In November 1840, Henry De la Beche, the Director of the British Geological Survey, sought the services of his old friend William Sanders. De la Beche and Sanders were fairly typical of those individuals who peopled the philosophical societies in their early years. Both were, or had been, shareholders in the Bristol Institution. Sanders was a man of independent means (though still living with his parents), a reliable geologist, and according to Samuel Stutchbury, the curator of the Bristol Institution’s museum at this time, he had ‘cautiousness large’. But now their roles had changed: Sanders remained the local philosopher, De la Beche the head of the science’s new corporate body. It was in this role that the latter asked his friend if he would complete the Survey’s work around Bristol as well as draw sections of the strata exposed during the laying of the Great Western Railway. This was a project of the British Association but would also derive useful results for the Survey. The Bristol philosopher immediately realised the implications of this request and of the Survey’s presence in his district. He had long intended to complete a geological map of the country around Bristol but had seen no imperative in doing so, and had completed only that part in the immediate neighbourhood of the city. What choice had he now? As he told De la Beche, ‘it is now manifest that my time & labour will under present circumstances be thrown away unless they are made subservient to your more finished mode of proceeding.’ He had no choice: ‘there is no course to take between wholly abandoning an investigation which has hitherto afforded me much gratification, & consenting to devote myself entirely to the matter as you propose’.[1]
Sanders agreed to join De la Beche’s ‘most noble company of peripatetics’ for expenses only, though these were intimated to be in the region of £100 per season. With the work completed, the provincial philosopher looked forward to another season with the Survey, replacing his pony for the grandeur of a horse and phaeton in preparation. But to his great surprise and disappointment he was told that his services would not be required and, indeed, that his expenses for the previous year had been too much. His employment had simply been a contingency. De la Beche, wrapped up in the strategic efficiencies of the Survey, lacked any empathy with the philosophical world he had once inhabited. He was insensible to the hurt caused to Sanders, whose expertise had simply been acquired, used and discarded. But it was not the loss of the appointment which concerned Sanders so much as the loss of opportunity. The type of work the Survey now undertook was, until its arrival in the area, within the grasp of any man with time, money and application.
When Phillips detected the storm brewing within Sanders he advised De la Beche to retreat from his position and re-employ him. Sanders rejoined the team on reduced expenses – the horse and phaeton were to be replaced by foot. His relationship with his friend De la Beche, now in the form of director and directed, had taken a dent. De la Beche would, from now on, continue to require assurances of the quality of Sanders’s work from Phillips and Stutchbury, while Sanders continued to give De la Beche reason for concern, particularly as it appeared that he would not release his documents to the Survey.[2]
Sanders was fighting the last battle for the provincial gentleman geologist. As Murchison announced to the Geological Society in 1843, and Sedgwick also commented in 1846, De la Beche’s ‘little army of good observers’ were transforming the fieldcraft of geology, establishing a new level of resolution in the enterprise of stratigraphy which usurped the role of the local philosopher and the distinguished gentleman geologist.[3] With its patronage of talent, sheer manpower, professionalised approach, higher resolution in fieldwork, better base maps and use of surveying equipment, the Survey was a formidable force.[4]
The Bristol philosophers were experiencing the beginning of a change in the culture of geology, a creeping professionalism resulting from increased government support for the science. It was a change which the York philosophers would not begin to feel for another five years, and then only in the form of their scientific associate, Phillips, then working for the Survey. The Survey had first been established to resolve the stratigraphy of landscapes very different from those which surrounded York or Bristol. In Devon and Cornwall, Wales and the Welsh borders, the rocks had a perplexing complexity. It was here that the Survey and some of the leading lights in the gentlemanly art of geology – particularly Murchison and Sedgwick – found their apotheosis. It was here also that the nature of fossil collecting and its relationship to the creation of knowledge was to be transformed. In the process the true nature of the science was revealed, epitomised by rivalry, jealousy and controversy. It was this which drove forward the science’s professionalisation.[5]
By the early 1840s it was increasingly apparent to its participants that the future of geology lay not in the gentlemanly science of past decades but in an entirely different culture. The transformation the Survey brought about more generally in how geology was investigated in the field can also be seen in the way fossils were gathered into collections and analysed in the museum. Ultimately the Survey would take absolute control of the primary resources it needed to create geology. In the scientific marketplace, which geology had become, all participants were striving to increase their power and level of control. But, in a mirror of the wider commercial community, the laissez-faire culture of geology was to be invaded by a system of government control and even monopoly.
The root of this transition can be detected in the early years of the Survey, when Henry De la Beche attempted to utilise in Devon that same kind of philosopher network which had proven so successful for Smith and Phillips in Yorkshire. As will be seen, this networking model had its limitations when local observers began to question the need to serve the science’s doyens and instead saw openings for their own philosophical enterprises.
Progress by symbiosis
The Survey’s exploration of Devon and Cornwall was essentially a collaboration between two men: Henry De la Beche, who had been engaged in a geological survey of the two counties since 1832, under the Board of Ordnance, and John Phillips.[6] The two men first came into correspondence in 1830, De la Beche apparently taking Phillips at face value as a sophisticated man of science, unaware of his financial circumstances. But it was some four years later, when De la Beche’s survey seemed besieged on all sides, that he most needed a palaeontological ally of Phillips’s calibre. De la Beche’s claims for coal plants in the Grauwacke of Bideford in Devon had met with ridicule from Murchison and Lyell; the latter were sure that the plants must indicate the later coal measures.[7]
De la Beche’s approaches came at a time when Phillips, exhausted by work on his Mountain Limestone volume, had resolved not to attempt any further treatises on regional geology. Indeed, the British Association had already asked him to compile a critical catalogue of British fossil species. This was a challenge of extraordinary scope. It involved the construction of a systematic catalogue of all described British and Irish species as well as those he might examine himself, together with their localities and stratigraphic distribution. The project was entirely Phillipsian in interest and scope, and was probably of his own manufacture. The crisis affecting palaeontology, which this catalogue sought to remedy, was one of nomenclature: synonymy was rife. Inadequate description, unavailability of texts, arrogance, laziness and monument-building had resulted in a multiplication and complication of names. Authors often had inadequate field knowledge; many believed that if ‘they occur in the same stratum’ they must be the same species. As Lonsdale told De la Beche, ‘as every one may call what he thinks fit a species, every one may be right’.[8] The consequences for geology, and particularly for stratigraphy, of not having a shared language of fossil names were particularly serious. The reliability of pronouncements on rock and fossil distribution could be undermined and the whole science turned into a farce.
The practical means for completing the catalogue, however, had not been determined and it was, at this time, in crisis – failing from a desperate want of support from collectors. To date only one (probably Gilbertson) had supplied him with material and Phillips was considering restricting his survey to northern England where at least he was assured of access to collections.
De la Beche’s rival, Lyell, was championing his own man, Gerard Paul Deshayes, to resolve this same problem. De la Beche regarded the Frenchman as a ‘mere conchologist’; some questioned whether he was even that. The Danish geologist, Dr Henrick Henricksen Beck, considered him ‘as very far from being a first rate conchologist!!’ Thomas Underwood reporting on one of Deshayes’s first papers, read in Paris, agreed with De la Beche: ‘a tolerable conchologist but no geologist’.[9] Deshayes had given up a badly paid medical practice to risk all on a career in natural history, keeping himself going by writing for encyclopaedias and selling his services. He had worked hard at making himself a proficient conchologist and Lyell had employed him in 1830 to teach him conchology. Lyell believed that a conchologist, possessing a knowledge of variation within extant species, would be better equipped to discern fossil species. As president of the Geological Society he was also able to award Deshayes £25 from the Wollaston Fund to assist his researches. But many saw Deshayes as one of the culprits of untamed description: ‘many of the identifications of fossil & recent shells are disputed & will be still more so, by eminent conchologists & Deshayes may earn a still higher title to the medal by answering them as I know he can.’[10] Lyell’s belief that fossils should be the province of the zoologist rather than the geologist was to gain ground over the next decade, culminating in De la Beche’s appointment of Edward Forbes to head the Survey’s palaeontological work. Fitton and Buckland were already encouraging botanists to take responsibility for fossil plants.
De la Beche harboured deep uncertainties about the stratigraphic utility of fossils. Like other European geologists, his faith in characteristic fossils had been dealt a blow by Elie de Beaumont’s notorious Alpine discovery of ‘a mixture or alternation of [Lias] belemnites with coal measure plants’. But the troubled De la Beche would ‘rejoice’ at Phillips’s open mind – especially on the utility of fossils. Phillips was, for example, entirely unsettled on questions of comparative rates of change in organic remains from terrestrial and aquatic realms. He could see no reason why certain parts of the globe might not be immune from the faunal revolutions seen elsewhere; similar faunas and floras need not be synchronous because they were as much dependent on physical conditions as on time. And he had long been aware that one had to think beyond the definition of rock formations defined by restricted and simplified notions of characteristic fossils. He and Phillips were both field men who saw geology as a multifaceted scientific subject: ‘you and I do somehow jump much together – keep poking the physics and chemistry well into the old story – it is the only way to drive out that kind of geology which might be termed le Geologie des Dames.’[11]
In his correspondence with De la Beche, Phillips hinted at the practical utility of his most recent work on Mountain Limestone fossils, which was soon to appear. The fauna there described would be vital to understanding the rocks of Devon, where De la Beche was also finding fossils traditionally associated with the Mountain Limestone.[12] This formation with its rich fossil fauna was, principally through Phillips’s work, the best known of the older marine rocks and provided the key to distinguishing the biological indices of still older formations.
De la Beche saw the necessity of getting Phillips’s assistance and immediately determined that it might be gained by offering to solve his problems with the critical catalogue. In carrying through his survey of Cornwall and Devon, he could send, at no cost to Phillips, specimens as ‘remarkable’ as those arriving from America. In return Phillips would provide information on these fossils for the Survey report. The needs of both men would thus be met, though De la Beche encouraged Phillips to move beyond his catalogue and begin work on a definitive volume on organic remains. Phillips already had a view of what the latter should comprise: ‘a satisfactory view of the natural affinities of organic remains, their relation to existing forms, their dependence on physical conditions, & the legitimate inferences from the facts known concerning them.’ The arrangement, which involved an exchange of favours and not of cash, as mutually beneficial as it seemed, demonstrated De la Beche’s ignorance of Phillips’s financial circumstances. Phillips no doubt carried himself with a level of education which belied his relatively humble background; but the two hardly knew each other. De la Beche was on a salary for this work; Phillips was not. Nevertheless, Phillips responded positively, but warned De la Beche that such an undertaking would take many years. ‘You will do the work philosophically, it has hitherto been done in a way which marvellously lacks philosophy’, De la Beche encouraged.[13]
Overcoming the collectors’ resistance
De la Beche’s Devon investigations were two-pronged: field investigation and collecting gave him a framework for interpreting structure; examination of existing collections would enhance this interpretation by providing palaeontological keys to the age of the rocks. He immediately began to search out existing collections from which Phillips could extract useful data. Those of worth within the two counties were few. Amongst them were the collections of Revd David Williams of Bleadon and Revd Richard Hennah, the latter containing important Plymouth material. An appeal for material was also made at the British Association meeting in Bristol in 1836.[14]
All seemed to be going well but as the geology of Devon took on an increasingly controversial slant, views became polarised and collections became political territory. At first, Phillips was promised the breadth of material represented in the best private cabinets in Cornwall and Devon. Then, suddenly, it seemed that Hennah’s collection would only be available to Murchison and Sedgwick. Murchison believed that it was interpretation of the fossils, rather than the fossils per se, which would prove whose reading of the geology was correct. However, the power of interpretation in some degree relied upon possession. De la Beche had earlier understood this, and taken exclusive control of Revd William Bilton’s collection of fossil plants from Bideford which De la Beche had told him to send to the British Museum rather than the Geological Society. Williams also began to see increasing scope for his own scientific progression, utilising his collection to ‘determine the disputed question’ himself. Major William Harding of Tiverton also had an important collection of North Devon fossils which De la Beche hoped to borrow. The best collection, however, lay in the hands of ‘a rising geological man’, Robert Alfred Cloyne Austen who, like Williams, also planned to exploit it for his own ends. ‘I now have all that the quarries produce’, he proclaimed, a suggestion which, since he continued to add new species in the following months, probably meant that he was assured of all new finds through the patronage of quarrymen. Likewise, the philosophical aspirations of Samuel Rowles Pattison, a Launceston solicitor, might remove another collection from Phillips’s view. For De la Beche and Phillips to succeed in their respective projects the collectors needed to be brought under control: ‘To advance science we must allow men to work from all sorts of motives, but “every gentleman for his peculiar fame” is sad work.’[15]
What De la Beche needed was a mechanism for extracting collections from individuals who were excessively possessive of the information they contained. His solution was to tell them that Phillips would return collections with specimens determined and labelled. Unfortunately, he made this assurance before talking to Phillips, unaware of the additional labour involved.[16] In practice, the idea was a good one, as it at least released the collectors’ grasp on their treasures. If collectors were to pursue their own theories then accurate identification, which they had no way of achieving themselves, would assist them greatly and, by this means, cost them nothing. De la Beche was to continue to worry about the supply of collections but his tactic worked, helped no doubt by Phillips’s ‘bland’ and apparently impartial disposition.
Progressively the collectors were won over and by early 1839 Phillips had seen most of the collections and even borrowed material from Sedgwick. De la Beche continued to fret over access to Austen’s fossils, but need not have done so. Austen had suggested that local rocks at Newton Bushel were of Mountain Limestone age. This was based on a comparison of his fossils with those illustrated in Phillips’s book, which provided Austen’s only access to information on the fauna of this formation; his rich and extensive collection contained not a single comparative specimen. But book illustrations alone were far from ideal and he sought a higher authority. Amongst those he approached was Murchison, who suggested that he ‘could not be better referred’ than to Phillips, and contact was made. Austen and other collectors now sought exchanges with Phillips and his Yorkshire contacts in order to build comparative collections of Mountain Limestone fossils, seeing these as crucial to local interpretations of the contested strata.[17]
In December 1837, as Austen began to make his own views on the geology of Devon known, De la Beche moved into South Wales in search of further clues, and promised to continue to send material to Phillips. ‘During our progress through Wales I will take great care to procure fossils for you. I suspect we shall not be long before we turn out some private collections of Carboniferous Limestone shells, corals, etc.’[18] Phillips had only begun work on De la Beche’s specimens in the previous month, a year after the deal had been struck. It was then that he realised the difficulty of his task. His past work had relied upon a limited number of well-curated collections, to which he had fairly open access. He had visited the collections personally, compared the contents with his existing knowledge, and then organised them and selected what needed to be appended to his cache of description-worthy specimens. The new arrangement was far less manageable. It involved the haphazard arrival of a miscellany of specimens which the collectors themselves, in isolation, had determined as potentially important. De la Beche, for example, had instructed Williams ‘to select those specimens which he considers most determinable’.[19] But how was Williams to know what Phillips considered ‘determinable’? Phillips had never in the past expressed a preference for ‘cabinet quality’ specimens, and indeed he and Smith frequently utilised mere fragments. Fine specimens would certainly make the task easier but were unlikely to give a comprehensive view of either fauna or stratigraphy as they often represented rarities. Thus useful comparative material remained out of reach.
Each box or basket arriving in York contained perhaps 100 fossils representing miscellaneous types, localities and rocks. If they were well preserved and conformed to an existing species, the task of identification was relatively simple. But inevitably each basket contained new forms. These could not be identified or described in isolation; similar species, from which they needed to be distinguished, had to be determined concurrently. And yet these might not turn up for many months, if at all, relying as the arrangement did on the collector’s ability to discern importance.
The effectiveness of the arrangement depended on the connoisseurship of the collector. While this protected Phillips from being swamped with specimens, it introduced a source of omission caused by the relatively parochial interests of the collectors. Most collected from an extremely limited area which they knew well but which encouraged the asking of questions concerning the internal consistency and relationships of rocks and fossils within that area. In particular they were interested in finding new species, and species indicative of age, without being equipped to assess how these might be determined (especially as they had little comparative material). Phillips was most interested in the comparison of similar species across wide geographical areas, which might reflect changes of lithology as well as temporal difference.
Beating the collector at his game
De la Beche had expected Phillips to meet his part of the bargain and supply a report on the fossils early in 1838 when he was putting the finishing touches to his report on the geology of the two counties. However, Phillips’s need to earn an income from other enterprises had made this impossible. After some negotiation, Phillips was given a contract to complete a work on the fossils of Devon and Cornwall for the sum of £250 which was to be paid on its completion a year hence.[20]
Immediately, De la Beche instructed Phillips to visit the area to see the sites from which the fossils came. Austen urged Phillips to locate himself at Ogwell House. De la Beche supported this plan, telling Phillips, ‘He is a thorough good fellow, very talented and in everyway the right thing. He will work with you in all honesty and good faith – something now-a-days’. Austen had been of great assistance to Murchison and Sedgwick but had found his local knowledge, collections and ideas used without proper credit. If Phillips located himself at Austen’s residence it would do more than assure this local gentleman of some intelligent discourse on geology. Phillips’s mission was ‘zoological’; it would place Austen’s collection at the heart of any published solution, illustrated and fully determined. The latter was particularly important to Austen as he was keen to take advantage of the celebrity of local fossils in foreign exchanges. He was already in negotiation with Philippe Édouard de Verneuil, a Parisian lawyer and gentleman geologist, who had gathered a notable collection of fossils from Eifel and Belgium, but until Phillips saw the collection, there was always a risk of sending away fossils which might be important to the story. However, in the late summer of 1838, Austen did acquire some comparative material, returning from a trip to France with 600-700 Transition fossils. As the battle for Devon continued, Phillips continued to delay in entering the field. The local collector could wait no longer. In 1840, just before Phillips entered Devon and before the two had met, Austen exchanged material for well-localised ‘Silurian’ (according to De Verneuil) fossils from the Rhine and Eifel districts. [21]
It was essential for Phillips to enter the field if he was to truly understand the affiliations of the species he was isolating in these collections. His aim was primarily to overcome the biases of cabinet collectors: ‘my plan of operations inevitably required me to study specimens, before going to examine their situations in the rock’, he told De la Beche. ‘My object in this latter task was not in fact geological, I wished to remedy the very evident incompleteness of the evidence contained in the collections shewn me by inspection of the thousand fragments in the quarries, which are often decisive as to changes of form with age &c., though little valued in the drawers.’ Despite his intention to ‘plunge into the mystery of organic remains of Devon and Cornwall’, Phillips still did not depart for Devon.[22] Committed to other work to pay his day-to-day bills, and unlikely to receive payment for his Devon work until the report was complete, Phillips found his financial situation again interfering with his science. Now Austen was about to exit Devon for Guildford and De la Beche was in the embarrassing situation of Phillips’s report being due in one month but as yet nowhere near complete.
Phillips’s apparent inactivity had also been caused by a keenness to complete his work on the collections he had been sent. He had by this time drawn all the species in his own and his friend, John Edward Lee’s, collection, and sketched some of those from Dr Harvey B. Holl, Pattison, and Harding. He also had ‘a general notion of the meaning’ of Williams’s collections, and had seen a few of Hennah’s fossils. But now the collectors were becoming frustrated and uneasy about Phillips’s slow progress; their collections had great potential in settling a major geological controversy but would have no part to play in it if locked away in York or London. Harding was now asking for the return of his collections and in November, Williams also withdrew his fossils, which Phillips had held for a year. These were no longer to be available to the Survey and Williams asked Phillips not to publish any figures of his best specimen as these would add impact to the work he now planned to produce himself. Holl had taken his to America; Hennah’s remained under the control of Sedgwick and Murchison; and Austen’s were in three places. Lonsdale informed Phillips that Austen too urgently wanted his specimens returned, but added a word of encouragement: ‘The subject of Devon geology … wants to be treated zoologically which is not attempted in the past but which I should be much pleased to hear we might expect from you.’ Despite the withdrawal of his material, Williams, who was not a man who naturally held high opinions of others, also wrote to encourage Phillips to visit Devon, perhaps fearing the rivalry of Austen: ‘You can do it – Prof Sedgwick and Mr Murchison I think will not and Mr Austen (who lives there) can not. The latter has not sufficient power of comprehension and combination for the great scale on which matters have been adjusted there.’[23]
By the end of February 1840, Phillips had ‘disposed of many troublesome affairs’ and intended to set off for Devon and Cornwall in the following month, stopping in London on the way to view some of the fossils which had been sent there, and at Bristol to pick up his friend William Sanders. ‘You will have a famous companion in Sanders’, De la Beche’s remarked.[24] Phillips was also encouraged to visit Austen’s residence at Gosden House, Guildford, where his collection was available for inspection together with much information regarding localities. Again he did not take up this opportunity, wishing to get into Devon as soon as possible.[25]
Before leaving he tried to ensure that the fossils he still had on loan were returned to their respective owners. He asked De la Beche’s advice on hostelries, contacts, collections and localities; the only person he knew in North Devon was the Revd William Bilton. De la Beche did better than to send the information and met up with Phillips in London with a supply of maps and suggestions for the itinerary. He then accompanied the pair into North Devon so as to show them some of the more obscure localities. The tour, which was to be undertaken mainly on foot, would take two months – ‘the minimum of time adequate to see well so embarrassing a region.’ It would focus on the field. Collections were visited but fewer than perhaps De la Beche intended; many were, anyway, incomplete, specimens having been sent away to other workers. And while collectors were extremely keen for the two to visit, Phillips’s preference, as ever, was for the field where considerably more could be learnt. Here he was surprised at how easy it was to gather fossils: ‘At last I am sure I can get more fossils than most of my friends’, he told De la Beche excitedly. It appeared to Phillips that valuable time had been lost in utilising Ordnance men and others to find fossils since they lacked the geological expertise to undertake the task effectively: ‘I have so many new things that I wonder what your young men were about in not supplying you with more of the organic fossils’. There was considerably more potential here than Phillips had anticipated; the planned sortie now appeared rather inadequate. ‘However, as it is, I suppose I shall add 50 species to what are commonly known or thought to be known & many interesting localities’, he told De la Beche reassuringly, ‘I suppose the Devonian trilobites will be fully half a dozen, and in the Coddon hill grits &c. ten species of shells. Think of that!’[26]
Phillips and Sanders left Devon with two or three hundredweight of fossils, Phillips returning with them to York where every species was to be described. He had entered Devon with little expectation of finding new species, so ‘indefatigable’ had been previous searches, doubtless believing that the rocks were less productive than they actually were.[27]
His relative success resulted from two flaws in the fossil-collecting model De la Beche had brought into being. The first was the reliance upon local collectors and untrained Ordnance men, the assumption being that anyone with sufficient time could make an effective job of finding fossils. Some, at least, lacked more than a basic understanding of fossils and rocks but to a truly inductive science this should not matter; rocks were simply a ‘black box’. Phillips found that he could out-collect his collector friends because his collecting was not purely inductive, despite his own claims. He entered the field equipped with an increased understanding of the occurrence of fossil species in time and space. This enabled him to maximise his collecting. The second flaw arose from the dependence on collectors to select and send material. So often decisions were made on completeness and not on the morphological characters Phillips considered important. Phillips knew that this had led to omissions.
Phillips and Sanders had further increased their productivity by collecting bulk samples for splitting back at base, and by making casts from cavities left by shells and corals in rocks so weathered they defied collection. Weathered heaps of fossils were used as indicators of the productivity of quarries, enabling the two to maximise their collecting by focusing on those beds which were obviously fossil rich. Pattison also gave them some tips, showing them how lime nodules from the Petherwin Group could be partly burnt in a kiln to induce fracturing and reveal numerous goniatites and Clymenia. Those fragmentary remains which Phillips knew would not be found in cabinet collections also gave him an insight into the distribution of species which could not be obtained in any other way. His prior examination of collections had assisted his fieldwork, enabling him to easily identify what he found, and distinguish that which was novel: ‘Mr Williams, among some valuable specimens which he sent me in 1837, enclosed a fine example of this curious fossil. In 1840, I found at Brushford, in North Devon, the young of the species, and impression of the plates about the pelvis; and in Major Harding’s collection saw with delight the full-grown body and arms.’[28]
Figures and descriptions
The aims behind Phillips’s activities in the cabinet and in the field were two-fold. The first was to satisfy his need to create a definitive catalogue of fossils and the second to provide De la Beche with a palaeontological key to the stratigraphy of the two counties. The former would be resolved by figuring all the species of fossil found in Devon and Cornwall regardless of whether they had been figured previously. This might at least clarify problems of synonymy. It also meant that he need not worry that his competitors had already published species which he had drawn and described during the long gestation of the new work. His figures remained valid and he would happily give priority.
This comprehensiveness was important to Phillips; he felt it had been the key to the success of his Yorkshire treatises. This was not simply to satisfy stratigraphic ends but also in recognition of changing perspectives on fossils. De la Beche certainly supported the idea of a specialist fossil work as an entity separate from the standard Survey report; fossils were increasingly seen as zoological subjects worthy of study in their own right without the burden of stratigraphic and lithological detail. His ambitions with this work had grown in grandeur: ‘This volume should be the commencement of a national work on the organic remains of GB & Ireland.’[29]
However, such discussions were premature. De la Beche’s attempts to entice funds from the government had only succeeded in commissioning Phillips to undertake the work. He waited before approaching the government for yet more money. By the autumn of 1840, Phillips was heavily committed to publication and would carry the project forward himself, if no other means of funding were acquired and provided that the Ordnance department subscribed to an adequate number of copies. The volume was to have upwards of 500 figures showing some 240 species, and as such represented a considerable financial risk.[30] However, by early December, with the completed figures and descriptions ready for the press, De la Beche had secured the necessary funding.
Phillips remained concerned about some of the names he had applied, aware that many new species, named in manuscript in Germany, were now in frequent use in Britain unsupported by any published description.[31] Pattison, Harding, Ottley, Lee and other collectors continued to correspond with Phillips, and send specimens, right up to the publication of what was to be called Figures and Descriptions of the Palaeozoic Fossils of Cornwall, Devon and West Somerset. It finally hit the presses in the summer of 1841. For these men it was a most important opportunity to see the products of their efforts in print, although Phillips reverted to his old style and gave collectors only general acknowledgement and very occasional reference.
The much sought-after collection of Austen, so late attended to, was still being sent to Phillips as the final plates were being engraved. Austen, having at last had material returned from Phillips, was now able to send specimens he was sure Phillips had not seen. As the text was already set and the figures complete, Phillips added these to a short supplementary section. However, as he had to insist on a deadline, it was more important that the work look professional and properly arranged than complete; ‘I really cannot introduce any more species without quite spoiling the book by irregularity of arrangement’. At the end of May 1841, Phillips had 274 species for the volume; another 40 had not reached him. Austen had a long time earlier talked of 300 species; Lonsdale was more convinced there were truly only 100; De Verneuil, working with Murchison, included 1100 species in the Devonian. Phillips, in the end, published 277 species or ‘animal structures’ but believed the list would grow to 350, 400 or even more. He had good evidence for this: ‘my own examination (with very good assistance from a diligent fellow-labourer) led to the detection of new forms in almost every locality we visited.’[32] He immediately made plans for a supplement.
Phillips was most anxious that his work should not suddenly become outmoded by stratigraphic uncertainties. His fossils were catalogued according to local lithological names which lacked any wider correlation. Indeed such an arrangement was essential to the statistical study which followed the description. He also hoped that his plates would survive changes of nomenclature and thus the figures were numbered but no names were added to the plates themselves, ‘the greater part of the species will I hope thus become permanently identifiable.’[33] The fossil descriptions themselves made no direct link to strata, being bound to the only immutable truth – geographical locality.
In terms of acknowledging variation within species – ‘It is very difficult, however, to be satisfied with any distinction among the numerous forms which these fossils exhibit, in endless variety’[34] – and the influence of depositional environment, this was a work of great maturity. The quality of the plates, with each species given in several views to highlight the salient characters, remained of key importance to the success of his descriptions. His second Yorkshire Illustrations had introduced one-line descriptions to his figured fossils as an aid to identification and he had improved the figures considerably to include multiple views of fossils, something he had discussed with Brongniart in 1829. With Palaeozoic Fossils he attempted to delve further into taxonomic description. In producing a work which sought to give species a name he had undoubtedly been very successful but he had hoped to achieve more than this. His aim had been to establish a definitive nomenclature, a taxonomic framework which would stand the test of time and eradicate the overlapping and inconsistent systems of the past. In this Phillips was less successful. His designations and descriptions were often hesitant and tentative; a mixture of uncertainty resulting from inadequate material and his natural diplomacy. He made a case for revision of a group and then stepped back. For example, for the Brachiopoda, he proposed a new classification of genera but then remarked: ‘I propose to make no use of the new terms which shall embarrass the reader who prefers any other view’. His crinoids remained littered with question marks, his descriptions sunk by a lack of definitive reference collections and the ambiguity of previously published figures and descriptions. On Adelocrinus hystrix, a species of a ‘provisional’ genus: ‘Hardly any fossil among the many difficultly intelligible fragments which fill the rocks of Devonshire, has caused me more trouble … supposing it to be a crinite … I give it a name which will affect only itself’. The urge to be comprehensive meant that such material, however unreliable, had to be included. Where a previous author was dead, Phillips lost his timidity, on Cyathocrinus (?) megastylus: ‘Miller’s figures … are like these; but his reference of the columns to the bodies cannot be depended upon.’ Phillips’s transparent exposition, however, sought to hide nothing. Some species were supported by too few specimens which made it difficult to define their morphological limits. On fossil brachiopods, he remarked: ‘The inconvenience of a single term for a vast heap of disagreeing forms has been removed, but there is great fear of our falling into an opposite degree of confusion from the adoption of genera insufficiently examined, or founded on too few examples.’ These problems became all too apparent as specimens drifted in over the years from various collections, showing that some initial judgements of affinity and diagnostic characters were incorrect, and warranting re-description. Some brachiopods he found extremely variable and therefore difficult to define as species; other species seemed more fixed. Phillips laid down clear rules which enabled him to decide whether he was seeing two species or one. Species which, for example, seemed to change subtly over time he gave one name.[35] Phillips did not have the comprehensive perspective for which he had hoped; there were many published species, such as Lonsdale’s corals, of which all he could do was copy the previous descriptions. Other species he had not seen and could not compare with his own. The whole business had been something of a rush; the restricted and difficult geographical provenance of the fossils had been a straitjacket; economic and networking constraints had also caused problems and his conversion from stratigrapher to palaeontologist had not been easy. Phillips had taken all these problems head-on and dealt with them valiantly. Though not as definitive as he had originally conceived the project it did demonstrate his usual clear and logical thinking, and in utilising other published descriptions at least made a move towards eradicating the problem of duplicated names.
Fossils as numbers
The second objective of the work was to provide De la Beche’s much needed palaeontological key to unlock the stratigraphy of Devon and Cornwall. The rocks of the two counties were problematic because they had no physical link to strata of similar age, either geographically or in geological section. Phillips felt that the only solution lay in a detailed analysis of the fossils. The statistical approach he adopted was developed rather late in the project’s history,[36] although the reasoning behind it was discussed in an article Phillips published in Penny Cyclopedia in 1840 entitled ‘Organic Remains’. It is doubtful that this should be seen as a shift derived from Lyell’s widely regarded Principles of Geology(1830-1833). Elements of a palaeoenvironmental approach can be found in the literature even in the early 1820s.
The zoological approach to fossils, for which there were widespread calls, was part of a wider change of emphasis from seeing rocks and their contents as structures (structural explanation) to viewing them as the products of processes (causal explanation).[37] Both viewpoints had co-existed and all workers were imbued to different degrees with each. Phillips, like Smith, was principally a ‘structuralist’ but one whose views were informed by a wider understanding of process and environment. But since it was his ‘palaeontology’ which had attracted greatest interest he was undoubtedly seen as one of a new kind of specialist and was expected to meet these growing demands for the biological approach. The transition to a causal mode of thinking provided the keys to unravelling the causes of his uncle’s ‘bald & trifling notion of the “identification of strata” by their organic contents’ as he had described it some three years earlier. While Phillips still saw this as a ‘powerful instrument of research’, he warned that ‘geologists must not overlook ascertained facts which limit the extent and modify the rigour of the application’. These facts were: that the geographical extent of a species was limited and rarely extended beyond a few degrees of latitude and longitude; that genera (and other larger groupings) were more geographically and stratigraphically widespread than species, and therefore gave a lower resolution in stratigraphic studies; that correlation over extreme distances was difficult as few species may exist which are common to each locality; and thus for distant correlations several species of characteristic groups will be needed to establish contemporaneity – species demonstrating a diversity of organisation, that is ‘characteristic combinations of organic life’. Phillips was enveloping ‘strata identified’ within a theoretical framework, though he saw it as factual and objective. Smith’s use of fossils as ‘characteristic’ indicators used (and indeed was in advance of) the Baconian inductivism espoused by his gentlemanly contemporaries in the second decade of the century. The science had matured and Phillips’s views of its underpinning concepts had, he felt, become better understood. However, it is worth noting that Phillips remained wedded to Smith’s basic concept: ‘The scale of strata, in almost every district, is found to be marked and symbolised in almost every part by some characteristic organic form’.[38] The problem remained: how are these reliable species to be located and distinguished from those of less restricted distribution?
Given De la Beche’s antipathy towards certain of his rivals, it is perhaps ironic that Phillips should have adopted statistical methods similar to those used by Lyell and Deshayes in their study of the Cainozoic. But for the previous twenty years numerical methods had rapidly been gaining ground as a means of establishing ‘laws’ on the basis of thousands of facts represented by species. Phillips’s introduction to the fashion for botanical arithmetic can be traced back to his contacts in the 1820s and perhaps earlier. Indeed Phillips claimed that he began to make inferences about fossil distribution in 1817. From this he developed the notion that fossil type and lithology were often related, inferring an environmental relationship between seabed composition and associated species. He had also noted that within any one period species diversity increased in time (that is vertically in stratigraphic section) to a peak and then declined. These ideas can be seen in the discussions of the York society in the 1820s, and were epitomised in the generalised tables of distribution he published in his ‘Treatise of Geology’ in Lardner’s Cabinet Cyclopedia in 1837 and subsequently in his work for the Survey. Whether they can be claimed as his own ideas is doubtful as they also appear as established notions in Young and Bird’s book of 1822. However, these authors could find little evidence to support such ideas. In the 1830s Darwin read Phillips’s Cyclopedia tables in search of clues on species distribution, and found Phillips’s assertion in a Royal Institution lecture that shells decrease in number with depth ‘most strange’. [39]
Other early influences on Phillips’s arithmetic probably included the honorary curator of botany in York, the lecturer William Hincks, a strong advocate of the work of Augustin de Candolle, who pioneered these methods. Both Phillips and the society also had close links with Adolphe Brongniart who used similar tables in his Histoire des Végétaux Fossiles in 1828. In 1829 De la Beche, Mantell and, most interestingly, Richard Cowling Taylor, Smith’s former student, produced distributional tables of organic remains. Phillips’s book of the same year contained prose discussing the distribution of fossils which intimated the same kind of statistical thinking he would later put in his encyclopaedia articles.[40]
The approach of Deshayes and Lyell, which sought to trace the diminishing numbers of existing species in Tertiary fossil assemblages as an indication of age, had been criticised by Charlesworth for ignoring inconsistencies of nomenclature and the mixing of successive deposits. Phillips, however, thought the technique valid. His article of 1840 shows that while existing species cannot be traced into the older rocks, the geologist could expect to find larger groupings such as the same classes and genera in these rocks, and should see an increasingly modern fauna appear over time. This he referred to as ‘transitions from one system of life to another’.[41]
In order to use fossils as temporal markers he needed to eradicate the influence of geographical variables; he did this by referring to an internally consistent framework based on the known succession of life: the great eras – Palaeozoic, Mesozoic and Cainozoic – that he had previously described in Penny Cyclopedia and which embodied his notion of faunal transitions. Phillips could only suggest that this framework had value in Europe; the tropics were still relatively unexplored and the extant fauna there when compared with the extinct fauna in Europe (elephants, lions, etc.) suggested that such broad generalisations might come unstuck. Phillips knew that such intellectual inventions were more likely to stand the test of time if they were built on a secure foundation rather than speculation.[42] With this scheme of biological transitions in mind he used the broader conceptual term Lower Palaeozoic in preference to Murchison’s Silurian. Both Phillips and De la Beche considered the Silurian a local system.
Phillips recognised, and sought to correct, any errors likely to arise from the simplistic inductivism which underpinned the statistical approach. He knew, for example, that the occurrence of some species often depended upon environment, others were poorly defined, and still others were diachronous in their distribution (that is, appearing at different times in different places). However, he believed that if large numbers of species were used and the geographical area was small such problems should be overcome.
Phillips began by comparing Devon with similar outlying areas both in terms of age and lithology, on the established assumption that younger rocks contain more species.[43] His two outlying districts had to have been treated in similar depth to Phillips’s current work and so he chose Murchison’s Silurian (here called Lower Palaeozoic) and his own Carboniferous Mountain Limestone descriptions (referred to here as Upper Palaeozoic). Phillips knew from stratigraphic relationships that Murchison’s rocks were older than the Mountain Limestone and he expected the number of known species to place the Devon and Cornwall rocks somewhere between the two. It did not. The two counties had produced fewer fossils than either. This problem Phillips dismissed as resulting from the exceptional richness of the strata Murchison had used and the inadequacy of collecting in the South West which he had himself witnessed.
His first methodology having failed, he then compared the number of species in each class of fossil in each of the three districts. This showed that the extreme richness of Murchison’s collections in corals and brachiopods had caused the earlier analysis to fail:[44]
| Lower Palaeozoic | Devon/Cornwall | Upper Palaeozoic | |
| Corals, sponges and bryozoans | 63 | 34 | 41 |
| Sea lilies | 14 | 16 | 43 |
| Bivalves 1 | 21 | 26 | 32 |
| Bivalves 2 | 6 | 20 | 24 |
| Brachiopods | 107 | 83 | 100 |
| Gastropods | 34 | 36 | 91 |
| Bellerophon, etc. | 11 | 7 | 10 |
| Cephalopods | 41 | 44 | 69 |
| Trilobites | 37 | 9 | 10 |
| 336 | 275 | 420 |
As raw numbers these too were inconclusive so Phillips converted them into percentage form – thus giving, for example, the proportion of the Upper Palaeozoic fauna composed of corals (9.7 per cent) and so on.[45] This removed any problems caused by the differences in total number of species found in each the three districts. For example, this gave the following percentages for corals, the Lower Palaeozoic (19.3 per cent), Devon and Cornwall (12.4 per cent) and Upper Palaeozoic (9.7 per cent). This he felt characterised the fauna as a whole, the figures best being interpreted as the relative importance of the different faunal components. Thus brachiopods make up an important part (nearly a third) of the Lower Palaeozoic fauna but less than a quarter of that of the Upper Palaeozoic. Here Phillips could show that for most classes of animal (seven out of nine) the Devon fossils fell between the two comparative areas as he had hoped. Only the bivalves failed in this analysis:
| Lower Palaeozoic | Devon/Cornwall | Upper Palaeozoic | |
| Corals, sponges and bryozoans | 19.3 | 12.4 | 9.7 |
| Sea lilies | 4.2 | 5.8 | 10.2 |
| Bivalves 1 | 6.2 | 9.5 | 7.6 |
| Bivalves 2 | 1.8 | 7.3 | 5.7 |
| Brachiopods | 31.9 | 30.2 | 23.8 |
| Gastropods | 10.1 | 13.1 | 21.7 |
| Bellerophon, etc. | 3.3 | 2.5 | 2.4 |
| Cephalopods | 12.2 | 16.0 | 16.4 |
| Trilobites | 11.0 | 3.3 | 2.4 |
| 100.0 | 100.0 | 100.0 |
He then, for each animal group, subtracted these figures from each other and summed them to give ‘the sum of the differences’ between each district or period. The table below shows this calculation for the Lower Palaeozoic and Devon groups giving the sum of the differences as 34.3. Phillips obtained a similarly low figure (27.0) when he compared the Devon and Upper Palaeozoic groups but a much higher figure (54.3) – indicating a much greater difference – when he compared the Upper and Lower Palaeozoic groups. This too seemed to confirm the intermediate position of the Devon rocks. Just to prove his assertion that this was a sound methodology he showed that a similar table could be constructed to compare the Upper and Lower Palaeozoic, and Jurassic or ‘Oolitic’ strata. His notion of faunal transitions underpinned these inferences. It was widely known, for example, that brachiopods were very diverse and common in the older rocks but became less diverse and even rare later on. Phillips had taken this and converted it into something theoretical to the point of abstraction.
| Lower Palaeozoic | Difference | Devon/Cornwall | Upper Palaeozoic | |
| Corals, sponges and bryozoans | 19.3 | 6.9 | 12.4 | 9.7 |
| Sea lilies | 4.2 | 1.6 | 5.8 | 10.2 |
| Bivalves 1 | 6.2 | 3.3 | *9.5 | 7.6 |
| Bivalves 2 | 1.8 | 5.5 | *7.3 | 5.7 |
| Brachiopods | 31.9 | 1.7 | 30.2 | 23.8 |
| Gastropods | 10.1 | 3.0 | 13.1 | 21.7 |
| Bellerophon, etc. | 3.3 | 0.8 | 2.5 | 2.4 |
| Cephalopods | 12.2 | 3.8 | 16.0 | 16.4 |
| Trilobites | 11.0 | 7.7 | 3.3 | 2.4 |
| 100.0 | 34.3 | 100.0 | 100.0 |
However, in converting fossil numbers into percentages their behaviour became more complex. The problem with using percentages, rather than true numbers, is that a decline or increase in any one group creates an increase or decline in all the others. The true dynamics of changing faunas become lost in this kind of representation.[46]
Phillips repeated this approach at a finer resolution attempting to order the rocks of sub-regions of Devon and Cornwall. However, he admitted that local circumstances, such as the prevalence of limestones, could seriously affect the result. He obtained the following ranking, creating a scale of time from the sum of the differences:
Upper Palaeozoic
Interval 316
North Cornish period
Interval 132
Later North Devon period (anterior to Carbonaceous group)
Interval 264
South Devon period
Interval 374
Lower Palaeozoic
On completing this analysis he had written to De la Beche to tell him that North Devon and North Cornish fossils were nearly coeval adding that he must see upon what this depends: ‘I think you will be amused by the arithmetic at all events’.[47] In this analysis the fossils were no more than numbers – but numbers suffused with theoretical notions of fossil distribution. The identity of the fossil had been lost, the notion of characteristic forms or combinations had no place here.
It revealed perhaps the most curious aspect of Phillips’s mathematical approach. Inevitably, he sought to arrange the rocks of these two counties in temporal succession. But suppose, as is the case, that the rocks of North and South Devon and North Cornwall are in varying degrees contemporary; that they contain rocks laid down over a very long period of time and extend up into strata contemporary with Phillips’s Yorkshire Mountain Limestone. Consider also that within this region rocks of similar age but different location look dissimilar due to differences in seabed conditions at the time of deposition. And that the sequence is incomplete in all of the sub-regions. What then do these faunas mean once they are lumped together in large sub-regional groupings? Phillips’s gross mathematics assumed that they still indicated differences in age (faunal transitions) but this new form of stratigraphic analysis only served to conceal contemporaneity. Yet Phillips and De la Beche clearly understood that faunal and lithological changes might well reflect differences in seabed conditions rather than time. Indeed, none of the main factors which determine the geology of Devon and Cornwall were alien to their thinking. In this light Phillips’s mode of analysis becomes all the more curious and all the more revealing. It appears that the tremendous constraints placed on his participation, particularly those resulting from his need to earn an income, without forgetting the extreme difficulty of the geology of these two counties, led him to gamble that these different regions were not contemporary – that these numbers might well be informative.
It should be no surprise, to us at least, that when he came to check this arrangement of sub-regions, expecting similar species to be largely restricted to adjacent periods, he found numerous inconsistencies. While, for example, a large number of the South Devon corals had Lower Palaeozoic affinities, the brachiopods were as a whole more indicative of the Upper Palaeozoic. Phillips had hoped that by this simple and logical process the gross stratigraphy of Devon would be revealed but thus far the results were far from secure.
He sought to explain these problems by suggesting that species numbers might more reflect the labour put into gathering fossils rather than the true number occurring. The idea was not new. In 1829 Fitton had remarked upon Taylor’s contemporary attempts at fossil numeracy; commenting that the proportion of known species is ‘in great measure accidental’, he went on:
the industry or success of collectors, and the greater or less extent to which the contents of the conchiferous strata are brought to light by human labour, or naturally disclosed: and all these sources of inequality must for a long time affect the different strata so unequally, that any general inferences now derived from the enumeration of species must be received with considerable qualification.[48]
As each Devon collector tended to be geographically restricted in terms of his collecting activity, the species lists for each area were likely to reflect the industry and biases of the local collector. This had not been a problem in Yorkshire because the stratigraphic sequence extended across large areas and its order of superposition was clear and simple. In Devon and Cornwall the rocks were very different in each of the sub-regions. Austen, for example, collected from the quarries around Newton Bushel; his research focused on extending fossil lists. The rocks of Newton Bushel fell within what Phillips designated the ‘Plymouth Group’. In terms of number of species this Group accounted for nearly 50 per cent of all those collected. Of all the Devon collectors only Austen was to progress in geological circles and De la Beche had a high regard for him intellectually.[49] Since the method of locating new species relied to a large degree on the connoisseurship of the collector it could be argued that the dominance of the Plymouth Group in the fossil lists was really a reflection of Austen’s dominance as a collector. Austen could also use his considerable wealth to further his fossil gathering. If Phillips had relied upon characteristic species rather than bald numbers, excessive collecting could have aided correlation but unfortunately this was not the case.
Of the other collectors, Pattison and Holl contributed Petherwin fossils; Harding’s material came from around Barnstaple; and Williams, who lived on the edge of the Mendips and did not have the luxury of patronising local quarrymen, collected from North Devon. Other collectors, such as the impoverished Cornish coastguard Charles William Peach, neither had the wealth to patronise quarrymen and artisan collectors nor leisure for long sojourns in the field, though he had the intellect and application to seek out and interpret his finds.
Phillips was aware of some of these biases in his own work including his use of the Mountain Limestone as an upper comparative standard. The limited nature of its lithology produced a fauna unrepresentative of the period as a whole; more material was needed from shales and other rock types.[50] But there were other biases which Phillips ignored such as the availability of exposures. Quarries had been essential to collecting inland. These, however, were most prevalent in the hard rocks used for building and road construction, and lime-rich rocks utilised for agriculture. Strata which were particularly favoured for these purposes were exploited in numerous quarries and thus increased the chances of representing the breadth of their fauna. There were also commercial operations where fossil content was integral to extraction, such as the industry then quarrying and polishing coral-rich limestones for the production of decorative tables. These added to the over-representation of limestone fossils.
Phillips’s final analysis, however, relied much less on the blind manipulation of figures, involving instead comparisons of the occurrences of specific species with known Upper or Lower Palaeozoic affinities. This was a return to characteristic fossils as broadly defined in his Yorkshire work. From his own work, and that of Murchison and his followers, he drew up a list of 76 species occurring in the region, which appeared aligned to one of these two groups.[51] These would surely provide the long-called-for zoological key to the Devonian problem. This immediately showed a strong affinity between South Devon, Eifel and Bensberg (but then all were predominantly calcareous rocks). This might also have reflected collector bias. Austen had been keen to seek out comparative species. He actively sought Mountain Limestone species and was the only Devon collector to have acquired a large collection of Eifel fossils. All other districts, however, showed mixtures of fossils of seemingly Upper and Lower Palaeozoic affinity. Even at the finest level of resolution, Phillips’s analysis had provided no reliable results.
The statistics of collecting
To understand this Phillips tried to explain the apparent deficiencies of his methodology by enumerating the statistics of collecting: ‘The chances of occurrence of identical species at different localities in the same range of strata are much less than is commonly imagined, and in a considerable degree depend on the earnestness and completeness of the search.’[52] Many of his readers probably never read this section of his book but the mathematics is not complex and draws fossils and collecting into an even deeper level of abstraction.
Firstly, he considered the probability of finding the same fossil species at each of two sites; his reasoning was as follows.[53] Suppose the rocks of two districts contained identical species randomly distributed throughout rocks which were equally exposed and therefore equally productive. Suppose also that the total number of species buried in these rocks as a whole is N (say for example 200). We can never, of course, know this number because we can never know if we have collected all the species that exist there.
In the museum at one of the sites there is a collection containing all the species gathered from there. This number can be referred to as a (for example say 100). If a collector now goes out into the field the probability of him or her finding one of those species already in the museum is
a/N or in our example 100/200 = 1/2
If the number of species found at the second site is b (say 50 species), then an equivalent probability can be found for that site
b/N or in our example 50/200 = 1/4
Multiplying these together gives the probability of finding the same species at both sites.[54]
ab/N squared or (100×50)/(200×200) = 1/8
Multiplying the total number of species existing in the rocks (N) by this probability gives the actual number of identical species found in the two areas:
ab/N or 1/8 x 200 = 25
Of course, we can never know N, the total number of fossil species which have been preserved in the rocks of a region, but we do know how many fossil species have been found in each of two sites and how many of those species are identical. This latter figure, will be referred to as d. In the example above it amounted to 25. From this N can be calculated as:
ab/d in our example, (100×50)/25 = 200
So if both South Devon and Eifel have each produced 166 fossil species, and 57 of these are identical, as Phillips determined, then the total number of species predicted to exist in the region is 483.[55] Phillips used this to suggest that whenever one-third of the species at two equally productive places is identical the affinity between the two sites is strong, though it is unclear why he felt this was so.
Assuming that the total number of fossil species preserved in Devon and Cornwall to be around 500, Phillips used these ideas to calculate the expected number of identical forms likely to be found between his sub-regions. South Devon had produced 166 species and Upper North Devon 72. He predicted that 23.9 species should be identical but only 20 were. He also compared South Devon with North Cornwall which had produced 67 species, predicting that of these 22.25 should be identical, only 18 were. But when he compared Upper North Devon with North Cornwall he predicted just 9.6 identical species when 21 had actually been found. This, he concluded, meant that these two sub-regions had a strong affinity. Of course the problem here is that Phillips used a figure for the total number of species which had, of course, been guessed. It also assumed that all species occurred evenly over the region as a whole.
In applying his mathematical analyses Phillips had to make assumptions regarding the purity of the data which can be derived from natural phenomena and the collecting process. He knew any one rock at two distant places would show spatial variability in its productivity and particular concentrations of certain families, as well as be unequally explored.[56] But such mathematical purity also ignored the differing probabilities of finding the poorly preserved, small, difficult to distinguish (perhaps from similar species), unfashionable, obscure, and so on. These factors further reduced the chances of finding identical species at two sites.
Phillips believed that N – the total number of species actually preserved as fossils in a region – could be found and throughout his book encouraged collectors to continue the search: ‘I have added some notices of localities, which, if well searched by diligent hands, may probably yield good fruits.’ ‘The resemblance of the fossil figured to Cypricardia impressa, of Sowerby, is such as to make me suppose that further research and better specimens than my fragments from Devonshire may prove them identical; but there is some difference …’. ‘Nearly every part of this region will repay further research.’ ‘Altogether this series has yielded by far the largest proportion of the fossils of North Devon; and it is probable that further search would at once discover many more species, and prove the nearly uniform character in respect of organic contents of the whole range of the same beds.’[57]
For the vast majority of collectors, who placed a premium on rarity, Palaeozoic Fossils became a connoisseur’s guide: ‘It seems a well defined species, but I have only one specimen, which was found with several other rarities’; ‘It has rarely occurred to me in South Devon, and perhaps not at all in any other district of England’; ‘At present I am acquainted with one instance of the occurrence of Clymenia in the British Isles out of the district under review.’[58] It finished with lists of localities and the fossils obtained.
Palaeozoic Fossils clearly reflects Phillips’s diplomatic and hesitant approach to publicising his own ideas together with the developing and controversial nature of contemporary Palaeozoic geology. It was an approach which meant that in a very factionalised conflict he remained largely unaffected by rivalries. His comprehensive and authoritative volume had on publication become simply a statement of work in progress, published at a time when the debate was all but over. In 1839, Murchison and Sedgwick had established the Devonian System to encompass the most problematic rocks – rocks which lay between the Mountain Limestone (above) and the Silurian (below). Phillips had believed that it was possible to collect and figure all the known species but, as research progressed, a once apparently barren region became extremely productive with perhaps only 60 per cent of its fossils known. Matters were further complicated by the mechanism for viewing the fossils – so dependent on the whim of collectors and the inadequate resources of Phillips. The resolution of collecting was absolutely dependent on the collectors; Phillips’s own fieldwork was also constrained by time and he had expressed an intention that his visit to the counties be zoological and not geological. Rather than collect from individual measured sections his finds could only be localised to individual quarries and coastal sites. The aim was to increase the species list and his knowledge of the geographical distribution of forms, on the assumption that geographical variation would reveal geological relationships.
His statistical interlude had superimposed the purity of mathematics onto the inconsistencies of nature. The approach might work for plant geographers but it remained unproved for the purposes of stratifying ancient rocks. Here there was little or no stratigraphic differentiation of the Devon and Cornish rocks but rather large geographical categories had been chosen which may have represented strata which in terms of age were contemporary, overlapping, intermittent, incomplete, or of slow or rapid deposition. Phillips remained perceptive and inventive. This was certainly one of the most theoretical works ever published in stratigraphic geology. But the fashion for numbers had not proved conclusive, though it is doubtful that many contemporary geologists paid much attention to the mathematics the book contained. As Phillips later admitted to De la Beche:
In 1841 came forth the Palaeozoa Damnoniensia which contains views some of which were assailed & are now adopted by the same pen; & others which I believe to be of general (mathematicè loqu) value, but which none of this race of Palaeontologists are likely even to look at, because they contain some uncommon multiplications & ratios. You see I can laugh at what I have done – perhaps others may do the same.[59]
Certainly few found fault, preferring instead to pull out apparent conclusions without seeing what underpinned the premise that the rocks of Devon and Cornwall could be divided into three. It was typical of Phillips to give it all – to hide nothing – rather than just give the conclusive results. Of the latter there was really only evidence that the South Devon rocks were probably contemporary with those at other European locations, which was already known.
Circumstances had prevented his use of his standard armoury of techniques: sections, superposition and characteristic fossils (in one form or another). Had Phillips had sufficient funds to undertake his fieldwork independently, in the style he had adopted in Yorkshire, more might have been achieved. As it was, his research was far from comprehensive, it utilised methods which were largely new to him, and sidelined him into a palaeontological appendix. Phillips was first and foremost a field geologist and stratigrapher; work for De la Beche cast him principally as a palaeontologist and museum worker, the former being a role in which he had yet to prove himself according to the standards he had set for the palaeontological monograph.[60] Phillips pursued his task with remarkable clarity of mind, turning the whole problem into an abstract intellectual construct so removed from, and yet so rooted in, the observational science through which geology had always sought explanation. The achievement of Palaeozoic Fossils is to be found not in its results but in its innovative methods which synthesised so much of contemporary thinking on fossil distribution. It was an experiment or prototype which would do much to influence Survey practice. It also epitomised the misunderstanding of the much praised notion of induction. Here facts were reliant upon established ‘laws’ of fossil distribution.
Collectors continued to send Phillips fossils. Harding sent a small collection from North Devon, of which Phillips remarked to De la Beche: ‘It is interesting, & not less so because it has almost no novelty, but several pretty repetitions’; perhaps the collectors were approaching N, the total number of species preserved in the rocks. The following February, 1842, Austen finished a paper for the Geological Society’s Transactions which was to list all the Devon species in his collection; he asked Phillips for a list of those awaiting representation in the supplement. Austen was still trying to achieve a collection containing every species. However, many of his specimens were to remain wrapped up with other Survey parcels, not to be rediscovered until 1850.[61]
The results of the Survey’s pursuit of the ‘Damnonians’ was not reflected so much in reports but rather in how it modelled De la Beche’s thoughts as to how the Survey should proceed. On three counts the free enterprise, which had driven geology to its current heights, was flawed. Firstly, collectors could not be relied upon to pull in the same direction as the Survey. They were intent on pursuing their own ends, either as rivals or in pursuit of unrelated or tangential objectives. In moments of controversy they had the power of supply and used it manipulatively. This open market encouraged controversy and this put unnecessary pressure and restraint on how a survey was undertaken, particularly as the Survey was vulnerable to political criticism and had no assurances of continued funding. Secondly, Phillips had, to his great surprise, shown how an informed geological knowledge could increase the productivity of collecting considerably. And thirdly, the problems of Phillips’s participation caused by economic insecurity – where geology, for him, existed as a private business – needed to be eradicated. The new Survey would now remove itself from the philosophical network and the culture of jealousy which divided a common effort. It would internalise and monopolise the whole process of stratigraphy and its associated palaeontology. Phillips later summarised his understanding of the matter: ‘We are not to publish again a work like the Devonian Palaeozoics, with a very limited sum, & tied down to a difficult & inconvenient area’.[62] While Phillips would continue, as ever, to produce his comprehensive tables of species, he would never again use them as the basis for blind mathematical abstraction. The future lay in a return to proven methods.
[1]. William Sanders (1799-1875); Secord (1986b: n.70). For cautiousness (an admirable philosophical attribute), Stutchbury, Bristol to De la Beche, 7 September 1841, NMW 2002. For choices, Sanders, Bristol to De la Beche, 13 Nov 1840, NMW 1858.
[2]. Sanders, Bristol to De la Beche, 28 January 1842, 21 February 1842, NMW 1861-1862.
[3]. Secord (1986a: 208). A point also made by Allen (1997: 211). Neve (1983: 197) uses Sanders’s failure to publish his own Bristol geological map until 1862 as indicative of the failings of the local scientific élite and its Institution. The true situation was rather different – Sanders had no choice in the matter and the final map would certainly be greatly influenced by his Survey work.
[4]. For this, Secord (1986a: 202-8).
[5]. Rudwick (1985) on the Devonian controversy and Secord (1986a) on the Cambrian-Silurian debate are definitive and eminently readable accounts. Secord (1986b) gives great insight into the development, operation and social history of the Survey, and lists earlier work. Morrell (1988a) gives rich context to Phillips’s early years in the Survey, stratigraphic beliefs and relationship with his peers. Herries Davies (1983) provides key detail on developments in the Irish wing of the Survey, including Phillips’s career.
[6]. Rudwick (1985: 90) for background to this Survey.
[7]. First contacts recorded in De la Beche, London to Phillips, York, 22 January 1830, OUM Phillips DLB1 and Phillips, York to De la Beche, London, 30 January 1830, NMW. Later contact and brewing controversy see Rudwick (1985: 100-105, 121) and De la Beche, Tiverton to Phillips, York, 27 February 1835, OUM Phillips DLB3.
[8]. Phillips (1836). Critical catalogue see Vernon Harcourt, Wheldrake told Whewell, Cambridge, 17 October 1831 in Morrell and Thackray (1984: 88) and Fifth Report of the BAAS, (1835), xxvi. Taxonomic issue came to a head in 1842 when the BAAS appointed a committee, which included Phillips, to construct rules for nomenclature including that of priority, See Annals & Mag. Nat. Hist., 11, 259-75. For ‘they occur …’, Lonsdale to De la Beche, 30 November 1835, NMW 891.
[9]. For these opinions, De la Beche, Tiverton to Phillips, York, 27 February 1835, OUM Phillips DLB3. For Beck’s (1799-1863) comments see Lonsdale to De la Beche, 30 November 1835, NMW 891. Underwood to Webster, 21 December 1822, (Challinor 1964-1965).
[10]. Lyell to Mantell, 10 October 1830, in Lyell (1881: I: 306). Rudwick (1978; 1985: 236). For ‘many of …’, Lyell to De la Beche, 5 April 1836, NMW 900.
[11]. For stratigraphic uncertainties, De la Beche, Tiverton to Phillips, York, 27 February 1835, OUM Phillips DLB3. Rudwick (1985: 106, 121) explains that the Alpine mixture was a notorious and famous anomaly. Austen (see below) echoes this distrust ‘I am not as you know one of those who would settle any such question by reference to organic remains’ in a letter to De la Beche, Swansea, 15 December 1837, NMW 26. For ‘you and I …’, De la Beche, Falmouth to Phillips, 31 May 1837, OUM Phillips DLB6; ‘rejoice’, De la Beche, Tavistock to Phillips, 13 November 1837, OUM Phillips DLB10. On comparative rates, Phillips, York to De la Beche, 4 November 1837, NMW. On faunal revolutions, Phillips to De la Beche, 1 April 1835, NMW. On rock formations, Phillips, York to De la Beche, 4 November 1837, NMW on which see chapter one for further comment.
[12]. Rudwick (1985: 222) and Morrell (1988a) on the relevance of Phillips’s former work.
[13]. The offer, De la Beche, Helston to Phillips, York, 6 November 1836, OUM Phillips DLB4. For ‘a satisfactory …’, Phillips to De la Beche, 1 April 1835, NMW. ‘You will …’, De la Beche, Falmouth to Phillips, 31 May 1837, OUM Phillips DLB6.
[14]. Rudwick (1985) provides biographical sketches of these collectors.
[15]. For Murchison’s concerns over fossil possession and his wish ‘not to part with our weapons’, Rudwick (1985: 272), and p. 211-12 for De la Beche’s use of British Museum. For ‘determine disputed …’, note from David Williams, Geological Society to Phillips, Kings College, 19 May 1837, OUM Phillips 1837/25. On Austen, De la Beche, Tavistock, to Phillips, 13 November 1837, OUM Phillips DLB10. See also Horace B. Woodward, 1885. ‘Robert Alfred Cloyne Godwin-Austen’, Geological Magazine (NS Decade 3), 2, 1-10. Austen (1808-25 November 1884), from 1854 Godwin-Austen, became a leading geologist of the ‘second generation’. Rudwick (1985: 224) for provincials entering into publication. ‘I now have …’, Austen, Newton Bushel to Phillips, 24 January 1838, OUM 1838/2. Austen to Phillips, 25 March 1838, OUM 1838/9. Samuel Rowles Pattison (d. 27.11.1901). ‘Every gentleman …’, De la Beche, Tavistock to Phillips, 13 November 1837, OUM Phillips DLB10, see also De la Beche, Tavistock to Phillips, 29 October 1837, OUM Phillips DLB9
[16]. Williams, Bleadon to De la Beche, Tavistock, 21 November 1837, NMW 2097. De la Beche, Tavistock to Phillips, 5 December 1837, OUM Phillips DLB11.
[17]. For Murchison’s recommendation, see Austen, Newton Bushel to Phillips, York, 18 January 1838, OUM Phillips 1838/1. For exchanges see, for example, Austen, Newton Bushel to Phillips, 24 January 1838, OUM 1838/2 and Pattison, Launceston to Phillips, 14 January 1841, OUM 1841/4. See Rudwick (1985: 226) for Sedgwick’s opinion on, and therefore the implications of, Austen’s lack of knowledge of the Mountain Limestone.
[18]. De la Beche, Swansea to Phillips, 12 December 1837, OUM Phillips DLB12.
[19]. De la Beche, Tavistock to Phillips, 5 December 1837, OUM Phillips DLB11.
[20]. De la Beche (1839), discussed by Rudwick (1985: 265-8). Contractual arrangements, Morrell (1988a: 16), Knell (1997).
[21]. De la Beche’s opinion of Austen is given in a letter to Phillips, 21 April 1839, OUM Phillips DLB30. For failure to credit see Rudwick (1985: 300) who also (p. 254) describes De Verneuil’s interests. For the transfer of fossils see Austen, Newton to Phillips, 28 September 1838, OUM 1838/47; Rudwick (1985: 257).
[22]. ‘My object …’, Phillips, York to De la Beche, 26 January 1839, NMW. For ‘plunge …’, Phillips, Newcastle-upon-Tyne, to Vernon Harcourt, Isle of Wight, 10 August 1838, in Morrell and Thackray (1984: 273).
[23]. On publishing his own specimens, Williams, Bleadon to Phillips, York, 29 November 1839, OUM Phillips 1839/50; 26 December 1839, OUM Phillips 1839/58; 27 May 1840, OUM Phillips 1840/24; 30 June 1840, OUM Phillips 1840/30. ‘The subject …’, Lonsdale, London to Phillips, York, 8 December 1839, OUM Phillips 1839/54. ‘You can …’, Williams, Bleadon to Phillips, York, 26 December 1839, OUM Phillips 1839/58.
[24]. De la Beche, London to Phillips, York, 26 February 1840, OUM Phillips DLB34. Remember this is before Sanders was asked to undertake work around Bristol.
[25]. Austen, Guildford to De la Beche, [1841FJN should be 1840], NMW 38.
[26]. Accompanying their entry into Devon, De la Beche, London to Phillips, York, 26 February 1840, OUM Phillips DLB34; ‘the minimum …’, Phillips, York to De la Beche, 22 February 1840, NMW. ‘At last …’, ‘However …’, ‘I suppose’, Phillips, Chudleigh to De la Beche, 20 April 1840, NMW. Phillips, Taunton to De la Beche, 27 April 1840, NMW.
[27]. Phillips (1841: v).
[28]. Ibid. p. vi, casts; p. 188, weathered heaps; p. 195, kiln burning; p. 29, ‘Mr Williams …’, describing Cyathocrinus (?) macrodactylus.
[29]. De la Beche, Bridgend to Phillips, York, 3 May 1840, OUM DLB37.
[30]. Phillips, to De la Beche, 20 October 1840 and 20 November 1840, NMW; De la Beche to Phillips, 4 October 1840, OUM Phillips DLB42.
[31]. Phillips, London, to De la Beche, 1 December 1840, (Phillips’s draft copy), OUM Phillips DLB45.
[32]. ‘I really …’, Phillips to De la Beche, 22 May 1841, NMW. Numbers of species, Phillips to De la Beche, 31 May 1841, NMW; Phillips, Llandovery to De la Beche, 25 December 1841, NMW and Phillips (1841: x). ‘My own …’, Phillips (1841: 182).
[33]. Phillips, York to De la Beche, 31 January 1841, NMW.
[34]. Phillips (1841: 43) on Posidonia.
[35]. Ibid. p. 55, ‘I propose …’; p. 30, ‘Hardly any …’; p. 32, ‘Miller’s …’; p. 52, ‘the inconvenience …’; p. xi, re-description; p. 164, species change.
[36]. Phillips, London, to De la Beche, 1 December 1840, (Phillips’s draft copy), OUM Phillips DLB45.
[37]. Rudwick (1978: 226-7).
[38]. Rudwick (1985: 122) describes Phillips as an environmentalist, though such a definition is rather limiting. His interest was in fossil distribution. For ‘characteristic combinations …’, Phillips (1840a: 491). See Morrell (1988a: 11) for Phillips’s earlier proclamation of this principle in 1835. Smith’s use of fossils is clear in a letter from Farey to Banks, 11 February 1802, quoted by Eyles (1985: 42). ‘The scale …’, Phillips (1841: 161).
[39]. Rudwick (1978) discusses the development of Deshayes’s and Lyell’s numerical methods. Rudwick (1985: 373) suggests a link to Lyell and Deshayes methods. For botanical arithmetic, see Browne (1983: 95) and Secord (1986a: 133). Smith’s distributional instructions, Phillips (1860: xxxviii). On Darwin, Barrett et al. (1987: Notebook B: 167-73); ‘most strange’, ibid., Notebook E, p. 122.
[40]. Taylor (1829); Mantell (1829); De la Beche (1830); Phillips (1829: 112-13). For other contemporary examples of distributive tables and the use of statistics in the natural sciences see Rudwick (1978), Browne (1983) and Laudan (1989: 159-60).
[41]. Charlesworth’s arguments are discussed by Strickland (1837). By suggesting the mixing of deposits and faunas, Charlesworth was giving the earliest account of what Flessa (1993) refers to as ‘time-averaging’, for which see Cadée (1991: 3-21). It is yet another example of the environmental approach to geology becoming prevalent at this time, and which Torrens (1993: 268) dates to De la Beche’s cartoon Duria antiquior. Rehbock (1983) suggests this had its roots in statistical biology; the two are undoubtedly intertwined. Phillips (1840a: 490) describes the system of ratios used by Deshayes but see Morrell (1988a: 12-13) for Phillips’s opposition to Lyell’s methods. For ‘transitions …’, Phillips (1841: 160).
[42]. Phillips (1840a; 1840b); see also Rudwick (1985: 372ff) for comment on Phillips’s new ‘trinity’.
[43]. Phillips (1840a: 489).
[44]. Names used in these tables have been simplified. Phillips used the terms ‘Polypiaria’ for corals, etc. and ‘Echinodermata’ for the group here containing only crinoids or sea lilies. His bivalves or clams were divided into two groups ‘Conchifera Plagiomyona’ and ‘Conchifera Mesomyona’. Brachiopods were ‘Conchifera Brachiopoda’. Gasteropods or sea snails were in part allocated to ‘Gasteropoda’ but Bellerophon, for example, was classed as a cephalopod in ‘Cephalopoda Monothalamacea’ even though several contemporary authors had recognised it as a gasteropod. All other cephalopods (Goniatites, Nautiloids, etc.) were grouped into ‘Cephalopoda Polythalamacea’. Trilobites were grouped into ‘Mollusca Crustacea’.
[45]. Phillips does not use percentages as such but rather ‘the proportion to 1000’ which is the same thing but offers the same accuracy without recourse to fractions.
[46]. As an example imagine there are just two groups of species and each makes up 50 per cent of the fauna. Now one group remains unchanged while the other declines to half its former size. The unchanged group will now make up 75 per cent of the fauna. This same group could have reached this same proportion if it had diversified and the other group had remained constant. Such simple interdependence does not reflect nature.
[47]. Phillips, Tenby to De la Beche, 21 June 1841, NMW.
[48]. Fitton, W.H. 1829. Presidential address, Proc. Geol. Soc., 1, 129; Taylor (1829).
[49]. Phillips even used Austen’s descriptions and figures without seeing the specimens himself (see Phillips 1841: 135-7).
[50]. Phillips (1841: 172): ‘In fact, it is chiefly from the limestones of that series, that the species have been largely and diligently collected, and neither the shales above nor those below the great mass of the limestone have been fully examined.’
[51]. Phillips (1841: 173).
[52]. Ibid. p.178.
[53]. These variables are not exactly as used by Phillips but simplify the explanation of his methodology.
[54]. This can easily be understood if one thinks about tossing two coins. There is a one in two chance of getting a head on each coin, but getting a head on both is a one in four chance (the possible combinations are, of course, TT, TH, HT, HH).
[55]. Phillips (1841: 178).
[56]. Ibid. p. 179.
[57]. Ibid. p. xi, ‘I have …’; p. 37, ‘The resemblance …’, on Cypricardia impressa; p. 182, ‘Nearly …’, referring to the likelihood of gathering new species; p. 186, ‘Altogether …’, on the Pilton Group.
[58]. Ibid. p. 49, ‘It seems …’, on Avicula anisota; p. 75, ‘It has rarely …’, on Spirifer disjuncta; p. 124, ‘At present …’, the specimen in question came from Ireland and was in Gilbertson’s collection.
[59]. Phillips, London to De la Beche, 13 January 1844, NMW.
[60]. Phillips to De la Beche, 1 April 1835, NMW.
[61]. ‘It is interesting …’, Phillips to De la Beche, 10 September 1841, NMW. Austen’s desire to make number of fossil species known (and inevitably in collections) = number of species actually existing in the field. Austen, Guilford to Phillips, York, 27 February 1842, OUM Phillips 1842/4.1. Also Salter to Phillips, 19 October 1850 OUM Phillips1850/11.
[62]. Phillips to De la Beche, 23 January 1843, NMW.
From: Simon J. Knell. The Culture of English Geology, 1815-1851: A Science Revealed Through Its Collecting (Aldershot/Burlington USA/Singapore/Sydney: Ashgate Publishing, 2000).
Simon Knell 