by Andrew Brown
As soon as Astbury heard this news, he congratulated Bernal effusively, but could not quite conceal his chagrin at Sage’s virtuosity: ‘Your letter was great news, and I can only hope you found my letter to Nature on the structure of keratin equally plausible. I have only one grumble to make, and that is that you have done the diketo-piperazine after all, and just as I was meditating an attack on it. I have had the crystals of this extremely fundamental substance quite a year now, and since my structure of wool is more or less based on an indefinite repetition of this ring, I had come, in a way, to regard it as my own private property, to be analysed simply and solely by myself. However, I suppose I must not be selfish, especially as the structure which you have got out of it is so thrilling. You undertook to examine all the a.[mino] a.[cid]s with the exception of dkpip., so I shan’t forgive you now until you carry out the investigation to the bitter end, and tell me the exact dimensions of the molecule and everything about it… I cannot rest now until I know the exact dimensions of the molecule itself. The important fact is that the main periodicity of alpha-hair is about 5.15 Å. I shall now also continue to urge you, night and day, to get me the dimensions of the molecules of glycine, alanine, leucine, glutamic acid and arginine.’44
Faced with such a brilliant collaborator, Astbury was anxious to reserve proteins for himself and wished to relegate Sage to a supporting role:
It is a fine bit of work which you have done, and I feel that together, me at the proteins, and you at their constituents, we should be able to knock a considerable hole in the subject. But I insist that you stick to the constituents, and don’t ‘snaffle’ any proteins. I am not sure whether you ought to take any poly-peptides, because the amino-acids have simply got to be worked out with all the intensity details. You have the apparatus and ‘flair’ for that sort of thing, while I have not.
With regard to cystine, I regret to say that I have a paper in press with the Roy. Soc. in which I have put forward a [unit] cell, from the powder photo only, which is not the same as yours…it seems that I am wrong – which is a pity, damn it! I argue that cystine, being the chief constituent of hair, supplies the ‘rungs’ of a polypeptide ladder into which it is incorporated. I shall be glad to learn all about cystine, as I have a lot of cunning theories as to how it functions in the structure of hair.
I should like to know what you think of my solution of the terrific mass of facts that constitute the problem of the structure of keratin… still a number of points which my conscience tells me are a little hazy, but for all that… I am convinced that structures quite analogous to that of hair are the basis of all proteins.45
Exactly one month later, Bernal, showing no irritation with Astbury’s attempt to limit his activities, replied that he had ‘finished alanine quite plausibly… I have also done aspartic acid and phenylalanine, both of which show molecules associated in pairs’.46 A month after that he had submitted a paper to Zeitschrift für Kristallographie giving his results on the crystal structures of eight natural amino acids as well as diketopiperazine and some peptides. He included details of the crystals’ optical properties as well as the unit cell and space group symmetry derived from X-ray oscillation photographs. By considering ancillary chemical evidence, he attempted ‘to fit plausible molecular structures to the cells and symmetry conditions found’.47 Bernal emphasized that these findings were preliminary, and alluded to the difficulty of obtaining precise structures because of the variability of the crystals and the limits to the resolving power of the available X-rays, amongst other problems. He urged anyone else ‘possessing such crystals of a magnitude of 0.1 mm or over’ to send a few milligrams of them to him for examination.
Bernal’s concentration on the X-ray analysis of organic substances was established within a year or two of his return to Cambridge, and was bolstered by an affinity with the researchers in the new Dunn Institute of Biochemistry at the University. The inspirational head of the Dunn Laboratory was Sir Frederick Gowland Hopkins, who, as was mentioned earlier, had been the first to state publicly in 1913 that enzymes were vital to all metabolic processes in living tissues. He went on to share the 1929 Nobel Prize for Physiology or Medicine in recognition of his discovery of the ‘accessory food factors’ or vitamins. Known inevitably as ‘Hoppy’, Hopkins was a slightly built, elegant, man of Edwardian demeanour and deep tolerance; he gathered a team around him who could stand comparison with Rutherford’s physics department as scientists, but who were far more active on the political left. Prominent amongst them were J.B.S. Haldane, Reader in Biochemistry, Joseph and Dorothy Needham, Conrad ‘Hal’ Waddington, as well as Bill Pirie and R.L.M. Synge. They were natural allies for Sage, and his close interactions with them over the next few years led to many mutual benefits, as well as a profound effect on the tenor of British science over the coming decades.
If the Cavendish and Dunn Laboratories were the two powerhouses of Cambridge science, with Hopkins succeeding Rutherford as President of the Royal Society in 1930, Bernal had the status of an urchin living on the street between them. As a married man, now with two young boys, Bernal could not live in college, but nor was he made a Fellow of Emmanuel. Many believe that this was a deliberate snub, a disapprobation of his libertine sex life and radical politics. C.P. Snow tried to get him elected to a Fellowship at Christ’s, but the cause was lost when an elderly don remarked: ‘No one with hair like that can be sound.’48 Even Bernal’s department, the Department of Mineralogy, lacked a secure place within the University. When Hutchinson was appointed in 1926, the chair of mineralogy had become vacant on the death of the previous incumbent in his eightieth year, after an unproductive 45-year tenure. The statutes were changed so that Hutchinson would retire at the age of 65 years, after a term of just five years, but the Council to Senate also set up a syndicate to plan the future of mineralogy in the University.
The syndicate included Rutherford, but not Hutchinson, who complained that he found out about it only ‘from the casual conversation of friends in London.’49 The syndicate recommended splitting mineralogy into two departments: one for crystallography and one for mineralogy and petrology, each with its own chair. To Bernal’s horror, the crystallography department was to remain true to its Victorian geological foundations and not be primarily an X-ray based enterprise. When the report came up for discussion in January 1930, he, Hutchinson, Wooster and Sir William Bragg argued strongly for the ‘new’ crystallography. Bernal tried to persuade the authorities that it would be folly to maintain an emphasis on the superficial, descriptive approach to crystallography at the expense of X-ray methods, which were not only more revealing about molecular structure, but also predictive in the sense that ‘there were now means of saying a priori that such and such an arrangement of atoms was possible and would produce a crystal of such and such properties’.50 The Council to Senate announced that they could not afford to create two chairs, but merely agreed that the teaching of crystallography should be separated from that of mineralogy and petrology.
It is not hard to imagine Bernal’s frustration and despondency at this turn of events. He viewed X-ray crystallography as fundamentally a branch of modern physics, but one that had enormous implications for chemistry and biology, as well as offering the promise of practical advances in the textile industry and metallurgy. Indeed he wrote a review article for an American medical journal at that time which concluded with the prediction that the chief utility of X-ray crystallography ‘will always be the fundamental assistance it gives to the process of knitting together all physical science into a harmonious whole’.51 Yet here he was at the epicentre of physical science in Great Britain, and he could not convince those in charge that this endeavour was worth supporting other than in the most dilatory and begrudging fashion. He did have at least one influential ally, Lawrence Bragg, who was moved to write the following letter52 from Manchester, dated 5th March 1931.
Dear Rutherford,
I have just had a letter from Bernal enclosing the report of the Coun
cil on Mineralogy, Petrology and Crystallography. He is very distressed at the subsidiary role assigned to crystallography and its prospective severance from Physics and Chemistry.
I cannot say or write anything publicly about the report, because if a Chair were founded and offered to him in the future I would be very glad to hear it. But I gathered that there was no immediate prospect of that for two years or so. Quite apart from that, as regards Bernal’s own position, is there no chance of giving him more independence? Universities over all attach the greatest importance to X-ray laboratories. They give them very fair facilities, and I have been very much impressed by the new lines which are being worked out in the ones I have seen. I do think the work has a tremendous future, and all the more so because it lies on the borderline between Physics and Chemistry, and so is often not directly backed up by either.
Bernal is extraordinarily good, though he talks rather much sometimes, and he keeps himself aware of the latest ideas in crystal work better than anyone else I know. He seems utterly discouraged by this report, and I think the best way I can help him is to appeal to you, and tell you how very good I think he is in his own line.
Yours very sincerely,
W.L. Bragg
Bragg’s letter had no immediate effect. The University did identify a bequest that they used to establish a chair in mineralogy and petrology, leaving the crystallography research group to fend for themselves in their four-room hut, which was not even weatherproof. The one advantage afforded to the crystallography group was their central position, only a few minutes’ walk from the Cavendish or the departments of zoology, chemistry and anatomy; as a new subject which needed to interlope onto the territories of older scientific disciplines, the topography could hardly have been better. Sage had acquired a colourful reputation that made him a much more prominent figure, in Cambridge and beyond, than if he had just been an extremely talented and productive scientist. There was really no range to his interests – everything put to him was likely to be considered as grist for the mill, and he was perpetually open for the cross-pollination of ideas. Other scientists faced with a new problem, especially involving a question of molecular structure, would naturally think of Sage and have no hesitation in asking him for help, even about a topic on which he had not previously worked.
So it was, over a matter of a few days in the spring of 1931, that Sage was approached about two apparently separate questions that would eventually turn out to be closely related. One enquiry came from J.B.S. Haldane and concerned vitamin D, the remedy for rickets. The 1928 Nobel Prize in Chemistry had been awarded to Adolf Windaus, a German professor, who had proposed structures for a family of compounds known as the sterols (cholesterol being the best known example); Windaus had also shown that vitamin D is produced by the action of ultraviolet light on ergosterol, a sterol found in yeast and many fungi. More recently, four different substances, each having antirachitic properties, had been crystallized. There was some argument amongst the chemists involved as to whether the different crystals really represented alternative forms of vitamin D or just resulted from one or more impurities. Haldane put Bernal in touch with the research group at the National Institute for Health in London, who were working on the problem, and they provided Bernal with some tiny, needle-like crystals of the different vitamin D candidates to study by X-ray.
The second challenge was also delivered to Bernal by an intermediary, Solly Zuckerman. Zuckerman, who was a year or two younger than Bernal, came to London as a medical student from his native South Africa, but after qualifying quickly disavowed clinical practice and embarked on a research career, building on his talents as an anatomist and anthropologist. He was an ambitious young man of limited means, who used his personal charm to enter many social circles, centred in both Bloomsbury and the Home Counties. One weekend he crossed paths with Bernal, who was probably the only active scientist he had ever encountered outside laboratory walls. The two struck up a friendship, and Zuckerman quickly came to respect Sage’s probing intellect. Indeed, he decided that Bernal was such an entertaining figure that he should be exposed to a wider audience. Zuckerman arranged for him to be invited to some country house weekends and London dinner parties. Bernal, who never let his political principles interfere with having a good time, was used to travelling light, and arrived for one weekend ‘with just the clothes he was wearing, carrying a briefcase which, in addition to papers contained only a razor and a slide rule’.53 The maid who unpacked his bag was flummoxed. Bernal did not disappoint the other guests at dinner, who included the novelist Evelyn Waugh, as he ranged over a number of topics with fluency. Waugh at the time was aspiring to join the Catholic Church, and he and Bernal held forth over the development of Christianity, while the port circulated. The argument went on until midnight, and Bernal, no doubt emboldened by the return of the ladies from the drawing room, seemed the more convincing of the two protagonists, to Zuckerman at least.
On graduating from medical school, Zuckerman had been appointed as the Prosector of London Zoo, where his main interest was the behaviour of apes and monkeys. He naturally spoke to Bernal about this, and also told him that a colleague at University College Hospital (UCH), Guy Marrian, had discovered what appeared to be a female sex hormone in the serum and urine of pregnant women. Marrian had obtained the substance, then known as oestrin, in the form of crystals measuring about 1 mm × 0.1 mm × 0.005 mm. Several laboratories other than UCH were attempting to define this new hormone, and Bernal told Zuckerman that he was sure he would be able to help if Marrian gave him some of the miniscule crystals to study. Confronted by the persuasive Sage, Marrian reluctantly parted with a few crystals, and Bernal carefully carried them back to Cambridge on the train.
If Marrian realized what a tall order it was to make any sense of the molecular structure of oestrin by X-ray crystallography, he might well have refused to hand over any crystals. The diffraction of X-rays is not a nuclear phenomenon, but is rather due to the interaction of the waves of radiation with the negatively charged cloud of electrons around each atom in the crystal. As Bernal himself pointed out in 1930: ‘It has, up until now, proved impossible to determine atomic positions [in a crystal] by direct calculation from the intensities of diffracted beams of X-rays. The inverse process is, however, straightforward: from a known distribution of atoms it is possible to calculate exactly these intensities, and the problem reduces itself to finding by any means whatever a plausible structure and then establishing it accurately by means of numerical correspondences between observed and calculated intensities.’54
The crystallographer typically started with some idea of molecular structure from the known chemical formula of the substance. In the case of relatively simple compounds such as the amino acids, there were few arrangements of the atoms possible, and by looking at and mathematically analysing the diffraction pattern, often one proposed structure would emerge as the most likely. Even with amino acids, the diffraction spots were not that sharply defined because of the limited intensity of the X-rays available, and their interpretation was further confounded by the phase problem – there was no way of telling whether a spot on the film represented a maximum (crest) or minimum (trough).
A large organic molecule like oestrin would seem to be a most unpromising problem because there were no good pre-existing ideas about its structure, and its diffraction pattern would be indistinct and ambiguous. It was on 10th June that Zuckerman55 had written to Bernal suggesting that Marrian might provide some material for X-ray crystallography, and despite all the inherent difficulties it appears that Bernal arrived at an answer before the end of the month. He wrote to Astbury: ‘I have done a hormone, oestrine, and a compound related to uric acid, but I am having awful difficulties with the people here and may resign at any moment. They have completely let me down over giving me apparatus.’56 Astbury advised him not to let ‘those villains at Cambridge beat you down’.57 Working under far from ideal conditions and with his usual number of competing distractions, S
age was able to deduce the size of the oestrin molecule, and ventured that its structure involved a number of condensed rings with small atomic groups (O and OH) attached at either end of the molecule.
That same month, he began to analyse the crystals of vitamin D and other sterol compounds supplied to him by the group at the National Institute for Medical Research. He subsequently wrote a letter to Nature58 summarizing his preliminary results. The five compounds included cholesterol and ergosterol as well as the antirachitic substance, calciferol, that had been prepared by the action of ultraviolet light on ergosterol. Bernal’s main conclusions were that calciferol comprised a single molecular entity that was virtually identical in size and shape to the parent ergosterol molecule. But after considering the X-ray data, he suggested that the dimensions that he had measured ‘are difficult to reconcile with the usually accepted sterol formula’. In other words, Bernal was politely disputing Windaus’ model and his intervention forced a radical change of thinking amongst sterol chemists. Rosenheim, who was the head of the group at the National Institute for Medical Research, wrote to Bernal in April 1932 to let him know that he had been led ‘to a modified constitutional formula for sterols and bile acids, which seems in better harmony with your results than the old formula’.59 Rosenheim kept Bernal supplied with a steady stream of crystallized sterol derivatives to study, as did Windaus. Thus both the British and German schools relied on Bernal’s laboratory, while they were involved in a sharp public debate about the structure of cholesterol. He was happy to provide data to both and was not tempted to take sides.