by Andrew Brown
Within six months of receiving her RF grant, Wrinch had a second paper in Nature45 – one which sought to make a fundamental change in the way scientists thought about protein structure. What she suggested was an extension of an idea of Astbury’s about how molecules of α-keratin were folded in unstretched wool. Astbury had put forward a model of α-keratin that included regular hexagonal folds, which the Oxford chemist, F.C. Frank, had validated by suggesting a novel chemical bond between amino acids in the polypeptide chain or protein molecule. This transformed peptide bond had been brought to Wrinch’s attention, as she acknowledged in her Nature paper, by Bernal. She now applied the transformation to all manner of globular proteins, suggesting that they existed in regular hexagonal arrays, rather like those of a honeycomb. Instead of globular proteins comprising long poly-peptide chains, folded in some arbitrary pattern, portions of them consisted of closed hexagonal ‘cyclols’. She considered the cyclols as essentially two-dimensional structures that were one amino acid thick, and then explained how these layers might be built up into three-dimensional aggregates. Very pleasing to her as a mathematician was the existence of geometrical series of cyclols, in which different orders of symmetry were associated with predictable numbers of amino acid residues.
While she put the cyclol theory forward as a ‘working hypothesis’, Wrinch was convinced that such an elegant, logical theory explaining the known facts about native, globular proteins must be true. Astbury, always enthusiastic about ideas that echoed his own, saw the cyclol hypothesis as a way of treating the soluble globular proteins and the fibrous proteins under a unifying theory: ‘Wrinch has attempted to build up a consistent molecular system for the globular proteins by folding back, theoretically, these liberated poly-peptide chains into series of folds that are none other than those first postulated to explain the X-ray diagrams and properties of keratin.’46
Wrinch undertook a vigorous campaign to persuade the world that her cyclol hypothesis was correct, both by writing papers to journals and letters to leading scientists. She managed to convert one very distinguished American chemist, Irving Langmuir, who had won the 1932 Nobel Prize for his many investigations of surface chemistry. Indeed it was probably the prospect of protein films, one molecule thick, which attracted Langmuir to Wrinch’s theory. As a mathematician, she had no laboratory expertise and no way of testing her own hypothesis, but was entirely dependent on the data of others. In 1937, her theory received a tremendous boost when two chemists published their analysis of egg albumin suggesting that the molecule contained 288 amino acid residues – 288 being a number that was found in one of her geometric series. Such was the attention her ideas were receiving that Sir William Bragg decided that they should be aired at a Royal Society meeting. Before the meeting he wrote to Bernal asking his opinion and received the following response: ‘The present position of our knowledge of the structure of the protein molecule is one where our only method of advancing is by guesswork. If a structure is proposed it can be tested, but the difficulty is to get the leading ideas. It may be, of course, that these will be arrived at by orthodox analytical methods, but in the meantime I think speculation is legitimate.’47 Bernal thought that Dr Wrinch was proceeding along the right lines.
The Royal Society meeting did not take place until November 1938, by which time there had been a polarization of views between Wrinch and Langmuir, still vigorously supporting the cyclol hypothesis, and the rest of the biochemical world. Wrinch immersed herself in the mathematical techniques of Patterson analysis, and lacking her own crystal data, had reworked Dorothy Crowfoot’s published results on insulin to show that they were compatible with a cyclol model for the molecule. This had understandably infuriated Crowfoot, who did not believe that her data supported any particular theoretical model and certainly not the cyclol hypothesis; she turned to Sage for support. His original support for Wrinch’s ideas had largely dissipated by this stage as more discordant experimental findings were made, and perhaps because he knew that Linus Pauling had launched a devastating attack on the cyclol theory. Bernal wrote to Wrinch in the weeks before the meeting, trying to explain why her insulin model was invalid. Wrinch was impervious to argument, and the correspondence finished just four days before the meeting in London with Bernal saying he did not think it would be profitable for him to list all his reservations about her method of Patterson analysis. He did say that one objection but not the main one was ‘that you don’t take nearly enough account of the physical character of the molecules’.48
At the meeting, Wrinch gave a spirited presentation of her ‘geometrical attack on the problem of protein structure’, claiming that the cyclol hypothesis explained a number of disparate facts about proteins. In a nod to her ally, Langmuir, she predicted that ‘surface methods will enable us to investigate the nature of the bonds maintaining the specific structure of proteins’.49 Neither Crowfoot, who was heavily pregnant and giving her first lecture at the Royal Society, nor Sage wanted to challenge Wrinch directly at such an auspicious venue; neither of them even mentioned the cyclol hypothesis, while both were careful to stress the preliminary nature of all protein crystallography. Bernal’s forbearance was too much for Bill Pirie, who wrote to complain: ‘The ignorant people from this laboratory [William Dunn Institute, Cambridge] who went [to the meeting] came away with the impression that everyone, including you, now agreed with her. I wish you would develop the art of being rude.’50 Nor did Bernal’s restraint earn any thanks from Wrinch, who now viewed him as ‘jealous, brutal and treacherous’.51
The continued, energetic promotion of an increasingly suspect theory was becoming damaging to the nascent subject of protein crystallography. Bernal and Lawrence Bragg decided that they needed to make a concerted effort to discredit the cyclol hypothesis. Bernal told Bragg that he had had a long talk with Langmuir, in which he did not succeed in ‘getting across the crystallographic point of view’.52 At the same time, Bernal wrote to Langmuir, sending him a copy of a letter which he was proposing to send to Nature, saying ‘I very much regret having to do so because it must produce open controversy on a subject which I should prefer settled in private conversations.’53 For his part, Bragg promised ‘to thresh [sic] the whole thing out’54 with Langmuir when he visited Manchester. Bernal objected both to the tone of the Wrinch school, which was far too certain, and to their misuse of the Patterson vector diagrams which ‘needs to be pointed out because the ultimate effect would be very bad for crystallographers’ credibility among chemists when the fallacies eventually came out’.55
In January 1939, Lawrence Bragg wrote a long, patient letter to Nature, explaining why Langmuir and Wrinch’s theory was unsubstantiated and less than it seemed to its propagators. His letter was followed in the same issue by Bernal’s, which while not rude, was certainly blunt by his standards. Bernal characterized the additional assumptions made by Wrinch in her application of Patterson maps as being of ‘extremely dubious validity’, he repeated the objections of the chemists to the idea of amino acid residues being linked to four others (rather than to two others by peptide bonds), and demolished any notion that the ‘vector maps of Dr Wrinch’s hypothetical cyclol structure bear [any] resemblance to those which have been derived by Miss [sic] Crowfoot from her observations’. In rejecting the cyclol hypothesis, which failed completely to account for the X-ray scattering results for insulin, Bernal thought ‘it is not necessary, however desirable, to put forward another model – the problem remains an open one’.56
Within weeks of writing those words, Bernal was putting forward another model, which would eventually prove invaluable. At Sir William Bragg’s invitation, he gave a Friday evening discourse at the Royal Institution on the structure of proteins. After a very complete, polished, review of the subject, referring to recent findings by Astbury, Crowfoot and Perutz, Bernal suggested ‘working hypotheses’ of his own as a guide to future work. One idea he put forward was that large protein molecules are not unitary but consist of subunits: his objecti
on to a single long chain molecule was that it was difficult ‘to imagine any kind of fold or coil by which a single chain can occupy the observable space and at the same time not be so intricate that its formation by any natural process would be enormously improbable’.57 By invoking subunits, ‘the improbabilities of the coiling become much less’, and while the sub-units themselves might be non-symmetrical, they might be arranged together in a symmetrical way, explaining why ‘the symmetry of protein crystals is much higher than would be expected statistically from compounds of such great complexity’. There was, he thought, good experimental evidence for such subunits, and the next question was how they were linked together.
Whatever the means of attraction between the subunits, it had to hold the protein molecule together in water. The mechanism that most appealed to Sage was one suggested, in a different guise, by none other than Langmuir at a lecture in London the previous month. It is a reflection of Bernal’s open-mindedness and acute perspicacity that he would still pay attention to a man whose theory he was actively discrediting and be able to hear one inspirational note against a background of noise. Langmuir’s original insight was that in aqueous solution, the hydrophobic side chains of amino acid residues would become buried on the inside of the folded protein molecule away from the surrounding water – a profound variant on the old adage that oil and water do not mix. Although Langmuir presented this as part of a rearguard defence of the cyclol theory, Bernal realized that the hydrophobic interaction between water and protein molecules was an idea that could stand alone. After crediting Langmuir, Bernal stated: ‘In this way a force of association is provided which is not so much that of attraction between the hydrophobe groups, which is always weak, but that of repulsion of the groups out of the water medium.’58 The overwhelming influence of the hydrophobic forces in determining the way proteins fold into their native configurations was not generally appreciated by protein scientists until the mid-1950s. Even Perutz, who was struggling with the tortuous folding of haemoglobin, was unaware of it until then.59 The clinching piece of evidence for the mechanism came from thermodynamic considerations of the hydrogen bonds in water – a subject that Bernal had initiated with Fowler in 1933.
In The Search, when Constantine is proposed for Fellowship of the Royal Society, he is ‘put up’ not by his own professor but one from Manchester. About a year after the novel appeared, the real-life Sage received a letter from a professor in Manchester, who had ‘been thinking for sometime that you should be put up for the Royal’.60 Bragg told Bernal that he had sounded one or two people out and he thought there would be good support. This might have spared Bernal the anxiety experienced by his doppelgänger, Constantine, who took the pessimistic view that ‘the physicists won’t like me because I’m a renegade from physics… and the chemists won’t like me because I’m a physicist. And the biologists won’t like me because I do biology. And the mathematicians won’t like me because I don’t do mathematics.’61 Sage was elected FRS in March 1937, at the remarkably young age of 35 years. His ‘proud and loving mother’ wrote from Florence62 to congratulate him; she and her sister, Cuddie, were recapitulating the grand tour of their youth, having handed over the running of Brookwatson to Desmond’s younger brother, Kevin.
Desmond was now living, most of the time, with Margaret and the newborn Martin in her house in Hampstead. Margaret was still secretary to FIL and was spearheading their efforts to rouse public opinion on the dangers posed by the growth of fascism in Europe. She was finding it a struggle to keep the momentum of the organization going. Aldous Huxley resigned as President at the end of 1936, followed shortly afterwards by Leonard Woolf, the Vice-President, who succumbed to his wife Virginia’s growing irritation with ‘dirty, unkempt, ardent, ugly, entirely unpractical but no doubt well-meaning philanthropists at whom I should throw the coal scuttle after ten minutes if I were in his place’.63 When Bessie Bernal returned from her European trip, she asked to visit Hampstead so that she could see Martin, but was embarrassed about meeting Margaret.64 Margaret took no offence, and the house was left in charge of Agnes, her pretty brunette, cook general; Martin was presented by Ully (his youthful, German, communist nanny). Bessie came away impressed by the cleanliness of the house, but no doubt bewildered by Desmond’s living arrangements. It is said in the Bernal family that she devoted herself to raising gentlemen and scholars, but could never get the two to coincide!
Life with Margaret expanded Sage’s already wide social horizons beyond scientists and radical intellectuals and brought him into contact with many of her artist friends. W.H. Auden would come to stay in Hampstead and talk to Margaret for hours as they smoked their way through a packet of cigarettes. Sage was a non-smoker and suspected it was an unhealthy habit. On occasion, Sage would express surprise at Margaret’s comparative ignorance about a topic. She once asked him, ‘What do you expect me to know?’ He replied: ‘Well, more or less, what happened everywhere at any time.’ And Margaret retorted, ‘Do you know that?’ Sage thought for a minute or two and said: ‘I know nothing at all about the fourth century in Romania.’
The friend of Margaret’s whom Bernal really latched onto was Barbara Hepworth, whose abstract sculptures fascinated him. In October 1937, Hep-worth held a show of her work at a Mayfair gallery and asked Bernal to write the preface to the catalogue. Her work reminded him very strongly of Neolithic stone monuments – just as those monuments employed extremely limited symbolic forms that could be seen as reaction to the living representations in the art of the Cave painters, so Hepworth had reduced her sculptures to the barest elements as a reaction to the representational art of the nineteenth century. Her use of bare, unadorned shapes made it possible ‘to see the geometry which underlies it and which is so obscure in more elaborate work’. To Bernal’s eye, ‘the elements used are extremely simple: the sphere and the ellipsoid, the hollow cylinder and the hollow hemisphere. All the effects are gained either by slightly modulating these forms without breaking their continuity, or by compositions combining two or three of them in different significant ways.’65 Barbara Hepworth enjoyed Bernal’s analytical approach to her work: ‘When you criticised my carving, you were quite right, sculpturally, spatially and constructively… you always seem to me to be searching for, discovering and applying basic laws and principles.’66 Hepworth regarded Bernal as ‘the most inspiring person ever to come into my workshop as he had the amazing capacity for comprehending in an instant the nature (and even the formula) of every sculpture, and hours were spent in exciting discussion and drawing’.67
Hepworth encouraged Bernal to write a chapter on ‘Art and the scientist’ in a book, Circle, that was being compiled by her husband, Ben Nicolson, and a number of other artists in the Constructivist movement. This gave Bernal the chance to lament the separation of art and science – ‘the last official link was the annual reviews of the Royal Academy which used to be given in Nature, and in which the academicians were chided for putting the moon the wrong way up in the sky or for painting a flower with too many petals… In the great creative periods of science the artists and the scientists worked very closely together and were in many cases the same people… Leonardo da Vinci, though the greatest, was only typical of whole schools of artist-scientists. Gradually, however, with the development of bourgeois culture the useful and the ornamental were piously separated. Science was used to make the money, art simply as a means of spending it. The result of this separation has been the most incredible mutual ignorance.’68
As a crystallographer, Bernal constantly thought in three dimensions and his remarkable facility in doing so may explain why sculpture and architecture were the two arts that appealed to him the most. He gave a talk to the Royal Institute of British Architects in 1937, which included a wonderful exposition of symmetry:
‘The only aspects of symmetry that are formally considered in architecture are those of mirror symmetry in elevation, and, to a minor extent, radial symmetry in plan, but there are far more symmetries than th
ese. Any type of repeatable operation, whether it is a reflection, a turning or merely a translation in space, gives rise to a symmetrical structure … An equal-spaced arcade, for instance, is a particularly simple example of translation symmetry.’ He explained that there were only a limited number of possible symmetrical modes: ‘In surface repetitions, for instance, such as those for pavements or walls, there are actually only seventeen different rhythms, all of which have been used unconsciously in art, but many more of the subtle ones only in the textile work of primitive tribes. In three dimensions the complexities are naturally greater. Here there are no fewer than 230 modes, most of which have certainly not been used up till now in architecture, but which might be made to produce new and significant effects… the architect is no longer tied to the massive piling of rectangular blocks and can place his elements almost where he likes in three dimensions.’