Sir George Gabriel Stokes, Bart (1819–1903): his impact on science and scientists

1. Introduction

This paper focuses on the history, rather than the mathematical physics, of Sir George Gabriel Stokes at the Royal Society.

The epigraph above, an extract from Stokes's obituary notice written by Lord Rayleigh, implies that the impact of Stokes on nineteenth century science might be best characterized not by his mathematical physics work (as the historiography generally has it), but by the assistance he provided to other scientists in their endeavour to further scientific knowledge.

We might reasonably assume that a quotation from an obituary notice exaggerates by inclining to the hagiographic rather than providing a more considered reflection. The selection of evidence presented here however confirms that Lord Rayleigh was nothing less than honest and precise in his language. The key understanding required is the measure of Stokes's authority over scientists in addition to his selfless willingness to assist them. He became one of the two Secretaries of the Royal Society in 1854, only 7 years after Royal Society reforms had placed the emphasis of the Society (and its criteria for Fellowship) on scientific achievement and excellence rather than on patronage or privilege [2, p. 362], [3, pp. 216–217], [4, p. 104]. Stokes embraced this notion with an apparent natural (or, at least, undisputed) talent for recognizing scientific excellence, and with serious intent and stamina to improve substandard work presented to him. While he remained the junior of the two secretaries (by only 1 year) until the retirement of William Sharpey in 1871, he took sole responsibility¹ from 1854 for the so-called ‘internal scientific work’ of the Society [5, p. 98], [6, pp. 323–324]. This mainly involved reading papers submitted to the Royal Society's two main scientific publications, Philosophical Transactions of the Royal Society (Phil. Trans.) and Proceedings of the Royal Society (Proceedings), and corresponding at length with authors in order to suggest—and sometimes enforce—improvement before allowing the work to be considered for publication.

Stokes's original researches—his experimental and theoretical work—influenced Victorian science in ways that are undeniably profound. He formulated a theory to explain the phenomenon of fluorescence (he coined the word), allowing the then newly discovered technique of spectroscopy to be extended into the ultraviolet part of the spectrum [7].² The Navier–Stokes equations are regularly applied today in any investigation involving fluid flows.³ In optics, the Stokes parameters are universally employed to describe the state of polarization of light. In mathematics, the Stokes phenomenon is central in the modern theory and physical applications of divergent series. I argue here however that his most influential work is his far broader, longer-lasting and far less well-documented impact on Victorian science and on extensions of scientific knowledge generally throughout his long tenure as joint-Secretary and then President of the Royal Society, a total period of 36 years (1854–1890). It was during this era that seminal advances were made in the understanding of the physical sciences, including electrical and magnetic phenomena, the essential concepts underpinning the new science of thermodynamics, the kinetic theory of gases and (towards the end of Stokes's life), the beginnings of investigations into atomic structure.

For much of this long period Stokes acted as the Editor of Phil. Trans. and Proceedings. He maintained and forged new peer review practices, chose appropriate referees for each paper (assuming much of the work himself) [8], arbitrated between the differing views of referees and author in lengthy official communications between them (and in lengthier unofficial correspondence with the author) and ultimately (with very few exceptions over his tenure) judged the scientific and technical merits of papers to standards he defined and upheld.

By ‘technical merit’, Stokes meant underpinned by rigorous mathematics, preferably based on fundamental and well-established Newtonian laws. Stokes's mathematics was the product of the ‘Cambridge school’, the mathematical Tripos in which pure algebraic skills were forged and tested under gruelling examination conditions [9]. Stokes graduated in 1841 as ‘Senior Wrangler’—first in the mathematical order of merit for the academic year—and also as ‘First Smith's Prizeman’. The Smith's Prize was an examination of different character involving the application of mathematics to problems of natural philosophy, in which students were expected to show insights not required in the Tripos [10] and success in both examinations resulted in the reward of Stokes's immediate election to Fellowship of Pembroke College,⁴ where he remained for 66 years.⁵ Stokes's mathematical techniques and experimental outlook later influenced generations of British scientists, including James Clerk Maxwell, J.J. Thomson and Lord Rayleigh, to all of whom he acted as mentor, mathematical tutor and examiner. Many of the physicists of the Victorian era conversed in the language of wrangler-level mathematics, and at the centre of them all Stokes, Lucasian professor for over 50 years, was the master. To those scientists not so well equipped to deal with mathematical theory, Stokes selflessly provided the assistance required to underpin theory. This paper provides examples of both categories.

The primary sources from which the historical impact of Stokes on his contemporaries may be best appraised are, for the most part, held in Cambridge University Library (CUL) [11]. There are a variety of difficulties in accessing these documents. First, there are approximately 68 000 separate pages (equivalent to 25–30 000 individual documents, significantly larger than the Darwin collection), mostly comprising letters from his correspondents rather than Stokes's own words. Secondly, they are accessible only via reels of microfiche as the original documents are regarded as too fragile or precious for direct examination by researchers, and this adds considerable amounts of time (and some mechanical manipulation) to the effort of accessing any specific item. Examining documents via the projecting lens and screen of the microfiche reader, however good the original photographic record, adds its own complications to the task, especially as, thirdly, the handwriting of Victorian scientists (not least that of Stokes himself) generally offers no compromise to the eye of the interested historian.

While it seems likely that these inconveniences compound to frustrate historians, the lack of scholarly focus⁶ on Stokes [12] should not be so glibly explained. This brief paper also begins consideration of the broader question on why Stokes seems neglected in the historiography compared to many of his peers. The discipline of the history of science itself, the paths of its development since 1945, may have some significant bearing on this question and I will offer below some preliminary comments on how and why this might be so.

My research programme comprises the examination of specific contributions made by Stokes to the work of other scientists, tracing original submissions through refereeing reports from Stokes or other referees (relevant where Stokes has used his authority to amend or dismiss a suggestion) and correspondence between referees, authors and Stokes through to eventual publication in a Royal Society journal. The extent to which Stokes has influenced the published version of the relevant submission are assessed and compared to the historical credit afforded him. The measure of such influence is generally found to be far the greater of the two. I examine here only three case studies of the many available, two of which relate to eminent physicists of the Victorian era with extensive published biographical histories in which Stokes's part, given the evidence below, seems curiously understated [13].⁷

The final case study relates to a less well-known scientist—in fact a chemist, so outside Stokes's generally recognized sphere of expertise—but on whom Stokes's impact was nonetheless formative in terms of his published oeuvre. This case demonstrates the extraordinary lengths to which Stokes would go to support, guide and occasionally direct in ways that were not always ultimately fruitful nor (it might be inferred) wholly appreciated. The completeness of the historical record of this correspondence, which is as far as I know unexamined from this perspective, provides insight into the vicarious means (and the extraordinary investment of time and effort) by which Stokes sought to extend scientific knowledge, even sometimes in a minor way, through his work with other scientists.

2. Case study 1: William Thomson (Lord Kelvin) FRS (1824–1907)⁸

Of all Stokes's several cordial relationships with other scientists, undoubtedly the most significant was his friendship with Thomson/Kelvin evidenced by the warm correspondence between them over 55 years. This was transcribed and published with explanatory notes by David Wilson in 1990 [14], and provides a treasure trove of opportunity for historical analysis of scientific developments over the Victorian era. There was no contemporary scientific issue that was not, or could not be, discussed between them. The letters from Thomson/Kelvin are full of ideas, false starts, crossed purposes, requests for information or for confirmation of points of view, and above all, mathematical physics. Stokes's responses are (with minor exceptions) more sober, reflective, cautious, and (occasionally) calling Thomson/Kelvin to order. And they add yet more mathematical physics. Together they form an extraordinary resource to which many biographers have eagerly referred selectively, but no attempt has yet been made to set the correspondence in context as a background to the scientific developments of the day. It will be argued here, based on a few examples only, that the extent to which Stokes contributed to Thomson/Kelvin's many accomplishments is not yet fully appreciated. Further study will undoubtedly produce additional supportive evidence in due course.

Thomson was the younger man by 5 years. He had graduated as second Wrangler and first Smith's Prize winner in 1845—Stokes had graduated in 1841. They had similar backgrounds—Irish-born, schooled in Britain (Stokes in Bristol, Thomson at the University of Glasgow, where his father was Professor of Mathematics) and then through the Cambridge Mathematical Tripos, both being tutored by ‘wrangler maker’ William Hopkins (1793–1866). Wilson believed that they ‘certainly met by early 1845’ to discuss Thomson's Smith's Prize examination [14 (preface, p. xli)] In mid-1846, Stokes was one of the many contributors of testimonials supporting the appointment of Thomson to the Chair of Natural Philosophy at Glasgow University, a position to which Thomson was elected on 11th September 1846 (at the early age of 22) [15, p. 113].

Thomson/Kelvin is well known for his technological achievements, mainly the submarine cables allowing communication over vast distances, including between Ireland and the Newfoundland (and thus, with other cables already in place, between other outposts in an expanding telegraphic network including London to New York) [16, pp. 161–162]. His paper on relevant theory (dated 3rd March 1855) published in Proceedings was unequivocal from the outset concerning the contribution of Stokes:

The letter in question⁹—apparently the first between Stokes and Thomson for some months—concerned a paper by Michael Faraday examining the time taken for electrical conduction to pass through some tens of miles of submerged insulated copper wire in comparison to similar wire that was not submerged [18]. In his letter, Stokes theorized the reasons for the relative ‘considerable magnitude of the time concerned’ in conduction through a submerged wire and asked if Thomson knew of any other reasons for the phenomenon.

Thomson responded with great enthusiasm, writing two long letters of mathematical physics that treated electrical flows as analogous to heat flows, thus allowing for leakage of electrical force as ‘radiation losses’ [14, ref. 115]. Stokes's own immediate response no longer exists in the CUL archives, having presumably been removed for printing in Stokes's published Mathematical and Physical Papers where an extract thus survives [19, vol. 4: pp. 61–62]. His complex mathematical analysis leads him to state: ‘Your [Thomson's] conclusion as to the American wire follows from the differential equation itself which you have obtained’. He then reached additional conclusions on how the physical structure of the cable (as functions of electrical capacity, resistance and length) would impact on the time taken for electrical signals to pass along the cable. All of Stokes's mathematical analysis became grist for Thomson's arguments and practical designs.

I am not concerned here with the complex mathematical and scientific nature of the ensuing extended conversation between Stokes and Thomson. This has been discussed by Smith and Wise in their detailed biography of Lord Kelvin [15 (commencing p. 448)] I have little to add to their history here, apart from noting that their observations of Thomson's misunderstanding of some details of Stokes's questions and his immediate reformulation of the issue of principle that attracted him—in this case, the ‘practicability of sending distinct signals along such a length as the 2000 or 3000 miles of wire that would be required for America’¹⁰—are characteristic of Thomson/Kelvin's occasionally arbitrary responses to Stokes (or presumably anybody else) when lost in enthusiasm for the scientific matter at hand. (I also note here, parenthetically, that Smith and Wise's otherwise commendable biography tends to treat Stokes as a passive interlocutor provided to allow for necessary exposition on the thoughts of the protagonist—e.g. ‘In a letter to Stokes, Kelvin explained … [some matter of scientific significance]’. Smith and Wise are far from alone among biographers in such treatment of Stokes, as my continuing researches are revealing). It is the personal response of Stokes to this correspondence that is more interesting, at least for my purposes. For on 1st December 1854 Thomson requested some clandestine activity:

This seems bold enough, but Thomson requested yet more to enable him to write out the relevant theory:

Many in Stokes's shoes would have responded sourly to Thomson's stated intention to profit from ideas that were demonstrably founded upon (or at least benefiting from) Stokes's own contributions. But Stokes was not so easily upset¹³ nor apparently motivated by the opportunity for monetary return [20], and by return of post he provided the letters as requested together with more mathematical physics relating to the behaviour of electrically charged objects in a submarine environment.¹⁴ Over the next few months the conversation continued to extend the mathematical theory, although some of Stokes's contributions appear to be missing from the CUL archives.¹⁵ But soon, at least, Thomson released Stokes from his (one-sided) obligations:

The patent mentioned by Thomson was in place.¹⁶ It was the first of twelve patents on the electric telegraph taken out by Thomson/Kelvin (occasionally in association with others—although never Stokes) over the next 40 years [21].

Thomson/Kelvin never ceased to attempt to mine the valuable resource he had in Stokes. Almost forty years after their correspondence began, he reported that ‘I always consult my great authority, Stokes, whenever I get a chance’ [22, p. 168]. He was always eager to work more closely with Stokes, even from the very beginning. When the Chair of Mathematics at Glasgow University became vacant in January 1849 on the death of Thomson's father, Thomson urged Stokes to apply for the position. From Stokes's perspective this was an attractive proposition which he considered most carefully, but ultimately decided that he could not in good conscience undertake Glasgow's ‘religious test’—the requirement to worship within the Presbyterian denomination of the Protestant faith (Church of Scotland), rather than that of the Episcopalian denomination of the Anglican Church [15, p. 48], [16, pp. 88–91]. This was devastating news to Thomson—he declared his ‘state of agitation’ concerning the decision (an unusual display of emotion in their chain of correspondence)—and brought all of his considerable powers of persuasion to bear over the next fortnight, taking the view that conformity to the polity of the Church of Scotland ‘would be in no way inconvenient, or repugnant to your feelings'.¹⁷ It seems clear that Thomson was less scrupulous on religious matters, regarding the tests as no more than a necessary formality.¹⁸ Stokes admitted to being ‘…staggered in my purpose by your powerfully written letter’ ¹⁹ but in the end Thomson had to admit defeat. On such grounds were major scientific decisions made.

A second opportunity arose in October 1859 on the death of John Pringle Nichol (1804–1859), professor of Astronomy at Glasgow. Thomson again urged Stokes to apply for the position. While noting the importance to science to get Stokes ‘out of London & Cambridge, those great Juggernauts under which so much potential energy for original investigation is crushed’, Thomson was not being wholly altruistic: ‘…I feel it an immense advantage to myself to have you so near’.²⁰ The religious tests were no longer an issue—they had been discontinued in 1853. But by this time Stokes was well established as Lucasian Professor and at the Royal Society; it seems that the position was not attractive enough to pursue.

Thomson/Kelvin became one of the foremost scientists of the era. Other scientists looked to him for approbation. Knighted in 1866, ennobled as a baron in 1892, he gained fame and fortune from his technological achievements. Stokes, for his contributions to this and other of Thomson/Kelvin's projects, gained neither. Kelvin's oration at Stokes's funeral in 1903 was filled with his admiration of Stokes, and of his contribution to science. It ended with a most poignant tribute which illustrates the extent to which he acknowledged his debt to his old friend and confidant:

3. Case study 2: Sir William Crookes FRS (1832–1919)

In his well-researched and highly readable biography of Crookes, W.H. Brock names Stokes as ‘one of Crookes's many ‘invisible’ helpers…’ [23, p. 11] although, as I will argue here, there is ample evidence to show that Stokes was in fact the main and most significant influence on Crookes's scientific work. Crookes himself thought so when providing his record of correspondence²² with Stokes to Larmor:

The evidence below shows that this should not necessarily be regarded as a wholly exaggerated claim. In the same letter to Larmor, Crookes explained his longstanding practice of sending Stokes's letters (the handwriting of which he had difficulty making out) to his own printing office, for his head printer ‘could decipher almost anything’, and thus made his own private printed copies. (This possibly reveals as much about Crookes's wealth as it does the value he placed on Stokes's words.)

The intellectual debt owed by Crookes to Stokes is perhaps greater than that of any scientist except Thomson/Kelvin. The significance to science of their collaborative relationship is supported by the 100+ of their letters published in Larmor's Memoir, together with more than 200 others held in the Stokes collection at CUL. The extent of Stokes's labours on behalf of Crookes is not easy to characterize in general terms. Stokes was involved—as always—with details throughout. These included checking mathematics, suggesting the design of experiments and apparatus (it seems that Crookes's glassblower was remarkably adept), taking responsibility for the final version of theory underpinning experimental results and submitting Crookes's papers to a sympathetic and competent referee, the first choice usually being James Clerk Maxwell [24, p. 300]. Harrison reduces the role of Crookes to that of a mere coordinator of the work of Stokes and a paid assistant (and of the funds provided as Government grants) [25, pp. 115–116], but this is surely trivializing Crookes's extraordinary technical ability and experimental insights into new fields of investigation discussed below. Besides, Crookes regularly offered generous appreciation of Stokes's collaboration, even if this might be (as I argue here) not wholly acknowledged by historians of science.

The collaboration began slowly. In 1852, Stokes helped to explain anomalies in Crookes's photographs of polarized light through crystal [19, Vol. 4: pp. 30–37]. From 1856 to 1869, correspondence comprises minor notes on lines in the solar spectrum as revealed by Crookes's photographs, and Stokes's corrections to the precision of Crookes's calculations of the atomic weight of thallium (of which Crookes was the discoverer) [26, p. 160]. The resulting paper on thallium was read at the Royal Society on 19th February 1863 and Crookes's Fellowship of the Royal Society followed on 4th June of the same year.

By the late 1860s, Crookes had become enthused by the then fashionable entertainment of spiritualism.²⁴ Convinced that this had a scientific basis that could be studied, he submitted a paper to the Royal Society in June 1871 that was rejected by Stokes. Furthermore, the two secretaries of the Royal Society (Stokes and William Sharpey) refused to attend related experiments. On the suggestion that the Royal Society should examine the phenomenon, Stokes wrote:

Stokes's stated issues with studies of spiritualism included lack of experimental rigour and concerns about charlatans. Brock inferred that Stokes was protecting the reputation of the Royal Society [23, p. 144],²⁵ while Wilson [6, p. 326] suggested that the basis of spiritualism—the belief that spirits of the dead might be capable of communication with the living—conflicted with Stokes's heterodox religious conviction that the human spirit remained in an unconscious state until resurrection of the saved.²⁶ All of these views seem equally plausible, and probably all operated to some degree.

The correspondence on spiritualism, later set out at length and from a distinctly biased perspective by Crookes [27] brought a coolness to his relationship with Stokes. But a thaw was not long delayed. Crookes developed the ‘radiometer’—a simple device comprising a delicate balance with vanes allowed to spin freely inside an evacuated vessel—to test a new force that Crookes felt might have some bearing on his spiritual researches [23, p. 163]. The phenomena demonstrated by this device implied that beams of visible light possessed a momentum that could drive a ‘light mill’. Crookes demonstrated the radiometer to the Royal Society in April 1875 [23, p. 173]. His assertion on the momentum of light was eventually disproved by new papers²⁷ and demonstrations read to the Royal Society in March 1876 by Stokes himself, with Crookes in attendance [23, p. 225]. These overwhelmingly confirmed that it was residual air in the radiometer that caused the windmill effect, not the direct action of light. Despite this apparent setback, Stokes was there to emphasize that Crookes had discovered something truly new to science:

Crookes continued to experiment with ever more highly exhausted vacuums, testing with improved apparatus Maxwell's counter-intuitive theoretical discovery that the viscosity of a gas is independent of its pressure [28].³⁰ Crookes appeared to feel somewhat out of his depth on theory—a draft of his 1881 paper ‘On the Viscosity of Gases at High Exhaustions' was sent to Stokes on Christmas Eve 1880 with some trepidation:

Help was, as always, at hand. Stokes offered to add notes of his own to discuss Crookes's results, and to add mathematical theory and observations to relate those results to Maxwell's theory. Crookes accepted the offer with alacrity on 3rd January 1881: ‘I felt this all along but I did not like to suggest it to you knowing how much your time is occupied.' Stokes's notes appeared as an appendix to Crookes's paper in Phil. Trans. [29].

Crookes's experiments with the radiometer led him to investigations of electrical discharges in glass vessels in which Crookes had attained higher levels of exhaustion than realized elsewhere. New phenomena including the cathode rays thus produced impressed even Stokes:

In this new research, Crookes and Stokes proceeded together. The nature of cathode rays was then unknown, and Stokes suggested differentiating experiments to decide whether the rays were particulate or electromagnetic in nature [19, Vol. 5, p. 262], [30].³³ Together Stokes and Crookes carried the argument among British scientists that the cathode rays were negatively charged particles, while German physicists were convinced otherwise [6, p. 326].³⁴ Stokes set out the principal arguments, and descriptions of the experiments undertaken (mainly designed by him), in unusual settings—first, his Annual Address as President of the Victoria Institute on 15th July 1896 [31],³⁵ and then in his Wilde Lecture³⁶ 10 days later [30, esp.p49–53]. In both, he extended his talk to speculate on the cause and nature of Röntgen Rays (X-rays), discovered inadvertently only a few months earlier during work that duplicated Crookes's own experiments with cathode rays. But by this time, at the age of 77, Stokes's limitations in theoretical physics were becoming apparent—his explanations of X-rays were rooted in his belief that light phenomena could only be explained as ‘vibrations in the ether’ rather than in Maxwell's electrodynamics. He and Crookes were however correct, at least in terms that would be understood at the time, concerning the nature of cathode rays. On 31st March 1897 Stokes wrote to Crookes that the ‘Kathodenstrahlen’ (the German term for cathode rays) are ‘not rays at all, but streams of molecules’.³⁷ Thirty days later at the Royal Institution, J.J. Thomson announced his discovery of what is now called the electron.

This case study on William Crookes concludes with some more general observations on his continuing correspondence with Stokes, which lasted until December 1901, by which time Stokes was 82 years old and little more than a year from the end of his life. Crookes never ceased to rely on Stokes for mathematical assistance (e.g. a new and more accurate formula for interpolating lines in spectral analysis in mid-1895), for experimental advice (e.g. on a large spectrograph, about which over ten letters were exchanged from November 1896 to March 1897) and for some occasional plain speaking:

To the end, Stokes was checking Crookes's sums—which clearly needed checking:

With that, and a few more letters on radioactivity and spectra of rare earths under electrical stress, the correspondence petered out. Overall, Crookes's extravagant assertion that ‘precious little’ of his own work would be left over if what was owed to Stokes was deducted does now seem something of an exaggeration. Crookes supplied extraordinary technical ability and an indefatigable approach in applying it, especially in improving apparatus for producing and measuring vacuums at micro-atmosphere levels and better. His experimental insights led to new fields of investigation, particularly on cathode rays (if not on testing the claims of spiritualist fraudsters). Stokes, for his part, provided experimental suggestions, mathematical and theoretical underpinning of results, support with publication and in the awarding of grant funding, and steadfast advice to help avoid gross error. Both made significant input into their collaborative effort and, based on the evidence above, I suggest that the resulting extensions in scientific knowledge were beyond the potential achievement of either acting alone.

4. Case study 3: Arthur Smithells FRS (1860–1939)

The third case study explored here purposefully involves a lesser (although still well-respected) light in the scientific community. Arthur Smithells was a chemist (i.e. not a natural Stokes's correspondent) and at over 40 years younger, not of Stokes's generation either. Born in Lancashire, his education was mainly undertaken in Scotland. He attended Glasgow University for 2 years, attending some lectures delivered by Stokes's friend Sir William Thomson. Acquiring a passion for science and chemistry in particular, he enrolled at Owens College in Manchester,⁴⁰ becoming a pupil of Henry E. Roscoe FRS. After gaining some teaching experience at Manchester High School for Girls, and after post-graduate work in Germany (including some months spent in the laboratory of Robert Bunsen), he was appointed Assistant-Lecturer in Chemistry at Owens College, then to the Chair of Chemistry at Yorkshire College in Leeds⁴¹ at the early age of 25 [32].

In early 1892 Smithells (together with a co-author Harry Ingle) published an article ‘On the Structure of Luminous Flames’ in the Journal of the Chemical Society [33], a subject which had already drawn the attention of Stokes [34]. Smithells took the opportunity of writing directly to Stokes, commencing a very lively correspondence over the next 6 years. The collection of their correspondence held in Cambridge is helpful to historians in several ways. From about 1879, Stokes had been using a typewriter and keeping carbon copies of many of his letters, thus solving two problems in one. Smithells's handwriting, contrary to most in the archive, is a neat cursive script. Thus, there is a full and relatively accessible record of their discussion, and from this some idea can be gained of the relentless tide of scientific logic and suggested experiments that Stokes offered to his scientific correspondents on a daily (occasionally twice- or even thrice-daily) basis.

The collection⁴² listed in Wilson's catalogue [11] comprises 275 letters, 96 written by Smithells and 179 by Stokes. Analysis by numbers of pages simply highlights the imbalance—179 pages by Smithells, 394 by Stokes. Given that Stokes's work was typewritten, the number of words on each page would lead to even greater disproportion, but by now the point is established. Stokes himself admitted the disparity in a March 1894 postscript: ‘I am afraid I have been rather pestering you with too many letters’, to which Smithells responded with good grace: ‘Please do not suppose that your letters however numerous can ever be otherwise that [sic] a source of intense gratification to me. I prize your interest and advice as a real privilege’.⁴³

My purpose here is, again, not to examine the scientific content of the Stokes–Smithells correspondence, which has already been considered in some depth by Robert DeKosky [35]. I intend instead to examine the impact of Stokes on Smithells's published work, and the extent to which he has gained (or failed to gain) associated historical credit.

In December 1893, Stokes read a draft paper written by Smithells (with much assistance from Stokes) on ‘The Luminosity of Gases'. Stokes's reply (which included his usual raft of suggested amendments) demonstrates that, even in this late stage of his career (aged 74) and after he had retired from the Council of the Royal Society in 1892,⁴⁴ he retained some authority in the world of scientific publishing: ‘I have read this paper, and I should like to see it published in the Philosophical Magazine’ in which it did indeed appear (together with Stokes's amendments) in two parts soon afterwards [36,37]. Smithells's gratitude to Stokes was unequivocally stated in the second part of his paper: ‘I have to express my grateful acknowledgements to Sir G.G. Stokes, Bart., for the interest he has taken… and for his valuable and ever ready counsel’.

After publication, the Stokes–Smithells correspondence entered a quieter period. A request from Smithells (in December 1894) to proofread the second part of his paper⁴⁵ (published in January 1895) was followed by a brief flurry (April–September 1895), mainly concerning a paper by Prof. Vivian B. Lewes [38] that contradicted Smithells's theories on the combustion of hydrocarbon flames. Lewes's conclusions were quickly refuted privately by Smithells and Stokes to their own satisfaction, and Smithells suggested (on 29th April 1895) compiling the objections together for publication. This suggestion could barely have been stated with more deference:

Stokes, excusing himself on the grounds of ‘illness of a relative’, dealt with this suggestion politely but clearly: ‘I think you had best write your paper independently’⁴⁷ and Smithells duly produced a draft of ‘a paper which I propose to send to the Chemical Society’.⁴⁸ After incorporating Stokes's amendments, Smithells's objections to Lewes's theory were read at the Chemical Society on 7th November 1895 [39].⁴⁹ With a final couple of letters from Stokes suggesting new experiments (which Smithells undertook) to quantify the temperature of a candle flame, after September 1895 there followed a 2-year hiatus. In December 1897 however, communication recommenced at an even higher level of activity (figure 1):

• Download figure
• Open in new tab
• Download PowerPoint

The revitalized exchange of letters began with a 6-page effort from Smithells, which began:

Stokes was once again immediately receptive, and over the next three months over 70 letters were exchanged—as usual, the greater proportion (over 60% in number, far greater in words) from Stokes. These letters provide, in their rapid to-and-fro, clear evidence of Stokes's keen interest and input into Smithells's work at this time—chasing down and reading papers, providing points of view, questioning assumptions, suggesting experiments and helping with the interpretation of the results in a continuous cycle that led to yet further suggestions.

On January 22nd 1898, Smithells sent Stokes a draft of a new paper on carbon spectra that had been the main subject of their original correspondence:

By this time, Smithells knew what to expect from his audience. Despite a death in his close family,⁵² Stokes provided ten pages of typescript of closely argued queries and suggested designs for additional experiments, followed by further short letters dated 28th and 30th asking Smithells to proceed no further with his paper until another more substantial letter, then being worked up, had been received. In the meantime, Stokes had been carrying out his own experiments (together with the Cambridge chemist Prof. G.D. Liveing) to confirm Smithells's conclusions and he reported on these activities on 31st January.⁵³ In response (on 1st February 1898) after immediately accepting all of Stokes's suggested changes in detail, Smithells asked Stokes for advice on where this paper might be published.

‘As to your paper’ responded Stokes in a handwritten note on 4th February 1898, ‘I should be happy to present it to the Royal Society’,⁵⁴ an offer that Smithells accepted ‘gladly, and feel it a great honour’.⁵⁵ Stokes, in turn, provided specific instructions on rigour of argument, required referencing, and the requirement for a short abstract (to be immediately published in Proceedings) that, altogether, would not look out of place for submissions to today's academic literature (including the Phil. Trans.).⁵⁶ By the end of February, Stokes was suggesting that the finalised paper should be sent to him for communicating to the Royal Society, as:

Smithells's acceptance of this ‘as a very great honour’ was returned almost immediately.⁵⁸ But doubts about details meant that Smithells continued to tinker with his paper (encouraged by Stokes's ready willingness to engage almost daily). By the end of May Stokes was beginning to agitate for submission, because ‘there are only 2 more meetings … before the summer recess’, and ‘I think the argument is quite in a state to fit the paper for presentation…’.⁵⁹ Smithells continued to prevaricate, submitting a draft but requesting a delay in publication until the autumn to hammer down a particular experimental difficulty.⁶⁰ Giving Stokes additional time however served no useful purpose—having expressed ‘no particular object to presenting the paper before the vacation’ on 31st May 1898, by the next day he had ‘just one point on which I am disposed to take a different view from you…’ and by 10th June this had expanded to a 5-page letter of further observations and suggestions.⁶¹ Smithells, under pressure with College administration and fraught with doubts about the rigour of his work, was forced to delay once more. His response, for the first time, contains a touch of iciness—‘Let me now thank you sincerely for your valuable criticism’⁶²—and the correspondence fell silent until early October, when Stokes again urged submission, reminding Smithells that the paper needed a ‘chance of getting known before the discussion of the claims of the candidates for the [Royal Society] fellowship comes on’.⁶³

At this point, Smithells appears to have become exhausted by experimental difficulties and Stokes's apparently endless capacity, when given a very short time to ponder, for spotting and reporting on minute holes in an argument. Besides, his candidature for Fellowship of the RS was being pursued (presumably unknown to Stokes) in another direction: ‘With respect to your kindly interest in my election to the Royal Society, I may say I have in complete readiness a paper on the electrical conductivity of vaporized salts, which my former lecturer and dear friend Sir Henry Roscoe has kindly promised to communicate to the Royal Society.’⁶⁴ This is undoubtedly the paper mentioned above (see footnote 50) in the letter which, ironically, revitalized the entire Stokes–Smithells correspondence in late 1897. The paper was read at the Royal Society in November 1898 and publication in Phil. Trans. followed early in the new year [41].

It is interesting, and not a little poignant that, after all the extended correspondence and input from Stokes on suggested experiments and observations on the spectra of gold and copper compounds (acknowledged much later by Smithells in his letter to Larmor⁶⁵), Stokes is given no credit nor even mentioned in the paper published in Phil. Trans. in 1899. Perhaps Smithells had simply tired of it all and did not dare to have Stokes read it again. After a meeting at a Chemical Society dinner and handful of further letters on experimental methods and the spectrum of carbon compounds, the correspondence apparently ended with a letter from Stokes on 18th November 1898.⁶⁶

The belaboured paper on the spectra of carbon compounds was eventually published, apparently without further input from Stokes but this time containing generous acknowledgements to him, in the Philosophical Magazine in 1901 [42] and Fellowship of the Royal Society was conferred on Smithells later that year.

5. Conclusion

The premise of this paper—that there is something wrong with Stokes's historiography—seems somewhat ambitious. Why would this be so, after 100 years of biographical study of many of Stokes's contemporaries on whom (I argue) Stokes had such impact? Nevertheless, the secondary sources are indisputably rather limited. Only one book in that 100 years has been devoted to Stokes [43], and that allocates only 11 pages (out of over 200) to his 50+ years as a central character in the Royal Society, the remainder focusing otherwise (for the main part) on his mathematical physics and offering very few specifics on his wide-ranging impact on science and scientists over that whole period. Other biographical sources do not exactly treat Stokes dismissively, but nor do they go much beyond acknowledging his assistance as a sounding-board off whom ideas are bounced without significant modification. The broad sweep of his involvement in theoretical speculation, on mathematical underpinning of argument, in the design of experiments to confirm or disprove a particular detail of theory—Stokes was generally relied upon as an expert in construction of the novel instantia crucis—and, above all, insistence on rigorous scientific argument before allowing publication in a journal of the Royal Society, is seldom mentioned and has never (as far as I can ascertain) been explored in any depth.

The curious historian might ask why this is so. The examples set out here can only scratch the surface of the extent to which scientists, well known like Thomson/Kelvin or Crookes, or less familiar like Smithells, came to depend upon Stokes's counsel and encouragement to help publish their researches in significant scientific journals. The extraordinary volume of primary source material is, so far at least, much underused in historical analysis. Perhaps the related access and transcription issues described above would deter many diligent researchers from a broad investigation of Stokes, although biographers have certainly dipped into the detail relevant to their focus.

An additional factor might involve Stokes's own character. Notoriously taciturn in meetings, he freely admitted he was ‘naturally of rather a retiring character’.⁶⁷ It appears from the evidence of the Thomson/Kelvin relationship, and the Smithells correspondence and elsewhere in the archive, that in seeking to extend the boundaries of natural knowledge Stokes sought no advantage either in financial profit or in enhancements to his scientific reputation. Yet it cannot be, as Harrison concluded, that Stokes was possessed of ‘guileless conduct in all worldly affairs’ [25, p. 92]. His long tenure in such a politically powerful position as Editor of the Royal Society's main scientific publications, in what must be regarded as a disputatious environment, points to an administrator of rare skill and temperament. His input was mostly appreciated, but (as the Smithells case above demonstrates), this was not always the case. It seems however, from the evidence in Larmor's memoir, that the Smithells case may be regarded as the exception rather than the rule.

For the final component of my complex answer to the question of why Stokes's work at the Royal Society has not yet attracted the historical attention it might merit, I turn to the development of the discipline of the history of science itself since 1945. The extent to which the science—and thus scientists—has tended to dominate the history is discussed by my PhD supervisor, Prof. Frank James, in his paper ‘Some significances of the two cultures debate’ [44]. Without restating the argument here, the resulting focus on the scientific content of the history rather than on its context and on the processes involving interplays between associated actors has led to a historical record that is not yet fully developed. A further historical spotlight on the supposed origins of the so-called ‘modern’ science in the sixteenth and seventeenth centuries led to a neglect (until the late 1970s) of the nineteenth century, and the treatments of several major scientists including Charles Darwin, Michael Faraday, James Clerk Maxwell and Thomson/Kelvin have, to one extent or another, been produced with the focus on scientific content rather than a contextualized examination that does not prioritize the science above other activities. The consequences include the historical underplaying of the impact of characters such as Stokes.⁶⁸

I conclude that Stokes's formidable original physical and mathematical researches may not comprise his most notable contribution to science. His impact may have been even more profound, but was expressed through the efforts and publications of other scientists to an extent not yet recognized in the historical record (despite being apparently well known among his contemporaries). Stokes was the hub of a vast correspondence network with which historians must eventually grapple in order to disentangle and thus identify his true influence.

Data accessibility

All data are cited in the article with full details allowing access. A spreadsheet of supplementary material provides the data from which figure 1 is compiled.

Competing interests

I declare I have no competing interests.

Funding

This research is funded by the author.