1’, showing a different version of the Robinson cabinet and, in particular, the hinged front door and end doors. Despite the hinged full-width front door, the lantern is still clipped onto the drawer base. Note the way the front door closes neatly under the top board, and how the lower transom carries the L-shaped cut-out for the release lever, and the two very shallow cut-outs on the lower edge for the ‘Wollaston Apparatus’ handles. Note too the chamfers on the end doors ensuring a tighter fit when closed, and that the top board has no hole for a lifting handle or knob. This suggests that with a front door, it is not expected that the lantern will be removed often, so there is no provision for a knob. [© NATIONAL MUSEUM OF SCOTLAND. FIG. 127. A TYPICAL ROBINSON CASE FOR A 10½″ BALANCE. Many features of the Robinson cases mentioned so far are clearly visible in Fig. 127 of the ‘ex-Edinburgh University’ Robinson in this exceptionally well-lit picture. The end doors are visible from both inside and outside. In this case each end door has a hook and a knob; the latter is a rarity. The R pin to hold down the lantern has split the wood with rough handling. Note too the hooks and pins to hold the drawer in place, and the protruding release lever in its L-shaped slot. The ivory drawer pull has lost its button. Observe the correct use of the brass feet. The brass knob for the riderslide, a late addition, is visible high up on the R side, towards the rear. Doorslides. By far the most interesting aspect of cabinet design generally is the doorslide. The concept was ubiquitous, and it now tends to be forgotten that the very earliest scales and balances were housed in carpenter-made structures resembling small library bookcases with hinged doors. Some early precision balance-makers such as Breithaupt and Robinson followed suit in respect of the doors, but in polished mahogany cases made by true cabinet-makers. These outward opening doors, although necessary, were extremely inconvenient for the user.
153 So, who invented the doorslide? Who created a small sash window frame to replace a miniature bookcase door? Truly precise routine weighing would be well-nigh impossible without this simple invention. Superficially, this was a late 18th century innovation, and the earliest surviving doorslides known to the author, and by known balance-makers, are on the Whitehurst and Sisson balances noted in Chapter 4. Their dates are vague, from the 1770s or 1780s. The Troughton version of the Ramsden, Fig. 17 in Chapter 4, may also be of the late 18th century. One doorslide reported recently is in one of two balances seen on the windowsill of an 18th century chemical lab in an image taken from William Lewis’ ‘Philosophical Commerce of Arts…’ published in London in 1763, as shown in Fig. 128.1 This item also appears in Jenemann 1979.2 Beside a conventional chemical balance there stands a very large hydrostatic balance with a doorslide fully raised. The balance is mounted on a high stand in the case, with a liquid container under each pan. The balance cannot be later than 1763, and is probably a few years earlier, making this arguably the earliest known representation of a conventional doorslide. FIG. 128. AN 18TH CENTURY CHEMICAL LAB WITH TWO BALANCES ON THE WINDOWSILL. Now reconsider the couple of surviving balances noted above: the Whitehurst balance in the Boerhaave Museum has a suggested date of c.1765―1775, but could well be some years earlier, and the Sisson balance in the Science Museum is dated at pre-1789. One of these three is probably the earliest known conventional doorslide, with the Whitehurst and the Sisson as the earliest known surviving doorslides. As far as is known, the next doorslide made in England was by Troughton in 1812 to the requirements of the American scientific civil servant Ferdinand R. Hassler, who insisted not only on a glazed case for his custom-made balance, but with a glass front as a single piece sliding up from below! His thinking is worth quoting:3 1 Warner (2018), 34. 2 Jenemann (1979), 34. (Not in the 1997 bilingual version.) 3 Hassler (1832).
154 I rejected doors; they are in the way when the case is open and, in shutting, they occasion a motion of the air within the case, which I have avoided by sliding the frame gently up into its place from below. He is quite correct about the inconvenience of an open hinged bookcase door, and of the massive motion and exchange of air within the case on shutting the door. The glazed doorslide is the way forward, but why from below when the conventional vertical doorslide had already been invented? This perverse form preferred by Hassler seems very strange. One might ask ‘where would the user’s knees go?’ Obviously the user would need to stand at the balance mounted on a slightly higher bench or balance table than is currently conventional. Fortunately this inconvenient mode of action never caught on, but the basic concept of a doorslide did. There are grounds for believing that many early chemists stood at their benches for all operations, including weighing. Another criticism of Hassler’s demand is that the bench or balance table would have to be specially designed to have a slot or recess into which the doorslide could descend. Next is a cased portable balance by Pistor c.1815/1820. Certainly the ‘Berlin School’ of balance design, meaning Pistor and his ex-apprentices, all used doorslides, including August Oertling. As to the inconvenience of hinged doors, a Whitehurst balance in the Birmingham Assay Office is worst of all, with its door hinged at the top. It opened up and forward towards the user, and had to be swung right up and over to rest on the top of the cabinet if the user needed both hands for an operation. See Chapter 4, pp.22-23. John Field’s simple assay balance inhabits a cabinet with both front and rear door- slides; see Fig. 23 in Chapter 4. Its date is a little uncertain, but must be between 1818 and 1826, and probably between 1821 and 1826, exactly coincident with Kater and Robinson. It is potentially the first known balance cabinet with two doorslides. Kater and Robinson must have noted these English and European doorslides, but of the known Robinson cases with his ‘clip-on’ lanterns, none has a front doorslide. Only one by Barrow as ‘Robinson & Barrow’ has a doorslide, but that only at the rear. Thus it looks as if Ludwig Oertling was the first maker in England to specify the doorslide routinely, and that was in the 1850s, overlapping with Henry Barrow & Co. The survival of so few balances by Robinson and so very few early ones by Oertling makes further deduction difficult, but there is just one known relevant survivor, ‘The Giles’ of 1848.4 This is the only known surviving Oertling model 2* (‘two-star’), the precursor of the model 2. It resides in a mahogany case made generally in the Robinson style with a drawer base having a detachable lantern. This lantern has a solid wood top and ends, with glazed rear and a glazed front doorslide; it also has four locating pegs and four latches, V1 and V2; see Fig. 129. Curiously, it has no adjustable brass feet, but only four flat wooden ones. As this was a presentation balance it is unclear as to whether all model 2* had this case, or only this special one. We have no idea of the nature of the cases used to protect most of Oertling’s thirteen first catalogue models, but this surviving 2* raises the possibility of Oertling’s very earliest cases being made in the Robinson manner with detachable lanterns. By the time his catalogues are illustrated, in the 1880s, there is no sign of a detachable lantern, only conventional rigid and fully glazed cabinets with front doorslides, sometimes with rear doorslides too. There is still so much to learn. 4 Currently in the National Museum of Scotland, Edinburgh; inventory number T.1994.36.
155 [PETER DELEHAR. FIG. 129. OERTLING MODEL 2* IN A ROBINSON-STYLE CASE PLUS FRONT DOORSLIDE. The Kater―Robinson routine production balances were widely copied in the USA and Europe. One such American version is shown in Fig. 130.5 This copy by J.P. Duffey of Philadelphia does appear to reproduce the Cavendish cradles in full, but few if any others did, although some attempted a partial copy of the concept. It shows that Duffey preferred a rigid complete cabinet rather than a lantern clipped on to a drawer base; a very sensible step forward. Neither a front door nor a doorslide is shown, only the side doors, but it can be reasonably inferred that a doorslide has been omitted for clarity, as was often the case in the engravings in makers’ catalogues and in photographs. No indication is given as to how the forward projecting release lever was allowed for in the doorslide. The artist is probably not reliable, as there is no provision for latching; instead it looks like a see-saw. 5 Morfit (1849), 92-98. V2 V1 V1
156 According to Morfit, Duffey added a simple and clever spring driven panstop mechanism, not shown in this sketch. This was a highly original style of panstop, not taken up elsewhere as far as is known. The knob on the lower R side is of unknown function; it is not a part of this sprung panstop, but its location suggests it may belong to another type of panstop. FIG. 130. THE DUFFEY AMERICAN COPY OF A ROBINSON 10½″ BALANCE. To summarise: there was clearly little innovation in Robinson’s cabinets, only the hinged full front door, and the early abandonment of the lifting handle/knob on the lantern. It looks as if Robinson provided no choice of cabinets from which his customers could select. It fell to Henry Barrow, in the Robinson & Barrow period, to develop the Robinson cabinet with time, from a single inconvenient hinged door, to double doors meeting in the middle, to double doors meeting asymmetrically so as to avoid eclipsing the pointer and chart, and finally to double doors with a narrow fixed panel between. In this last style he is actually following the much earlier cabinet door designs of Ramsden (Fig. 8), Troughton (Fig. 17) and Robert Fidler. Barrow introduced numerous other improvements, notably glazed end doors and the slate insert, taking Robinson’s design so much further, yet perversely still retaining the detachable lantern even on the heavy instrument with the slate insert. It seems odd that he did not introduce a front doorslide yet did place a counterpoised one at the rear of the cabinet with the slate insert. The ‘Adie’ is the only Robinson-style balance in a cabinet with doorslides, not merely a single front one, but four in total: at the front, rear and both sides too. To date this seems to be a unique arrangement.
157 It is worth looking further back still, to the recently discovered Troughton balance in the Avery Museum, for which see Fig. 18 in Chapter 4, and Fig 131 below. This balance, seemingly late in Troughton’s balance offerings, is almost in the modern form, having a front doorslide, two front fitted drawers with a brass rotary release knob (an early ‘milled head’) between, and although now missing, adjustable feet. Where it harks back to earlier designs is the small doors in each end panel, and a full width rear drawer, fitted out to carry the beam and the two pans. The ivory drawer pull is of its time, with a black ring, slightly simpler than on the multi-ringed pulls of the cabinets of the ‘Wollaston’, Fig. 125 above, and Sci. Mus. 1927-1138 (Chapter 21). The L drawer is fitted for the brass masses but held, most unusually in an edge-on manner in slots. This is most impractical and not seen again. The one grain mass is missing, otherwise the set is satisfyingly complete and mostly in good condition. Small forceps are present, suitable only for the lightest masses or, more probably, for the ‘fractions’ which are missing. The big slot with the central depression is hard to understand, but may have been for some tool to manipulate the larger masses. Other empty features are not as yet understood; what else is missing? The R drawer presumably held a box of ‘fractions’, but there is space for much more. [© ANDREW P.B. LOUND. FIG. 131. AVERY 01.02.17 LEFT FRONT DRAWER. For each of his balances Robinson certainly took up the fitting-out of the drawers in the manner of Troughton, Field and others, for the beam, pans, tools, masses and boxes of ‘fractions’. It seems so strange that neither he nor Barrow ever took up the already-invented doorslide. Did they not realise its superiority over conventional doors? Did their clients not ask for it? Does that suggest that they did not discuss design with their customers, which feels unlikely? Robinson retained the end doors of his predecessors, essential if the lantern were to remain in place when weighing. Barrow greatly improved the front doors, and eventually glazed the ends of the cabinet to give better internal lighting, yet never, as far as is known, took up the front doorslide. Chemists around the world must be grateful that the next generation of balance-makers did.
158
CHAPTER 23 Robinson’s Weight Sets ERY FEW of Robinson’s commercial lab weight sets have survived, and he said little about them in his catalogues. All were naturally in grains at that time. As far as can be gathered from the limited number of survivors seen so far, the larger masses were all in brass, and those from 40 grains downwards were usually in platinum. To be more specific, the masses are in the now very rare ‘Old’ grains system 4-3-2-1. To date only one almost complete Robinson set has been found; and several part sets. Between them it has been possible to deduce which weight sets Robinson supplied for his three sizes of production balances. The principal weights were all single-piece knobbed brass masses, thus: BEAM LENGTH CAPACITY, GRAINS CAPACITY, GRAMS PRINCIPAL MASSES, GRAINS 5½″ 400 26 200/100 8″ 1000 65 400/300/200/100 10½″ 2000 130 1000/400/300/200/100 With the lower ‘fractional’ values totalling just over 100 grains (see below), these sets could supply weights up to the capacity of the balance. The conventional brass masses were housed in pens in the balance drawer: two for the 5½″ balances, four for the 8″ balances and five for the 10½″ balances. No pens were provided for the two principal masses on the early scales and on a few of the very early 5½″ balances. Fig. 133 of Rothamsted BA104, overleaf, shows a typical set of four for an 8″ balance. A ‘tyre lever’ weight lifter is provided to lift these four principal masses, and is more clearly shown in Fig. 45 in Chapter 5. The brass flat grains set of 6 to 1 grains in Fig. 133 are late 19th or early 20th century interlopers replacing lost platinum masses. Just above the twin forceps is a closed ‘fractions’ box of mahogany, shown open in Fig. 134. In this context ― grains masses ― ‘fractions’ begin below the lowest knobbed brass mass, usually the 100 grains, illogical as that sounds. This loose- topped box holds nearly a full set of geometric masses in platinum wire, one decade per recess. The intended layout is clear. Starting from the L, the first recess contains the four tens of grains, the next are the four units, then the tenths, then the hundredths, with the fifth recess for the thousandths although not one of these last is present. Not all masses are where they should be in Fig. 134. These have all been sorted, weighed and considered, and the findings are shown in table Fig. 132. V
160 [COURTESY OF ROTHAMSTED RESEARCH / LAWES AGRICULTURAL TRUST. FIG. 132. THE WEIGHT SET BELONGING TO BA104. This set of weighings has confirmed the relationship of geometry with mass, thus: square = 4, triangle = 3, V = 2 and a straight rod = 1, for each decade of mass. This was a very common idea amongst later makers but this is the earliest example seen so far; could this be another Kater/Robinson innovation? Clearly some values are missing and some have been duplicated, triplicated or more over the years. Those masses indicative of ‘2’ in their decade were intended to be V-shaped, but some were bent so slackly that they look rather more like the heraldic ‘bend’ or even L-shaped. Summing up, the ‘fractional’ masses as supplied to customers were as follows: 40/30/20/10/4/3/2/1/0.4/0.3/0.2/0.1/0.04/0.03/0.02/0.01, total 111.1 grains, plus a set of thousandths of a grain in some cases. All were usually in platinum, although one set has been seen where the knobbed brass pattern is carried down to 10 grains, to be shown below. In Fig. 134 the four masses in the tens decade were initially thought to be probably the only original wire weights, because of their matt surface texture and the poor finish on the ends. This assumption was incorrect, as will be shown later. They were obviously cut from a length of platinum wire, and as Robinson was working only from 1821 to 1841, these are of very early commercial platinum. It is known that Smithson Tennant and Wollaston, who made the first malleable platinum shortly before Robinson’s productive years, did so in a secret 26-step process. They had
161 eventually perfected a method by 1804/1805, such that by 1815 they had purified some 1½ tons of the crude metal, and gained a fortune thereby. Wollaston was able to draw the first fine wire in platinum by 1813; see his entry in Chapter 4. In his researches he used a 5½″ Robinson balance. What then could be more natural than for Robinson to latch onto Wollaston’s platinum wire for making corrosion-resistant small weights for use in chemistry labs? Robinson and Wollaston cooperated in other areas too; see for example the final page of Chapter 25. [COURTESY OF ROTHAMSTED RESEARCH / LAWES AGRICULTURAL TRUST. FIG.133. THE ROTHAMSTED BA104 DRAWER WEIGHTS. [COURTESY OF ROTHAMSTED RESEARCH / LAWES AGRICULTURAL TRUST. FIG. 134. THE BA104 ‘FRACTIONAL’ WEIGHTS BOX AND LID. Fig. 135 shows what now seems to be the standard Robinson ‘fractions’ box. This one belongs to one of the Hass balances and currently holds a few Continental metric masses from 500 mg down; ignore these. The under- side of the lid is shown; for the top side with the two brass hooks which engage the two brass pins in the base, see Fig. 133, or more clearly in Fig. 71 in Chapter 10. [IMAGE COURTESY OF MATTHIAS HASS. FIG. 135. THE USUAL FIVE-RECESS FRACTIONS BOX.
162 Some balances contain ‘fractions’ boxes with four recesses, such as the one in the Adie balance. In this case the knobbed brass masses continue down to ten grains, and the ‘fractions’ therefore begin at four grains. Adie followed Robinson closely in most respects, and it looks from this and other sightings as if Robinson used both layouts. Each of the four recesses in the Adie ‘fractions’ box is most helpfully labelled in the tiniest handwriting in white ink, thus from L to R: 1,2,3,4 1,2,3,4 1,2,3,4 1,2,3,4 Grains tenths of hundredths thousandths a Grain of a Grain of a Grain We now come to a most intriguing question: how pure was the platinum which Robinson used in his ‘fractional’ masses? The probability is that in the 1820s and 1830s he was using platinum prepared by Smithson Tennant and Wollaston, and extruded as wire by Wollaston, as described above. The Science Museum holds some Wollaston platinum wire in a set of his wires, notes and letters, inventory number 1946-228/1.1 The set has been dated by the museum as of 1800―1815, presumably on the basis of the letters and notes, which are not available on-line. No work appears to have been done on establishing the actual purity of this ‘pure’ platinum, or of determining which of the other platinum metals remain as impurities. Even if Robinson’s wire ‘fractions’ are not Wollaston’s they will still have been made of very early commercial platinum, and it would be most interesting to compare the purity and impurity patterns of the Wollaston and Robinson wires. The technique of choice for such analyses is XRF spectroscopy.2 Alas, neither Rothamsted Research nor the Science Museum has this facility currently available. However, XRF analysis was possible on the ‘fractions’ in the drawer of the 10½″ ‘ex- Edinburgh University’ Robinson balance; the results are tabled below. It must be understood that these are semi-quantitative figures, as no reference materials were available. The orders of magnitude are inescapable but the exact decimal values may be in some error. Possible traces of rhodium were not detected as this element is a part of the instrument, and osmium was not detected. Note n/d = not detected. MASS, GRAINS % PLATINUM % COPPER % RUTHENIUM % IRIDIUM square 40 99.2 0.3 0.1 0.5 triangle 30 99.3 0.2 n/d 0.4 bend 20 99.1 0.2 0.2 0.5 square 4 98.4 0.3 0.7 0.6 triangle 3 97.5 0.3 1.5 0.7 What is abundantly clear is that the three coarse wire masses are composed of a very high purity platinum >99%, and that the two smaller masses are still of a high purity, 97½% and 98½%. Thus the Smithson Tennant and Wollaston claim to have created very high purity platinum in industrial quantities appears justified. The copper content is presumably contamination from the brass barrel in which the platinum powder is compressed under extreme pressure in their process. 1 Besides the platinum wire are wires of palladium, gold, silver, copper, tin, lead, iron and zinc. The silver, copper and lead are said to be as alloys, unspecified. 2 XRF stands for X-ray Fluorescence spectroscopy, a very powerful non-destructive technique of elemental analysis, but requiring some very expensive instrumentation.
163 The question remains as to why the three higher wire masses have such a rough finish. It is now clear from the XRF results that it is not an indication of impurity, but may simply be an artefact of his extrusion process on what is a rather coarse wire. This would not have been by his published method for creating very fine wires of platinum, although the finer Robinson wire masses probably were.3 It is more likely to have been by a conventional extrusion process through a steel die. Wollaston gives no details of his preferred extrusion process for coarser wires toward the end of his paper on the creation of malleable high-purity platinum.4 Returning to the two Robinson scales in Chapter 7, the drawer of each contains a pen for tiny weights. The one in the ‘Steel Scale’ is empty, but that in the ‘Wellcome’ contains ten little rough grey metal balls, presumably lead shot. These were weighed on a portable electronic scale and found to be: 9, 9, 11, 11, 11, 11, 11, 11, 11, 12, grains, mean 10.7 grains; call it 11. Presumably these were intended to be roughly 10 grains, but for what purpose? Rough counterbalancing of small containers would seem to be the most obvious answer, in either normal or hydrostatic weighing. Their form ensures no confusion with the proper weights, and if they really were common lead shot, they were cheap and readily available. The observed forceps for handling the fractions are sometimes conventional, such as the two pairs seen in Fig. 133, all with modern curved points, but in the very early Robinson ‘Steel Scale’ and some others later balances, there are some very strange ‘old’ forceps. It has proven impossible to deduce what the original Robinson forceps were, and for now it seems best not to speculate. It is also unknown as to why Robinson usually supplied forceps in pairs, contrary to later conventional practice. The pre-Robinson balances by Troughton have thrown up a couple of oddments. First, in the Avery Museum balance 01.02.17, the principal masses are all brass discs set on edge in slots, from 2000 grains to 1 grain; see Fig. 131 in the previous chapter, and Fig. 136. [© ANDREW P.B. LOUND. FIG. 136. THE AVERY 01.02.17 WEIGHT SET. 3 Wollaston (1813). 4 Wollaston (1829).
164 This is a neat method of storage, but most awkward for retrieval when in use. The supplied forceps will handle only the lightest masses. Some type of tongs for the greater masses may have occupied the empty middle slot, but it is difficult to envisage a design that would not risk damaging the surface of the discs. The ‘obvious’ and easy answer is to use fingers, but as the masses are not lacquered, this would soon cause corrosion and errors in weight. To date this mode of storing masses has not been seen elsewhere. It is clear that knobbed brass masses, such as those by Robinson, are far superior and easy to handle with forceps, and by lifters for the larger masses as shown in Fig. 45, even double ended lifters for big, heavy sets with a different size of slot at each end of the lifter. It is unfortunate that the ‘fractions’ and their forceps are missing, although it may be that the forceps shown here were the ones for the ‘fractions’. Secondly, the Troughton balance T.1996.48 in the National Museum of Scotland has a capacity of some 16,700 grains, or just over 1 kg. A most unusual grains weight set was recommended for it, in an 8-4-2-1 sequence not seen before, thus:5 8000 / 800 / 80 / 8 / 0.8/ 0.08 / 0.008 4000 / 400 / 40 / 4 / 0.4 / 0.04 / 0.004 2000 / 200 / 20 / 2 / 0.2 / 0.02 / 0.002 1000 / 100 / 10 / 1 / 0.1 / 0.01 / 0.001 As with all makers, Robinson was happy to provide weight sets for use with other makers’ balances. For instance, there is a note in the Royal Society Council minutes for 11/11/1830 that T.C. Robinson was paid £2-12-6 for a set of weights for use with the Ramsden balance. Where is this set now? Further, in the same entry, Troughton & Simms were paid five guineas for repairing the Ramsden balance. As early as 1848 some leading chemists were advocating that students be supplied with metric weight sets, in terms such as this:6 As the French gramme system is universally employed on the continent, the student should be provided with a box of these weights which are fitted up by Oertling in precisely the same manner as the grain system. Despite these pleas, weight sets in grains continued to dominate weight sales long after Robinson’s time and Barrow’s time too, almost entirely in the ‘New’ grains 6-3- 2-1 system. This pre-eminence of grains sets slowly declined over time until the metric system became dominant in British scientific circles circa 1890. For more on grains and metric sets and the order of the masses therein, see reference.7 It should not be forgotten that, in addition to his weight sets, Robinson was capable of making extremely accurate standard masses, such as a copy of the Imperial Standard Troy pound which he made at the request of Schumacher through the good offices of Captain Kater.8 5 Brewster III (1830), 200. 6 Noad (1848), 122. 7 Oliver (2020), 864-866. 8 Schumacher (1836), 457.
CHAPTER 24 Robinson Revived in the 20th Century OST UNEXPECTEDLY, the author has found that the Robinson slit-end beam was resurrected in the mid-20th century by the instrument-making company William A. Webb. Initially at 223 St John Street, Clerkenwell, from 1894, Webb moved in 1919 to larger premises nearby at 4/5 Skinner Street. In 1946 they moved out of London to Ilford. They were taken over by Avery and merged with Oertling/DGS in July 1961. From that point on the Webb factory at Ilford became an extension of the Oertling operation at Cray, but to the outside world it continued to be William A. Webb Ltd for some years. It is necessary to delve briefly into the history of the Webb ‘White Monarch’ pharmacy scale before considering the derivative scales with slit-end beams, and their Oertling descendants.1 Those readers uninterested in this late development may safely skip to the next chapter, confident that they have missed nothing new in the 19th century story. The White Monarch was by far the best scale design by Webb’s somewhat eccentric designer A.W. Beal; see Figs. 137 to 140 overleaf. It was created somewhere between 1933 and 1937 as a superior dispensing scale for pharmacies. It was a genuinely stylish, original and highly practical design with much thought given to making the scale resistant to the effects of being out on a dispensing counter all day, and exposed to spilt powders. The stand was a wide and tapering construction of rectangular cross section in ivory bakelite. The seven-inch Dural beam, steel knives, agate planes and roof-type stainless steel suspension pieces, were all completely hidden within a matching ivory bakelite beam cover alias ‘shroud’ or ‘umbrella’, which was a simple snap-on fit to the top of the stand. The pointer descended inside the stand, appearing behind a glass window at the base with an easily readable chart. Thus all the operating parts were well screened from dust, both general and pharmaceutical. The rated capacity was 2oz, and sensitivity 0.1 grain; in metric terms 57 g and 6.5 mg. The deep base, in a hammer finish of ‘Opaline Green,’ contained a quarter-circle drawer on the left side. Usually the weights pan was stainless steel, on the left, and the dispensing pan on the right was glass and easily removable, all following normal pharmacy scale conventions. The top of the base was overlaid with a plate of polished stainless steel.2 Thus almost every exposed part was corrosion-resistant and easy to wipe clean.3 The excellence of the design was soon realised, and Webb 1 For the full history of Webb, the White Monarch and its Oertling derivatives see Oliver (2020), Chapters 32 and 35. The term White Monarch was a registered Webb trademark well before WWII, but the exact date is unknown. 2 Firth Brown ‘Staybrite’ non-magnetic stainless steel. (74% iron, 18% nickel, 8% chromium.) 3 The form described is that of the late Skinner Street and the Ilford periods. The earliest form had the stand and the beam cover in white porcelain with stylish curves, and the drawer base had two conventional drawers. It appears in lab supply house catalogues of the period, e.g. Townson & Mercer 1937. No surviving example is known to the author. M
166 sold as many as they could make, which was not very many due to their inefficient and mainly manual production methods. FIG. 137. THE WEBB WHITE MONARCH. [MIKE SHARPE. FIG. 138. THE WHITE MONARCH BEAM, FRONT, WITH POINTER AND FLAG IN POSITION. [MIKE SHARPE. FIG. 139. THE WHITE MONARCH BEAM, REAR, WITH SUSPENSION PIECES IN POSITION.
167 FIG. 140. THE FORMAL SPECIFICATION FOR THE WHITE MONARCH. 3FIG. 141. THE TWO AGATE CENTRE BEARINGS. 5FIG. 142. A FEW WHITE MONARCH PARTS. [BOTH: MIKE SHARPE. Some points of the specification need expansion. The beam is uniformly 3 mm thick, and the length k―k is 180 mm, 7.1″. The top edge of the beam carries a crude form of flag as shown in Fig. 138. The steel centre knife is 15 mm long. The centre bearing consists of two 5 mm agate cubes each machined to a curved self-centering working surface, as seen in Fig. 141. This is a scale-maker’s traditional style and is a source of error, not being reliably self-centering. Thus the instrument is not sufficiently reproducible for high-precision weighing, but is quite adequate for pharmacy dispensing and for rough weighing in a lab. The steel cheek plates are also shown. It needs to be stated clearly that the quality of the finish of many of the parts is distinctly rough, as can be seen in Figs. 141 and 142. The suspension piece on the right is particularly bad. The selection of Duralumin alias Dural for the beam was a good one, with the density of Dural being only 2.8, compared with 8.4 for the more conventional brass. Dural has the advantage of giving a lighter beam, therefore a greater sensitivity as indicated by beam theory, but it is quite strong enough for simple pharmacy weighing. Note too that Dural is only slightly worse than brass in terms of response to temperature changes, and it is also, like brass, non-magnetic. All three hardened steel knives were swaged into the alloy beam. The little gouge just to the R of the centre knife, as
168 seen in Fig. 139, looks very clearly as if the knife has been swaged in with a small chisel. This crude technique was not unknown on aluminium beams but uncommon on fine scales and balances.4 With all three knives seemingly locked solidly in position, the arm length is not readily adjustable once factory set, although it is possible.5 As it is a cheap scale, the knives would not have been drifted in, as on a finer balance. This is a rising plane scale but the three bearings are not fully relieved, they merely have the greater proportion of weight removed. Nevertheless Webb always promoted this model, as did Oertling with its descendants, as a balance even though it is technically a scale. The suspension pieces are of the roof type, consisting of an agate inverted V held in a stainless steel body, as seen in Figs. 139 and 142. The company later conceived the idea of a superior version intended for the general laboratory market. This they continued to name as a White Monarch, but it was additionally described as an ‘intermediate laboratory balance, chain model.’ Major changes were made, of which the most important was covered by the description of the beam in the new publicity sheets, where only one word was added: instead of ‘beam 7″ Duralumin’, it now read ‘beam 7″ cut Duralumin’, meaning that the Robinson slit had been added, as shown in Fig. 144. Although the adjustable slit has been added, the fine adjustments of parallel by pressing screws have not; refer to Fig. 31 in Chapter 5. The pushing screw enters from the underside as can just be seen in Fig. 144, and is adjusted with nothing more than a tiny screwdriver; this is much clearer in later versions shown in Figs 146 and 148. This was obviously held to be an adequate adjustment of co-planarity and arm length in this simple scale, somewhere between the fixed Webb beam and the precision adjustability of the original Robinson slit. As to who introduced this change at Webb, the ‘obvious’ answer is Beal, but surviving staff of the period suggested it was their managing director and the salesman. The much more interesting question, to which there is now no prospect of an answer, is where and how did they discover the Robinson slit, with Robinsons being so rare? [MIKE SHARPE. FIG. 143. UPPER: THE BEAM OF AN ORIGINAL WEBB WHITE MONARCH. FIG. 144. LOWER: THERE IS NO USABLE IMAGE OF A WEBB SLIT-END BEAM, SO A J10 IS SHOWN INSTEAD; IGNORE THE DIFFERENT MOUNTING OF THE CENTRE KNIFE; OTHERWISE IT IS AS ITS WEBB PREDECESSOR. 4 Swaged: where brute force has been used to create a distortion of metal to trap one part inside another. Knives drifted in means they have been forced into their triangular slots. First a small hole was drilled in the beam. It was then filed out to a triangular profile of the correct size, but with one end marginally narrower than needed. The prepared hardened steel knife, with a matching slight taper, was then hammered in or, better, pressed in, from the wide end until it wedged solid. This explains why some knives stick out a little more on one side than the other; this defect is harmless. 5 Oliver (2020), Appendix 11, section Knocking-Up; pp.859-860.
169 ‘Fluid damping’ was also added, hidden within the pillar. An oil damper dash pot was built into the push rod, and a vane dipping into the oil was riveted to the pointer; as shown in Fig. 147. Oil-damped scales had a small hole in the rear of the bakelite stand through which the dashpot could be topped up with oil; the hole was stopped up with a small black rubber plug. The stand was fitted with a chaindial attachment, to eliminate the handling of fractional weights. The type used here is more correctly known as a Grainger drum. The drum paid out chain in nine 0.1g steps, and the chart covered 100 mg in 10 mg divisions. Standard laboratory conventions were now used; both pans were fixed, and the pivoted ‘drawer’ was moved to below the weights pan on the right. This ‘drawer’ comprised a hinged wooden block quadrant with recesses containing a set of knobbed brass weights from 100 g to 1 g, complete with forceps. The specification was both revised and re-expressed in metric terms: 200 g capacity and 5 mg sensitivity. Several options were available: plastic or stainless steel pans, and L or R ‘drawer.’ As there was no change in the thickness of the beam, or the metal, it appears that the original plain beam was devised to be as simple and cheap to make as possible, being totally over-specified for a mere 2 oz or 50 g, but perfectly capable of carrying 200 g with only a very slight deformation and loss of sensitivity. A second model with oil damping but without a Grainger drum was also offered, entitled as an ‘intermediate laboratory balance.’ This made it essentially an updated White Monarch, but again following laboratory conventions with a quadrant ‘drawer’ holding a weight set on the right. Lacking the drum, the weight set carried fractions down to 100 mg. The two revised designs went into production only a year or so before the Oertling takeover. There is no doubt that these were good little second-class lab scales for rough weighing. The Oertling take-over. The first and most important effect of the Oertling/DGS take-over in 1961 was the immediate abandonment of all the Webb balance designs, by then considered to be uncompetitive in design and price. Most scales were abandoned too. The main exception was the White Monarch. It was this that Oertling wanted to ‘get their hands on’ and was the real reason for the purchase of WAW Ltd. They saw the new Grainger drum version not only as a first class dispensing balance, an area of the market which they had not recently occupied, but they also saw its potential as a cheap schools balance, another area in which they had recently been weak. Additionally, it had potential as a very cheap small rough lab balance alongside three of their own models with which they would now cover almost all possibilities at the lower end of the lab weighing market. All this meant that it was worth spending some time and money in optimising its production. A survey of the Webb operation before purchase had shown up some obvious inefficiencies. A senior Oertling engineer, Mike Hollott, was sent to Ilford to assess and revise the production of the White Monarch. He soon found the primary problem to be that the parts were inadequately specified and, before they could be fitted together, they frequently needed ‘a touch of the file’. This was frustrating and time-consuming for the two skilled operatives on assembly. Next, there was no Inspection function; it was ‘up to the final adjuster to give it the OK’ he was told. In reality the one girl on packing duty frequently had to point out obvious defects such as scratches.6 By re-specifying the parts, now made with new tools and with formal inspection specifications, equipment and procedures, all parts became fully interchangeable and 6 This seems to have applied to all the other scales too. The situation with respect to balances is unknown.
170 could be fitted with no extra work on them. Some thirty revisions were made in the re-engineering programme for more efficient production. These, with improved bench layouts and seating, better lighting and fresh decoration, improved the production rate five-fold, from ten to fifty per week, some 200 per month. High- volume sales began from 1962. Another result of the re-engineering was to increase the capacity specification to 250g. The Oertling J10, J11, OJ10 and OJ11. The revised version of the Webb Grainger drum model was renamed the Oertling J10, and issued in the form shown in Fig. 145. There was also a simpler version, the J11, taking the place of the White Monarch traditional dispensing balance, having a glass pan on the R and a hinged quadrant ‘drawer’ for weights on the L. These were the ultimate instruments to use the Robinson slit, albeit in a restricted form, up to as late as 1981. The name White Monarch was supposedly dropped, but confusingly often appears on some price lists and publicity flyers for both J10 and J11. To add to the confusion, the J11 was also offered in a stampable version by 1974 (and probably earlier), and given the original 2 oz/0.1 grains specification. The original hydrocarbon oil was replaced by a silicone oil, but at some point c.1972/1973 oil-damping was abandoned and replaced by magnetic damping. The serial numbers thereafter had the suffix ‘M’ attached, with 7418M being the earliest observed to date. It is easy to spot a magnetically damped scale because the hole and rubber plug are no longer present at the rear. FIG. 145. THE FINAL PRODUCT (1): THE OERTLING J10 LABORATORY SCALE. (#14231) (The pointer rarely shows well in photographs, as in this case, but it is just visible in the reflection.)
171 With the closure of Ilford in mid-1970, J10/J11 production transferred briefly to Cray, still at 200 units per month, before moving again in 1972 to the recently acquired Milngavie factory of John Jamieson (Glasgow) Ltd. Here the J10/J11 were renamed again as the OJ10/OJ11, and created in further large quantities. [BOTH: MIKE SHARPE. FIG. 146. OJ11 beam and pointer, push rod and frame. FIG. 147. OJ10 OIL DAMPER POT, POINTER AND VANE. The beams for the OJ10 and OJ11 are identical except for the engraving on the OJ11. Originally the slit adjustment was by a tiny screwdriver working on a tiny pushing screw recessed into the underside of the beam end, as shown in Fig. 148. The later beams have the recessed screw replaced by a prominent hex-headed screw as shown on the OJ11 beam in Figs. 146 and 150. This will make the adjustment far easier, but the author suspects that Webb’s designer Beal hid the tiny pushing screw to make it difficult for non-engineers to meddle with. Figs. 149 and 150 overleaf are of a recently found and almost perfect surviving OJ11. The push rod and the frame carrying the centre bearings are unchanged, as are the pointer and the crude flag. Most curiously, this OJ11 is undamped, which makes it a late FIG. 148. A J10 ROBINSON SLIT. re-creation of the original White Monarch,
172 but far better-made. As before, for pharmacy practice, the glass pan is on the R and the metal weights pan is on the L above the hinged quadrant ‘drawer’. Note the late sans-serif lower-case Oertling nameplate, contrasting with the previously used serif upper-case nameplate, from which a very approximate dating is now possible. Their observed dates so far are: 1956―1973 and 1977―1979.7 Regrettably there are not enough dates to close the gap, but the new OJ11 must date between 1972 and 1981, the start of the OJ nomenclature and the closure of Milngavie, respectively. The most probable date for this OJ11 is therefore in the range 1977―1981. FIG. 149. THE FINAL PRODUCT (2): THE OERTLING OJ11 PHARMACY DISPENSING SCALE. (#15854) FIG. 150. THE OJ11 BEAM. 7 From Oertling researches reported in Oliver (2020), 686-687.
173 Fig. 150 has been taken from the rear, and shows a late practical improvement for fitting the beam cover or ‘shroud’. Instead of the original snap-on action over two pips, which shakes the scale badly, the rear pip has been replaced by an Allen screw and washer at the top of the pillar. The stem of the screw matches a new slot on the beam cover which now slips over the screw; with the cover in place the screw is then tightened up slightly. Some useful sources of information have appeared in Australia. First, there is a certificate of approval from the National Standards Commission of Australia, dated December 1973: This is to certify that the pattern and variants of the White Monarch Dispensing Scale (also known as the Webb Class B Beam-scale) have been approved under the Weights and Measures (Patterns of Instruments) Regulations as being suitable for use for trade when converted to read in metric units… All instruments conforming to the pattern or variants shall be marked NSC No 6/2/101. A similar certificate of approval issued by the National Standards Commission has been noted for the OJ11 in these terms where ‘scale interval’ means chart division: This is to certify that an approval has been granted by the Commission that the pattern and variants of the Oertling Weighing Instrument Model OJ11 submitted by… are suitable for use for trade. All instruments purporting to comply with this approval shall be marked NSC No. 6/1/2. Pattern. A semi-self-indicating equal-arm beam weighing instrument of 50g capacity by 0.01g scale intervals. Variants. 1. 50g / 0.02g scale interval 2. 50g / 0.05g scale interval This is the only known record of any White Monarch or its descendants being available in different sensitivities. This approval is dated 1981 and states that it replaces a previous approval of 1974, No 6/2/101. In other words this approval of the OJ11 replaces that of the White Monarch above. The individual Australian states issued their own supplementary approvals. One very singular and relevant survival was seen briefly on eBay, a stamped White Monarch from Queensland, shown in part in Fig. 151. FIG. 151. PART VIEW OF THE QUEENSLAND WEBB WHITE MONARCH, #14342.
174 The beam was engraved: ‘CLASS B TO WEIGH 50 g’ and bore an indecipherable stamped lead plug. The engraved rating obviously fits the above certificates of approval, and according to a paper label the Queensland government has clearly approved it as well. It had the empty hinged weights ‘drawer’ on the L, and a glass dispensing pan; it had no plug on the rear of the stand, so it was presumably magnetically damped. It also carried a Webb White Monarch label as seen in Fig. 151 and more clearly in Fig. 152, but the winged ‘WAW’ was missing from the front of the hinged ‘drawer’. Most curiously, it also bore an Oertling nameplate under the release knob (not the normal position), and a serial number 14342. There is no Oertling serial number plate on the rear. Fig. 152. THE MYSTERIOUS SERIAL NUMBER. What is this item? Uniquely it carries both Oertling and Webb insignia. Is it an Oertling J11 or a Webb? From a distance it looks like a standard Webb White Monarch. But with the Oertling 1956―1973 door label on the front it looks like a J11. One would hope that the serial number would enable a distinction to be made, but it is of no help at all. It is not inconsistent with the Webb serials seen on their fine balances, and it also fits into the Oertling J10/J11/OJ10/OJ11 serials, at the most recent end. Several interpretations are possible. The most obvious is that it was a Webb White Monarch made at Ilford just as Oertling took over in 1961, but the serial number is far too high for an Oertling J- serial in 1961. Therefore #14342 is a Webb s/n, but there is no Webb serial number shown on the roundel in Fig. 137, nor on the roundels in any Webb publicity photographs. Alternatively it could have been an elderly White Monarch in need of a factory overhaul, where it may well have received an Oertling 1956―1973 nameplate and had the winged ‘WAW’ removed; an Oertling serial number was then engraved on the old Webb roundel. But it is highly unlikely that such a simple scale would have been returned to London. Instead it would have been serviced by Avery Australia, where it is most improbable that an Oertling nameplate and serial would have been added. There is no point in further speculation, and the dating of all White Monarchs and their Oertling descendants must be left to providential finds of future researchers.
175 Protective Cases. Webb had offered a lockable glazed wooden case for transport and/or extra protection in the workplace. Here they made a serious error in this user’s view, in building the case out of thin plywood and down to the cheapest price possible. It was finished in a clear spray lacquer, looking every inch ‘home-made’, and not at all complementary to the neat little scale within. Nevertheless, many of these cases were sold. It was intended that the plywood sloping front door would just be lifted out at the start of the working day in the pharmacy or the lab, and be replaced at night. Used in this manner the box also acted as a draught shield. It was a good idea but poorly executed. Oertling continued with this plywood case, as shown in Fig 153. There the carrying handle is non-standard, presumably added as a customer request. Later cases had no side or rear windows. FIG. 153. AN OERTLING J10 IN ITS PLYWOOD CASE. (The pointer is visible in this photograph.) Production Quantities. Huge numbers of J and OJ scales were made on the three Oertling sites over the years, and they turn up on the market very frequently, but no J11s have been seen. Only one surviving OJ10 and two OJ11s have been recorded to date, compared with 30 J10s, but large numbers have passed silently through on-line auction sites. In the 1970s and the 1980s pharmacies were rarely making up prescriptions, so few pharmacy scales were needed, which may account for the low numbers of surviving OJ models. A retired Weights & Measures inspector tells of pharmacy scales languishing in pharmacies in the 1980s, very little used, but having to be formally inspected regularly. It seems that the frequency of inspections varied from one W&M Authority to another, anything from twelve months to five years. Old hands say that large numbers of OJs were made and exported, where there were still markets for cheap, simple scales, mainly in Africa and Australia.
176 Other factors were at work too. First, the industrial need for second grade lab weighing was increasing greatly in Britain, but these old designs were facing very great competition from top-pan balances, both mechanical and electronic. This competition came from both Avery and foreign makers. Secondly, a situation had existed for many years in that Avery subsidiaries, secretly held but promoted as independent to the outside world, were competing with each other as well as with the true competition. This wasteful practice had persisted for decades but had been solved by c.1955.8 Because of the Avery take-over of Webb and its insertion into Oertling/DGS, the old problem had been accidentally revived; here were Avery and Oertling competing head to head once again. This is almost certainly the real reason why in 1981 the OJ designs were abandoned by Avery and Milngavie was closed. Some final points. On performance, recent trials on a surviving J10 and an OJ11 indicated that there were significant reductions in sensitivity with loading, especially up to 250 g on the J10. This was not easily identified towards the lower end of loadings due to the disappointingly poor reproducibility in both cases, but it must be remembered that both instruments are rather elderly and their treatment in the past is totally unknown. In summary, observed Oertling J and OJ serial numbers range from 745 to 15854, indicating at least some 15,000 units were made. The number of Webb White Monarchs made is unknown and likely to remain so. One wonders what Kater and Robinson would have thought in 1821 if they were told that their novel concept of a slit-ended beam would still be in use 160 years later, and that the Oertling descendants of their balances were made in far greater numbers than either Robinson or Kater could ever have imagined. 8 Oliver (2020), 55-56, 503.
CHAPTER 25 Robinson’s Balance Designs, a Final Assessment HOMAS CHARLES ROBINSON is a most interesting balance-maker, standing as a real pivot point between 18th century scales and attempts at high-precision scales on the one hand, and the 19th century true high-precision balances on the other hand. He was far ahead of his time with his short lattice beam, descending arrestment frame and very high precision arrestment and, it may be argued, even his cabinet design. He influenced Bate, Ludwig and August Oertling, and later makers too here in Britain, in Europe and in the USA. It is clear that Robinson was not only a balance-maker, but more generally an instrument-maker of great ability, seemingly discovered and promoted initially by Captain Kater, and thereafter ably guided and advised by Kater. A final assessment of T.C. Robinson really needs to be divided into three parts: the influences upon him, the development of his designs with time, and his influence upon later makers. Let us take these in turn. The influences upon him. Robinson’s education and training in instrument-making are, after much research, regrettably still totally unknown elements in his story. His first appearance as a balance-maker is in the Kater double reference of 1821/2.1 Robinson and Kater were working together although not as a business partnership. In modern terms Kater was the ideas man, Robinson the artisan who converted Kater’s ideas into working realities. Such collaboration between a Natural Philosopher and an artisan has been shown, above, to be quite normal at this period. The partnership was highly successful. Nothing has appeared in the way of formal documents concerning this collaboration, nor any letters between them. All we have are occasional joint references to them in the literature; see for example Fig. 130 on page 156 from the USA, plus the editor’s statement in the Kater double reference. It must be remembered that this was only a very tiny part of Kater’s activities. The balances were often attributed in that era to Robinson and Kater; their collaboration was known and understood even though only Robinson’s name, as the actual maker, was engraved upon the instruments. Thereafter, Kater’s unrecorded involvement was slowly forgotten as the years passed. It should also be remembered that balances were only a small part of the Robinson workshop’s production, and Kater was influential in other areas of Robinson’s output. Thus, in the catalogues reproduced in Chapter 3, we find phrases such as ‘invented by Captain Kater’ and ‘contrived by Captain Kater’. 1 Kater (1822). T
178 The development of the designs. Working together, Kater and Robinson clearly pushed English balance design forward a long way. A clear progression of thought has now been demonstrated in the twenty known Robinson and Robinson-style scales and balances. The principal divisions are as follows. 1) The early scales, based on the 1821 hydrostatic design in the Kater double reference. The surviving ‘Steel Scale’ and the ‘Wellcome’ are almost identical with Children’s original sketch in Fig. 46. It follows that these are clearly the earliest surviving Robinson weighing instruments known to date. The single significant difference from Children’s sketch is the addition of Wollaston’s pan stabilisers, in brass, to the ‘Steel Scale’. This simple scale design is so very clearly the basis from which the true balances were derived. 2) The Robinson production balances were made in three sizes of 5½″, 8″ and 10½″, of which the earliest known example is the ‘Wollaston’ 5½″. This is very obviously developed from the ‘Wellcome’ scale, but now with full relief of all three bearings. It features the first Robinson slit-end beam, a first clumsy attempt at a J- hook suspension piece, and the linking of the two sloping arms of the ‘Wellcome’ with a horizontal element to create a triangular arrestment frame. The cradles on the arrestment frame were copied directly from the Cavendish balance, and were the best ever used in routine balance production. The highly adjustable slit-end beam design was innovative and an excellent complementary addition to the Cavendish cradles. These twin innovations greatly increased the precision of the weighings. This leap forward was rapidly followed by the 5½″ ‘Miller’ in which the true and elegant J-hook first appeared, setting the style for almost the next century. The centre knife was now working on a continuous centre plane through the centre light of the beam. Robinson beams were therefore captive from here on, and could be released only by slackening the two massive finger bolts. All three planes were of agate from this point forward. Numerous balances followed, all very similar to the ‘Miller’ — the design had now essentially stabilised. Naturally minor changes followed, such as the use of a larger gravity bob, and the two larger sizes of 8″ and 10½″ needed additional braces in the beam. A true brass flag replaced the original piece of wire on the flagpost at some indeterminate point. All of this excellence and progress was somewhat marred by the rough operation of the spring-loaded lever-lock arrestment. This could presumably be overcome by frequent use, thereby developing the necessary skill to release and arrest gently. The fact that Robinson was making precision beams as short as 5½″ some four decades before the short beam was rationalised and introduced by Bunge in 1867 is, it seems, a clear illustration of how many great instrument-makers have achieved an empirical understanding of their subject and craft without any theoretical under- pinnings. This is an illusion in Robinson’s case, the theoretical basis having been supplied by Kater, whose contribution has been forgotten in later years. Robinson’s production balances were essentially all analyticals for chemical research, with the highest capacity model being of only 2000 grains or 130 g. As the mid-19th century chemist C. Greville Williams pointed out in 1857, Robinson’s balances were of little use in chemical synthesis labs, being of too low a capacity and physically too small to carry the sizeable flasks used in synthesis work.2 The same was also true for the inter-comparison of national standards such as the copies of the International Prototype Kilogram, and our own Pounds Troy and Avoirdupois. It would seem that 2 Williams (1857), Section VIII, paragraph 96.
179 Williams was unaware of Robinson’s bespoke higher capacity balances, although they are mentioned briefly in the known Robinson catalogues. It has been shown that Robinson’s production balances were highly valued and often sent for repair rather than being replaced. They were therefore long-lasting. After Henry Barrow bought the Robinson business in 1842 he continued making the same balances as before, with changes of note in only two areas. He replaced the flag with a nut travelling along a short horizontal threaded rod tucked in one of the end lights of the beam, which was a far easier way of adjusting the poise. Barrow’s other contributions to progress were in his improvements to the cabinets. His doors were restyled to make reading the pointer and chart so much easier, but they were only somewhat less awkward to use, still opening outwards; he never achieved the practical reality of the front doorslide. His insertion of a heavy slate slab screwed firmly to the cabinet baseboard was, presumably, an attempt to prevent warping either with heat or damp, with the balance stand firmly bolted to the slate. As only one balance has been found with this feature so far, Sci. Mus. 1977-340, it is unclear whether this was a special or a new production type. But he still clung perversely to the clip-on lantern, which is most unsuitable for such a heavy balance, as anyone who has moved this instrument will testify. Nevertheless, it seems clear that Barrow was equally as good a balance-maker as Robinson. 3) The bespoke long-beam balances were used mainly for the comparison of national standards. Not much can be said on these as only one, Dr Moll’s balance, has been found and studied so far, and that has been much modified by another maker. The Natural Philosopher Schumacher bought a large Robinson balance, said to be a copy of one used by Kater in the construction of the original Imperial mass standards. It was initially reported in the Schumacher―Gauss correspondence, where he replies to a request from Gauss for information on his balance in a letter of 1828. This gave very rare data on price and performance of a Robinson large balance. 3 The same balance was later reported in an English paper.4 Professor Miller used a Robinson & Barrow 15″ balance for his preparation of replacements for the Imperial Standard Troy Pound and other standards lost in the 1834 fire at the Houses of Parliament. That one was reported to have a capacity of a kilogram. Neither of the Schumacher or Miller long-beam balances has been located. Currently it is unknown whether Robinson and Barrow were as consistent in their long beam designs as they were in their short beams. The probability is not, for all the long beams were bespoke items for individual clients. Whether Robinson long beam balances were used in chemical synthesis labs is unknown. English chemists were certainly using imported long beam balances from Europe and, naturally, used English large scales for weighing reactants for high-volume synthesis work. His Influence upon later makers. From 1846 onwards Ludwig Oertling simply picked up where Kater, Robinson and Barrow had left off in the advancement of English balance design, such that he soon dominated the English balance market. Oertling had noted the limitations of the Robinson slit in its twenty-five or so years of use, and decided to ignore it as did most other makers. Yet the Robinson style persisted long after his death in 1841, such that J.J. Griffin was still offering balances with Robinson-style slit-end beams by unspecified makers as late as 1866.5 3 Peters (1860), 195. Excerpt translated by Thomas Allgeier. See Chapter 15, p.120 for the details. 4 Schumacher (1836), 457. Excerpt translated by Thomas Allgeier. See Chapter 15, p.120 for the details. 5 Griffin (1866), 38-39.
180 The ‘Adie’ copy of Robinson’s design has already been noted, but this seems to be only a high grade single, not one of a production range. Also noted were the few Robinson style balances created by Robinson’s one-time apprentice, John Dover, as late as 1851. In fact, no production range in Robinson’s style is known other than those by Henry Barrow, yet Griffin had a source of Robinson-style balances for sale as late as 1866. The probability is that these were of Continental origin. In Europe, August Oertling copied the Robinson slit but not the Cavendish cradles or the Robinson Tulip, and Oechsle and Reimann produced some balances in the style of Robinson, as well as in those of other makers. So, what of Robinson’s influences upon later makers at a lower level of quality? Was his style attempted by second-grade copyists? Some slight evidence of low grade copies has been seen, but only one piece has been thoroughly examined. As this is a singular item, apparently of no importance although a real curiosity, it has been relegated to the appendices; see Appendix 7, A Pseudo-Robinson Scale. In Child’s well-known book The Tools of the Chemist, it is said that the first high- precision balances used in the USA were made by the London maker Robinson. 6, 7 However, not even a single Robinson balance has yet been found in the USA, despite great efforts by Professor John Stock and several other researchers. A few American copies of Robinson designs have been observed in their literature, such as the one by Duffey noticed in Chapter 22 on cabinet styling. The developments of Robinson’s case designs with a detachable lantern were also influential. His early lift-off lanterns, merely hooked to a drawer base, continued not only to Robinson’s death, but were also used by Henry Barrow in the Robinson & Barrow era. The later developments of an added front doorslide by various makers, were very soon displaced by fully rigid cabinets with front doorslides by all makers, both here and abroad. An unexpected consequence of the excellence of the Kater―Robinson designs is that in creating the first British high precision production balances, they also effectively created the rôle of service/repair man, as the Robinson fine adjustments would have been beyond the capabilities of most laboratory users. Ludwig Oertling’s fine balances simply perpetuated this development, and calling in the serviceman or sending a balance back for repair became conventional practice. A little more on Robinson generally may be found in Buchanan’s thesis, although some of it has been superseded by the more recent findings above.8 After the Kater―Robinson collaboration ended with the death of Kater in 1835, a similar duality continued to push English balance design forward – the pairing of Professor Miller and Henry Barrow. After 1846, it was Miller and Ludwig Oertling who continued to progress the development of English weighing; while Miller employed Oertling to construct exotic balances such as the Miller Vacuum, he appears to have had no detectable influence on Oertling routine production designs. Thus Robinson was forgotten for some one hundred years, until at an indeterminate date in the late 1950s, the Robinson slit was reintroduced by William A. Webb Ltd in a revision of their White Monarch pharmacy dispensing scale. How this came about is completely unknown, due to the deliberate destruction of company and family records during a house move. 6 Stock (1986), 239, 244-245. 7 Child (1940), 81. 8 Buchanan (1982), 156, 160-164, 169-174, 208-216.
181 Miscellanea. It is of course impossible to say how many more Robinson survivors still exist out there in the world, but there have been rumours of a few, all untraced as yet. For example, Peter Simmons, an Oertling senior service engineer, reported to the author that there had been one in the chemistry department of Queen Elizabeth College in Kensington, but it has now proven impossible to find. Schumacher’s large Robinson balance has not been located, likewise Miller’s large Robinson & Barrow ‘1843-1845’. Joseph Black (1728―1799) seems to be the first British Natural Philosopher to have indulged in fully quantitative chemical experimentation, in 1752 to 1755. From about this time onwards the demands of the Natural Philosophers in Britain and Europe drove their instrument-makers, such as Robinson and Bate here in Britain, and Pistor and the other members of the Berlin School of balance design, onward to ever greater refinements. These ever finer instruments allowed their users to push the boundaries of science ever further outwards. Research and instrument-making fed off each other in a cyclic process, as they still do. This was as true for balances in that era as for the better-known clocks, sextants, microscopes and telescopes. Additionally, the instrument-makers were competing with each other in a purely engineering competition, and as the industrial revolution really got underway here in Britain, so machine tools filtered down from the great companies like Whitworth and Maudslay to the medium size engineering factories and eventually into the smaller workshops of the instrument-makers. Finally, a note on a little-known and lucrative sideline which Robinson was running with William Hyde Wollaston. The work of Smithson Tennant and Wollaston in making malleable and ductile platinum available to industry, and their research work on the isolation of other platinum group metals, have already been recounted. Wollaston then found that having isolated a few pounds of palladium, there was no real market for it other than as a gold/palladium alloy for the graduated scales on a few astronomical instruments; this most notably in collaboration with Troughton. Meanwhile, the steel pen nib had been patented by Bryan Donkin in 1808, and eventually displaced the traditional quill pen. These split steel nibs were found to be rapidly corroded by the acids in the inks of the period. Gold tips were tried but were too soft. Eventually Wollaston found that a 4:1 alloy of rhodium and tin made a very hard long-lasting tip for gold nibs. Several thousands of these tips were sold by Wollaston to Robinson, making a significant profit by this sideline. Robinson made them up into gold pens as his sideline, to some unknown profit. 9 9 For more on this see Usselman (2015), 130-134, 371.
182
CHAPTER 26 Valediction HIS IS THE PLACE for the author to make a few miscellaneous observations before saying ‘goodbye’ to his readers. It has to be reiterated that this somewhat impoverished account of Robinson as a maker of balances is due partly to a shortage of records, but mainly to a shortage of surviving balances. Much has been established in this work but much more remains to be done. I want to take this opportunity to indicate how I feel future research in this area should go. The most pressing need is to find more surviving Robinson balances, especially the larger ones such as the Robinson & Barrow ‘Miller 1843―1845’, the two Katers A and B and the Schumacher. The origin and development of English fine balance-making has now been covered, or more correctly, much of it has been covered, as follows: The late 18th century and early 19th century pre-Robinson, by Stock and herein; Robinson, in Stock (1969), Stock (1973), Buchanan (1982) and herein; Bate by McConnell and in Oliver (2020); Oertling in Buchanan (1982) and Oliver (2020). For a start, what remains to be done is for some scholar or scholars to do a thorough study of the great missing gap, that of the very long-lived De Grave company in its many manifestations, which was covered only superficially in Oliver (2020). Further, I hope that other scholars will be inspired by the work here to write a full biography of Robinson to include all aspects of his output. Ideally this would be by one expert in each field ― telescopes, microscopes, sextants, theodolites and the magnetic instruments ― to add to my specialist contribution on his balances and weights, all pulled together and edited by one author. In particular, it is to be hoped that his ancestry, education and training will be uncovered at last. This is perhaps the place to emphasise the academic neglect of balances, weighing and weights generally by historians of scientific instruments. I believe that an appreciation will occur in years to come as Robinson, Bate, Oertling and their contemporaries recede ever further into antiquity. Another factor may well be that the supply of the more exotic antique instrumentation will be drying up, and newer instruments and different topics will be researched instead. One hopes that balances, weights and weighing will be included in those areas. I would like to suggest some useful investigations for future researchers to consider. T
184 1) An achievement of the current work was the establishment, via X-ray fluorescence spectroscopy (XRF), of the very high purity of the platinum which Robinson used for his ‘fractional’ weights. This indicated that Robinson’s geometric wire ‘fractions’ were from 97½% to 99% pure platinum, and some indication of the impurity patterns was found. The author believes that this should be taken much further, using each known part set of Robinson ‘fractions’ held at various institutions in a collaborative effort, with all samples examined on one XRF instrument by one researcher in one study. This would make a splendid PhD project, establishing the exact purity and the impurity patterns of the other platinum group metals, copper and any other metal impurities. For this an authentic standard of each platinum group metal and of copper would be required. Furthermore, the probability is that in the 1820s and 1830s Robinson was using platinum prepared by Smithson Tennant and Wollaston for a few balance pans, and using Wollaston’s extruded platinum wire for his geometric ‘fractions’ masses. Even if Robinson’s wire ‘fractions’ are not Wollaston’s they will still have been made of very early commercial platinum. It would be most interesting, as a part of this study, to compare the purity and the impurity patterns of Robinson’s wire masses with the samples of Wollaston platinum wire held in the Science Museum set of his wires, notes and letters, Sci. Mus. 1946-228/1. 2) A full and accurate translation of Magellan’s paper of 1781, discussed briefly in Chapter 4, has become really important and necessary, as I feel that it has been mis- represented. The scholar required would need to be competent in 18th century French, and have a keen grasp of the early weighing technology discussed herein. I can supply translations of some of the obscure French technical terms, but not all. 3) Not one of the many cone beams noted in this work has had its internal construction examined. In many cases this would seem to be impossible, but for the Ramsden cone beam, endoscopy via its tiny, covered, inspection port seems to be a practical possibility. The questions are basically: a) exactly how were the two cones held together; b) exactly how were the centre knives fitted; c) and an examination of the internal CG adjustment would be interesting. The reader has now before him/her all of the available information on T.C. Robinson the man, his company and their scales and balances, as of late 2025. Having completed this work as far as I feel able to go, I can now return to collecting and conservation, and I hand the wider research on to the next generation. For my part I will continue to advise and assist, where possible, all collectors of old balances, especially those of Robinson, Bate and Oertling. I will also continue to collect and collate any extra information which may arise on these three makers. Finally, I must lay claim to all of the mistakes within this book, however few or many. Numerous friends and colleagues have checked various sections, have advised and admonished and corrected at various stages in this work, but the responsibility for the truthfulness or otherwise of the final product is entirely mine. I will be pleased to hear of any errors detected by readers. Please contact me at: barryoliver.frsc@outlook.com Barry J. Oliver, LEICESTER, DECEMBER 2025.
