Home > About the Academy > Biographical memoirs

BIOGRAPHICAL MEMOIRS

John Stuart Anderson 1908-1990

Early life: London, 1908-1938 Melbourne (1) 1938-1946 Harwell 1947-1954 Melbourne (2) 1954-1959 The National Chemical Laboratory 1959-1963 Oxford 1963-1975 Aberystwyth 1975-1982 Canberra 1982-1989 The man and the scientist Appointments Honours Notes

By B.G. Hyde and P. Day

This memoir was originally published in Historical Records of Australian Science, vol.9, no.2, 1992. It also appeared in Biographical Memoirs of Fellows of the Royal Society of London, 1992.

Numbers in brackets refer to the notes at the end of the text.

Early life: London, 1908-1938

John Stuart Anderson ('JS' to most colleagues, and in this memoir) was born on 9 January 1908 at 134 Englefield Rd., Islington, London (U.K.) and died on Christmas Day, 25 December 1990, at the Woden Valley Hospital, Canberra (Australia). He was the youngest child and only son of John Anderson (1850-1916), a master cabinet maker who was born and brought up in Aberdeen, and Emma Sarah, née Pitt (born Ipswich, died 1937). Knowledge of his ancestry is limited and contains no suggestion of science, his antecedents being in business or teaching or, in the case of his maternal grandfather, a sailor. His parents, both widowed, married about 1901-2. Being of his father's previous marriage, two step-sisters were much older than he, by about twenty years; his sister Olive was about four years older.

His father was a good violinist and also played the 'cello; he enjoyed musical evenings with his daughters and friends, playing instrumental quartets and sometimes more ambitious efforts such as, on one occasion as JS recalled, 'Handel's Toy Symphony' (sic) (1). The house was well stocked with books especially the main classics of the Victorian period; but family circumstances, at first 'modestly prosperous', declined in the years just before and during the First World War, catastrophically so in 1916, when his father died. Just before that, his mother went to work in a munitions factory, his sister left school early and they suffered upwards of ten years of great difficulty and acute poverty. He believed that these years which, due to his mother's 'overdone ideals about gentility', were solitary and self-contained ones, left on him a permanent mark a lack of self-confidence, mainly manifested as shyness and an inability to get on with people. He was certainly a private man but, at least after he passed the age of 30, his younger colleagues did not detect this supposed lack of self-confidence.

His first schools were old-fashioned dame's schools but, at the age of 11, after two years at the 'Hugh Myddleton' L.C.C. school (Clerkenwell, 1917-1919), he obtained an L.C.C. Junior County Scholarship, which took him to the boys' school attached to the Northern Polytechnic (Highbury County School, 1919-1924). School did not impress him but the Islington Public Library ('exceptionally well stocked') did and, as school work came easily he had time for omnivorous reading including all the chemistry books he could lay his hands on. After passing the Schools Certificate examination in 1924, with Distinctions or Credits in every subject, he left school with a strong determination to study chemistry.

With an Intermediate County Scholarship he enrolled in the day BSc course at the Northern Polytechnic (1924-1926), but apart from the advantage of there being very few Science students he was not impressed with this establishment either. During his second year he attended third-year evening classes in physical chemistry at the Imperial College, and decided to enter for a Royal Scholarship. He sat for this during the 1926 General Strike, passed at the top of the list, and entered the Royal College of Science (RCS) in October. In 1928 he topped the First Class Honours list of BSc's and received the Frank Hutton Prize in Advanced Chemistry.

The crucial decision then was what to do for his third, research year at RCS – necessary before being awarded one's degree. Strongly influenced by his organic chemistry lecturers (Sir Jocelyn Thorpe FRS and E.H. Farmer [later FRS]), Anderson warmed to the logic of degradative and synthetic organic chemistry: Farmer tried to recruit him, and Linstead (later, Sir Patrick FRS, then 'a lecturer in his first year, with a pink-cheeked nervousness that was hard to recall in his later years') also – but he decided to turn elsewhere. He thought that physical chemistry (with J.C. Philip FRS and H.J.T. Ellingham) and inorganic chemistry (with Riley and Harwood) were both uninspiring, but records that 'some perverse streak made me decide that inorganic chemistry must present areas of neglected opportunity, and I decided to enter Professor H.B. Baker's laboratory'. Later he added, 'I took up inorganic chemistry out of cussedness, really'.

In spite of its disadvantages, Baker's laboratory appeared to attract JS because it was one of the few centres of inorganic chemistry research in England and recruited keen young people, including foreigners such as N. Grace from Saskatoon and Swart from Amsterdam. Among other things (e.g. his fixation on the 'Intensive Drying Controversy'), JS criticised Baker for not directing his research students(2): but by mutual self-help he and his colleagues learned a great deal, and established important personal relationships. He has recalled that with Cheeseman (subsequently at the Universities of Reading and Tasmania) he learned the difficult art of blowing soda-glass (this being well before the days of pyrex) and that they made their own vacuum systems – diffusion pumps, Bourdon gauges, ground-glass joints, the lot! Cheeseman also made red, liquid Cl0₃ before Schumacher et al. published their work but, because it detonated when he turned a (greased) tap – 'leaving the tap handle in his fingers and a little glass dust' – he and JS embarked instead on a study of the sulphur-fluorine system, reacting sulphur with lead fluoride. They failed, however, to fractionate and characterize the products.

H.J. Emeleus and R. Purcell were senior members of the laboratory. Emeleus, later FRS and Professor at Cambridge, was an 1851 Exhibition Senior Student recently returned from Stock's laboratory in Karlsruhe, and about to take up a Commonwealth Fellowship at Princeton: in his work on explosion limits he used Stock's type of grease-free vacuum line and vacuum-fractionation technique. His interest in chemical kinetics and chain reactions involved theory, which was apparently anathema to Baker (and the RCS). Purcell, later Head of the Royal Navy Scientific Service, was 'a fine experimenter and skilled glassblower', just returned from Smits in Amsterdam and then re-investigating, more effectively, Baker's research on the 'chemical' effects of intensive drying. Tea-time discussions between the young researchers provided the guidance that Baker did not, and led JS to start work on nickel carbonyl: purifying it and studying its physical properties and chemical reactions especially with NO. Baker was antipathetic, and JS had to devise the micro- and semimicro-analytical techniques for characterizingthe air-sensitive solid products.

At Imperial College there was, at this time, a firm tradition that to have a prospect of academic life it was essential to get experience elsewhere, by working on the Continent. His own work on nickel carbonyl compounds had led JS to the contemporary German work of Hieber et al. on the iron carbonyls and so, in 1931, he applied for a University of London Travelling Scholarship to work for a year with Hieber at Heidelberg. With this award he became Hieber's first foreign co-worker.

Before going to Heidelberg, JS carried out a seminal experiment on the structure of the metal carbonyls. Hieber's work had suggested that a cyclic structure was improbable, but it was still disputed whether the CO groups were CO molecules or covalently bound to the metal atom as C=O. Discovering (in 1930 or 1931) that a Raman spectrometer had been set up in the Physics Department of the RCS, JS borrowed time on it to see whether the Raman shift corresponded to CºO or C=O: only the former was observed – close to but significantly different from that for CO gas. Thus, as he put it, 'I think that I must have been one of the first people to use Raman spectroscopy to settle a clearly defined structural problem'.

After obtaining his PhD in 1931, he spent two semesters in Heidelberg, working in the Curtius Saal behind Bunsen's old Chemisches Institut in the Akademiestrasse. Only one of Hieber's group spoke English, and JS became proficient in German. He also learned to use oxygen-free atmospheres for experiments and, in turn, was able to introduce vacuum-line methods, glass Bourdon gauges and Sidgwick's ideas about the electronic theory of valence into the Heidelberg laboratory. Hieber was at first inclined to regard the valance ideas as dangerously speculative and refused to allow his name on a paper, the discussion section of which was based on them. But, according to JS, he soon adopted them and was probably the first German inorganic chemist of note to approach the problems of bonding in coordination compounds from a modern viewpoint. JS's work at Heidelberg involved the isolation of Fe(NO)₂(CO)₂ which, with the already known Co(NO)(CO)₃ and Ni(CO)₄ showed the existence of an isoelectronic series of tetrahedral molecules. The preparation of substituted derivatives of the first two showed their chemical relation to the pure carbonyls. After a canoe trip down the Danube to Budapest, and a stroll through the Balkans and Greece, JS returned to the RCS in 1932 as a Demonstrator.

The next few years were crowded ones. While some carbonyl work continued, his interests broadened to the more general question of the role of unsaturated molecules as ligands in Pt and Pd complexes but, according to his later assessment, the absence of suitable instrumental techniques and aprotic solvents was a hindrance. He started to think about means to study organic molecular compounds, the two obvious approaches being to examine association equilibria in solution or – 'more incisive, but also [with] extreme difficulties' – to determine crystal structures. Characteristically, he chose the second. As he pointed out, at that time the crystallography of organic compounds was little developed (in contrast with the situation for inorganic materials/minerals), and there was no British school with which to collaborate. The only possibility was to become a crystallographer oneself, and this indeed was the firm advice that Cheeseman had received from Kathleen Lonsdale when he tried to get her help. Knowing nothing of the technique, JS obtained two small grants from the Royal Society: with the larger (£50) he had a single-crystal goniometer built by an instrument maker, while the other provided a 30 kV transformer. He recalled, 'I built a Shearer self-rectifying tube and set the whole thing up in the open laboratory' (He realised the hazards much later!) 'It was then necessary to learn the rudiments of X-ray crystallography, self taught, as I went along' But he was never a professional crystallographer and inevitably, even though the equipment later accompanied him to Melbourne, this work, on phenoquinones and related compounds, was not very productive.

He was also peripherally involved with hydrogen isotope studies, which yielded his first research student, F. W. James. (Until then JS had worked singlehanded.) Together they studied the exchange of deuterium between solvent water and metal ammines including 'inner sphere' Werner complexes like Co(NH3)|+. Then, with two PhD students, he studied the kinetics of isotope exchange of Co(ND₃)³₆⁺, Co(D₄-en)³₃⁺, Pt(ND₃)²₄⁺, Pd(ND₃)²₄⁺ etc. with light water. The results showed two routes – ligand exchange of ammonia and water, and acid dissociation of ligand NH3 – and provided a unified explanation of some of the chemical properties of such complexes of all transition metals. His collaboration with L.O. Brockway was a fairly early determination of molecular structures by gas electron diffraction. (The first was seven years earlier, in 1930.) It preceded by thirty years JS's use of electron diffraction to study the solid state. Another sideline, pursued in collaboration with Linstead and A. H. Cook (later FRS), was the absorption spectra of dissolved phthalocyanines. These solutions also fluoresced and Sir Robert Robertson, the Government Chemist, had a suitable spectrograph and was already engaged in fluorescence spectroscopy. JS's request to use that instrument was facilitated by his ability to provide first-hand local knowledge of the Eastern Carpathians – where Robertson intended to go trout fishing – which he had gained on a recent walking tour there.

In 1935, at Enfield, Middlesex, he married Joan Habershon Taylor. Their children are Margaret Jean (born 1936), Elizabeth (1938), Ursula Ruth (1939) and Malcolm Robert Hugh (1941). All his family survive him. The birth of their first child necessitated his earning more than a demonstrator's pay of £260 p.a. or even an assistant lecturer's (£300). He therefore worked for British Chemical Abstracts in the evenings, his fluency in German standing him in good stead.

Shortly after, Emeleus returned from Princeton to a lectureship at the RCS and, while they were both demonstrating to first-year practical classes, he and JS discussed how inorganic chemistry should be taught. Dissatisfied with the standard texts such as 'Partington', which gave the chemical facts but made no attempt to fit them in a unifying and predictive way into a theoretical framework such as that provided by the existing, rapidly developing concepts of atomic structure and valency, they decided that these should be discarded and that a lot of material from the current literature – which Emeleus says they 'used to read avidly' – should be brought together instead. Emeleus goes on, 'This went down well and J.C. Philip, who was the editor of a series published by Routledge, suggested that we do the book'. Thus, they planned Modern Aspects of Inorganic Chemistry, dividing the contents between them according to their separate interests and, in the first instance, basing it on the lectures that both were giving in third-year courses. From start to galley proofs – including much reading, and writing on the train to and from Chemical Society Thursday meetings – took less than two years, and it was published in 1938. Invariably known as 'Emeleus and Anderson', it was a landmark text that went through many printings and several editions and translations. JS sometimes thought that it was his most important contribution to chemistry. On its runaway success, his characteristically dry comment was, 'Evidently it met a real need'. He noted that when Glemser was called into the German army on the outbreak of the Second World War, he took it with him to the front; and that a German translation was published by Springer in 1940! Later, he pointed out that the 1938 preface laid emphasis on the importance of solid state chemistry, a subject which hardly existed then in the English-speaking world but which, of course, soon came to dominate his own career. Parenthetically, royalties from the book enabled him to buy his first car.

Nevertheless, a fixed establishment of Lecturers at the RCS made JS believe that his future there was blocked and so, in 1937 he applied for and obtained a Senior Lectureship at the University of Melbourne, replacing Noel Bayliss (later, Sir Noel, FAA) who had taken the Chair of Chemistry at the University of Western Australia.

Melbourne (1) 1938-1946

The Andersons arrived in Melbourne in 1938 and JS immediately had to take responsibility for teaching first-year medical and dental students, lecturing to them on physical, inorganic and organic chemistry, and organizing and supervising practical classes for 170 people. A story is told about this time which, in the light of his heavy teaching duties and his character, is not surprising. A group of medical students went to his room to welcome him to Melbourne – his response was a terse, 'Well gentlemen, I'm sure you are as busy as I am. Good morning.' Thus was the foundation laid for his not-entirely-deserved reputation of being somewhat unapproachable.

The Head of the Chemistry School, Professor E.J. Hartung, he described as 'a paternalistic autocrat...of impressive intellectual power and wide interests', 'really an 18th-century enlightenment figure, who had strayed by mistake into the 20th century'. Because of administrative and outside commitments Hartung, at this time, did little research himself, although he had earlier done some very good work. However, within the constraints of a slender budget, he strongly supported those three of his staff who were thus active: JS (inorganic), Associate Professor 'Bill' Davies (organic) and Dr Erich Heymann (physical) Hartung quickly commanded, and always retained, JS's respect. In those days Australian universities had no PhD courses, but the final year of BSc and an additional year both involved course work and research, leading to the MSc degree. Later, JS noted that the standard of MSc theses was not far short of the typical PhD level, and that 'the steady trickle of research workers included a high proportion of graduates of great ability, who achieved very distinguished careers'.

He shared his office/laboratory (which also contained his open X-ray set!) with A.L.G. Rees, later Chief of the CSIRO Division of Chemical Physics, FAA, and sometime President of IUPAC. They became friends for life. Similarly firm friendships developed with Walter Boas, later Chief of the CSIRO Division of Tribophysics and FAA, a metal physicist who had fled Europe in 1938; E.S. Hills, later FAA, FRS and Professor of Geology at Melbourne; the mineralogist Austin Edwards; and the geophysicist K.G. Bullen (later FAA, FRS). From their several disciplines this group came together during the war years for seminars and mutual education.

Again feeling handicapped by insufficient spectrographic equipment and also by his lack of crystallographic expertise, JS cast around for alternative research activities, corresponding with David Mellor in Sydney on coordination compounds and, through Hartung, approaching Sir David Rivett ('always approachable and kind to younger men') about fields of work appropriate to the Australian environment. Consequently, 'I deliberately changed my direction of work in Melbourne...It seemed a good idea to find a field that fitted better into the local scene' He started to think about the chemistry of Australian mineral resources, and embarked on two lines of research. As a schoolboy and young undergraduate he had been excited by the discovery of hafnium in 1922 and rhenium in 1925, and now he decided to look for element 43, technetium, which had just (1937) been produced artificially (hence its name). In 1925 Noddack – who called it 'masurium' – had claimed to have found it in niobite, but his claim was never verified and that name was dropped. (It is now known that element 43 has no stable isotope.) JS's reading of Noddack's and various geochemical papers suggested that both technetium and rhenium might occur in trace amounts in sulphide ores, but that significant quantities could only be retrieved from large-scale processing operations such as those for lead, copper and zinc (at Broken Hill, Mount Lyell, etc.). However, his examination of residues and products from all stages in the working-up operations yielded nothing but a great deal of experience in emission spectroscopy, which later proved useful in tracing the course of indium in the refining of cassiterite and in recovering Pa from uraninite refining.

On the other hand, work on separating zirconium and hafnium was more profitable. With Ivan Newnham (an MSc student in 1939; later Chief of the CSIRO Division of Mineral Chemistry and, later still, Director of the CSIRO Institute of Minerals and Energy Research), and using kilogram amounts of zircon and monazite concentrates from Byron Bay beach sands, the lower halides of Zr and Hf were prepared and studied. The idea was that, by analogy with the Nb/Ta pair, ZrX₄ would be more readily reducible than HfX₄. Progress was made and, later, US zirconium producers tried to recruit Newnham to develop the process. However, the advent of ion-exchange methods made it redundant.

During the summer vacation, 1939/40, the Department moved from the Old Physiology Building to the new Chemistry Building. That and the Second World War interrupted crystal structure work and consolidated the change in JS's research interests. His leaning towards the chemistry of the solid state increased, and he became especially interested in the constitution of non-stoichiometric compounds and solid solutions, for example in how elements of different valence state were accommodated in minerals. This was a logical development of his geochemical interests, but it had been initiated by two earlier events, namely a lecture by Bernal in about 1936 at a Chemical Society symposium on developments in structural chemistry in which he discussed 'berthollides' like 'FeO' and 'FeS' that were of variable composition, and V.M. Goldschmidt's Hugo Muller Lecture. These, and his wide reading for 'Emeleus and Anderson', led him to think about the compensated and uncompensated replacement of one element by another in non-molecular crystals – a theme that underlay most of his research for the rest of his working life. He spent the early 1940s in extensive reading about the chemistry, reactivity and thermodynamics of solids – all the papers of Huttig, Fricke, Biltz, Tammann and so on. More than 100 from the Biltz school comprised a mass of sound, systematic information on the subject.

A major turning point in his career occurred when he lighted upon the classical (1931) paper by Schottky and Wagner on the statistical thermodynamics of ordered mixed phases. Appreciating the implications of point defect theory, his own ideas fell into a coherent pattern. He later wrote, 'for some reason, that seminal paper had not registered in the consciousness of chemists as a whole, and Wagner's papers of the 1930s, exploiting the concepts, were little known'. (Indeed they are not referred to in the first, 1938 edition of 'Emeleus and Anderson'.) In his biographical notes, JS continues:

In terms of the statistical thermodynamic, point defect model, absolute invariability of composition could not be true for any solid compound. On the other hand, if realistic values for the energetic quantities were used in estimating intrinsic defect concentrations, the inherent defectiveness and ease of compositional variation would be small, for typical crystalline compounds. To me the big problem was why some, and only some, crystalline compounds had a broad composition range, and what delimited that range? Later(3), the third aspect of my concern was how, in structural terms, really large variations in composition were accommodated in structures which, in many cases, were formally of simple types. Lacher had developed, for the palladium-hydrogen system, the statistical thermodynamics of defect structures in which there was an exothermic, pairwise interaction between defects on adjacent sites. I carried this a bit further and generalised it to define the accessible, stable composition range of a binary phase, insofar as it was controlled by the internal, intrinsic equilibria; it treated the nonstoichiometric crystal as a closed thermodynamic system, rather than considerations of the composition dependence of thermodynamic potentials, in an open system with a succession of phases. The defect interaction model was compared with such equilibrium data as were available for nonstoichiometric systems: these were very limited in amount and sketchy in quality, being mostly confined to the work produced by Biltz and his school in the 1930s. My paper does seem to have stimulated others, particularly when interest in stoichiometric variability grew stronger, around 1955-65.

Indeed, the theory JS developed in this paper determined not only most of his own research during the next twenty years, but also that of many other people in the field. (But, as we shall discuss later, some quite different ideas also appeared in the decade to which he referred, viz. 1955-65.)

This classical paper did not, of course, come out of thin air. For the previous few years, with a series of postgraduate students, he had been carrying out experimental work on relevant inorganic systems. M.J. Ridge studied Sn+SnS, melting the components together, equilibrating, and analysing the products by classical wet methods. But defect concentrations were much too small to be detected by such analyses; attention was therefore turned to measuring electronic properties. Merial Morton (née Clark) varied the sulphur content of PbS and of SnS by evaporating out sulphur in vacuum, and then observed the effect on the resulting n-type sulphides of chemisorbed oxidants, using 4-probe electrical conductivity and thermal e.m.f. measurements. N.N. Greenwood, later Professor at Newcastle-on-Tyne and Leeds and FRS, did similar studies of Cu₂O. But, at that time there was no relevant theory for interpreting the results. J.P. Shelton and D.J.M. Bevan (later Professor at the Flinders University, Adelaide) measured the semiconducting properties of metal oxides at high temperatures, particularly the dependence of electrical conductivity on oxygen pressure. It was recognised that, with refractory oxides, they were dealing with the electronic effects of surface processes – chemisorption and the primary acts in oxidation and reduction. The work led to the formulation of ideas parallel to, but independent of, those concurrently being developed by W.E. Garner at Bristol, especially in his work with T.J. Gray and F.S. Stone – ideas that JS then applied to the primary steps in reactions such as corrosion, reduction of oxides or roasting of sulphide ores.

In 1945, with J.R. Richards, and using radioactive tracer methods (first learning how to make geiger counters and scaling circuits), the self-diffusion coeffiecient of Pb in n- and p-type PbS (respectively stoichiometric, or more likely Pb-rich, and S-rich) was determined. The aim was to distinguish between two possible modes of intrinsic disorder, the Schottky and Frenkel types. Faster diffusion was observed in the S-rich material, consistent with the former.

Much wider ranges of non-stoichiometry appeared to obtain for the fluorite-related, rare-earth oxides PrO_2-x and CeO_2-x: PrO_2-x as suitable oxygen potentials (0 < p(O₂) < 1 atm), and so an attempt was made to obtain good isothermal data for equilibrium oxygen pressure versus composition. But first a reasonable quantity of pure praseodymia had to be prepared. Starting with Byron Bay monazite, they used the method of fractional crystallisation. It was an agonising process, as two persistent stories confirm. One involves T.A. O'Donnell (later, Professor at Melbourne) and a fellow student: it appears that while the former was away on holiday the latter, not realising the extreme importance and value of the partly-fractionated contents of some inadequately-labelled beakers, washed them out to use for some other purpose. The other is similar, but involves JS himself. His beakers – unlabelled (the paper labels had long since fallen off) but in an ordered row on a hot-plate – were removed by an enthusiastic cleaner who, after the dusting process, replaced them more-or-less at random. This catastrophe – which destroyed months of painstaking work – did momentarily, cause him to forsake his customary sangfroid. Afterwards, the best material they produced was used by R.L. Martin (later FAA, Professor at Melbourne, ANU and Monash, and Vice-Chancellor at the last) for the intended p(O₂)-x-T study. The results indicated continuous, bivariant isotherms [p(O₂) vs. x]. As it later transpired however, this was because the oxide was insufficiently pure: the presence of even quite small amounts of impurity (such as Nd³⁺) prevents long-range ordering of the anions (in PrO_2-x). Some years later, using material purified by ion-exchange, LeRoy Eyring and co-workers showed that, at lowish temperatures (mostly below 400°C), a series of ordered intermediate compounds appears between Pr₂O₃ and PrO₂ although, at higher temperatures, two widely-nonstoichiometric phases obtain.

At this juncture ( 1946) the wartime Manhattan Project was metamorphosing into post-war, civilian research on atomic energy; in particular, the Atomic Energy Research Establishment (AERE) was being set up at Harwell in England. JS was invited to join the embryo Chemistry Division there and, after some hesitation and a discussion with M.L. Oliphant (a participant in the Manhattan Project; later Sir Mark, FAA, FRS, Director of the Research School of Physical Sciences at the ANU, Governor of South Australia, etc.), accepted an appointment as Senior Principal Scientific Officer. As in the RCS earlier, there was little prospect of promotion at Melbourne – Readerships were few – and the needs of his family were still growing.

Harwell 1947-1954

JS joined AERE, Harwell in September 1947, his responsibility being 'general oversight of inorganic chemistry research of all kinds and to run, directly, a small and quite uncommitted research group'. Unwilling to work on the more applied, process-development side of the Chemistry Division's programme, he was not part of the plutonium-separation research. He and his senior colleagues Gluckauf and Wild were unofficial deputies to R. Spence, Head of the Chemistry Division (later FRS and Director of AERE) but, as JS was on the spot in Building 220, the main divisional laboratory, much of the interesting work of deputizing fell on him. At the time he joined, the skeleton Chemistry Division was in Canada: it arrived at Harwell during the next few months, as also did new recruits (R.W.M. D'Eye, K.B. Alberman, K. Saddington and, later, K.D.B. Johnson, K. Dawson and M.W. Lister). Building 220 was then still a hole in the ground near the main runway of Harwell airfield and, while Buildings 147 and 149 were being converted to laboratories, the group was accommodated in the old sergeants' mess. In due course they shared, with Seligman's Isotope Division, one floor in Building 149.

Continuing his now dominant interest in the solid state and non-stoichiometry, JS chose to work on uranium oxides and hydrides. From the pre-war work of Biltz and Müller, UO_2+x appeared to have a wide, but undetermined, stoichiometric range; and nothing more on this topic had emerged from the Manhattan Project. It was clearly an important material as a precursor for U metal and a future fuel material, and more needed to be known about its properties and chemical behaviour. A great deal of work was done, including the characterization of a phase 'U₄O₉' – an ordered, fluorite-related structure that, in fact, has only recently been solved(4). JS's group went on to study ternary oxide systems involving mainly U(IV) and U(V), developing high-temperature methods for experiments above 1,500°C. At first, X-ray facilities were concentrated in the Diffraction Physics Section of the Physics Division. An agreement for very extensive 'hands-on' work by JS's group was arranged, but it was not until 1950, 'after much argument at the Director level', that he was able to get an X-ray set for his group. The monopoly being broken, the Metallurgy Division and other sections also acquired their own X-ray facilities, 'greatly to the improvement of the efficiency of the Establishment's work and without detriment to the real task of Diffraction Physics [under J. Thewlis and G. E. Bacon, later FRS] in creating the new field of neutron diffraction'. JS met Gunnar Hägg at the 1952 'Reactivity of Solids' meeting in Gothenburg, and visited his department in Uppsala. As a result of this visit D'Eye was seconded to work for a few months with Hägg, and one outcome was the introduction of the Guinier X-ray camera into Britain.

Studies of surface and bulk diffusion processes in the oxidation of UO₂ and (Th,U)O₂ solid solutions were fruitful, particularly after JS was joined, in 1951, by L.E.J. Roberts (later, FRS and Director of AERE). The heat of adsorption of oxygen was measured as a function of oxygen coverage and, as the inward diffusion of oxygen was fairly rapid at room temperature, measurements were made at 90 K using a Bunsen calorimeter. This was quite successful until a serious fire involved the gallon or so of liquid oxygen in which the calorimeter was immersed.

Another research topic used the Hahn emanation technique. J.N. Gregory, a Melbourne graduate and member of the CSIRO team seconded to Harwell, checked the accepted interpretation of the method by observations on known structures, which he 'built' with Th-X atoms at determined depths in Langmuir-Blodgett multi-layers of long-chain Ba soaps. D.J.M. Bevan rejoined JS as a Junior Fellow at Harwell, and was also involved in Hahn emanation work. His results on diffusion in ThO₂ were unexpected, showing that radium (Th-X) was not incorporated but rejected to surface sites even at the vanishingly low concentrations involved in these labelled materials. A scientific assistant in that work was Stephen Moorbath, who left to study at Oxford as one of the first holders of a Civil Service Bursary, and who subsequently became a Reader in Geology there, and FRS.

As a Harwell Fellow, Bevan had considerable freedom and, in 1953, elected to study CeO_2-x by X-ray powder diffraction. CeO_2-x was more readily available than PrO_2-x, but the experiments were more difficult as its oxygen fugacities (dissociation pressures) are extremely low, and the lower oxides (x > 0) instantaneously oxidize in air at room temperature. However, by developing techniques used earlier by JS and D'Eye for handling ThI₂ and ThI₃ (and, earlier still, in Newnham's work on ZrI₃), Bevan devised vacuum methods for reducing CeO₂ and manipulating samples of the products for analysis and X-ray diffraction. In JS's words, 'Experimental enterprise, good X-ray powder work, sharp observation and judgement led him to recognise that he had uncovered a "homologous series" of intermediate oxides [later identified as] Ce_nO_2n-2. This was the start of Bevan's long interest in defective fluorite structures.'

Alberman and D'Eye recall that they joined JS at Harwell in late 1947, soon after he had arrived. Both were seconded there in lieu of military service, and were then raw, 20-year-old graduates. To each, in their first interview with JS, it was made clear that, although monolingual, they were expected to read the German literature and also, within a couple of weeks, to have learned glass-blowing and vacuum techniques. Alberman recounts, 'At this point J.S.A. produced, by way of encouragement, a glass mercury diffusion pump which he had blown as a form of physiotherapy whilst recovering from a broken wrist'. At first, they must have wondered if their secondment was a wise choice; but 'we were still technically in the army and, once accepted, could not withdraw from the positions without being classed as deserters'. Any doubts they might have had were rapidly dispelled. Alberman worked on uranium oxides, D'Eye on thorium compounds and on the uranium fluorides. The importance of the last was that UF₆ was the volatile compound used in separating the isotopes of uranium, and UF₄ was the direct precursor for uranium metal.

Although, as stated earlier, JS was not directly involved with the plutonium separation project, he was on various associated committees, and he was also responsible for the analytical work on fall-out from atomic tests. He recalled that he vividly remembered the first Russian nuclear explosion, for his group had to postpone work on a scheduled US test in Nevada in order to examine material from Siberia. The first Russian hydrogen bomb also brought him into contact with Lord Cherwell, and trouble, when the Paymaster General, with no prior arrangement, suddenly appeared at one of the analysis and intelligence team meetings at Harwell! JS noted dryly,

A D.C.S.O. (as I then was) cannot easily throw out an uninvited, senior Cabinet Minister. That meeting, and a later personal reporting to Cherwell (my only visit to a Ministerial sanctum in Whitehall) showed how difficult Cherwell could be. His discussion remained critical, based on scientific judgements, until the words "Russian" or "communist" cropped up: and at that, all objectivity went out.

Prior to the first British nuclear test, at the Monte Bello Islands in 1952, JS made a quick trip to Melbourne to arrange for urgent analysis of the air particulates that would result. Hartung, due to retire at the end of 1953, sounded him out as his possible successor as Professor and Head of the Chemistry School. JS had already received offers of senior positions in the UK. However, his deep apprehension about the danger of a nuclear war led him to accept the Melbourne position when it was offered, even though he had doubts about the research facilities, and the salary was no more than he was receiving at Harwell. After some haggling, he at least obtained two post-doctoral positions, then unknown in the Melbourne Chemistry School.

Melbourne (2) 1954-1959

One of these postdocs, B.G. Hyde, resumed Merial Morton's earlier study of the chemi-sorption of oxygen on PbS surfaces, but the results were still inconclusive. The other, K.J. Gallagher, following the Harwell work on the oxidation of UO₂ and the earlier Melbourne work on PrO_x, studied the kinetics and mechanism of the oxidation of fluorite-related C-type Pr₂O₃. This time the material was pure, prepared by building an ion-exchange column and separating Pr from Nd in an available sample of 'didymium'. With J.O. Sawyer, some work on UO₂ was also continued and, with C. Barraclough (later Reader at Melbourne), thermodynamic measurements [p(O₂)-x-T] were made on CaUO_3.5-CaUO₄ – apparently a continuous solid solution which also had a fluorite-related structure. Other studies of uranate chemistry were made with J.G. Allpress (later, at CSIRO with J.V. Sanders and A.D. Wadsley, distinguished by his key role in pioneering lattice imaging electron microscopy), and with Meta Sterns and Jill Kepert. Sterns also used p-x-T measurements and X-ray powder diffraction (5) to unravel the lead oxide phases. With Kepert a great deal of new, complex chemistry of the alkali metal uranates was uncovered. Thirty years later, JS returned to this work for his last years in the laboratory, at the ANU in Canberra.

Although PhD degrees were now possible in Australian universities, research funds and students were both in very short supply. (At one stage there was a crisis because the supply of clamps and bosses was insufficient for the needs of all the glass vacuum systems being built in JS's group: he suggested corks and wire instead!) Despite the fact that he revitalized research activity in the Chemistry School, it is clear that on this occasion JS's time in Melbourne was something of a disappointment. He left after only five years, before the beneficial effects of the Murray committee of enquiry into Australian universities (1957) became apparent. Attempts to obtain funds from industry for 'unfettered, fundamental work', while not entirely unsuccessful were, as ever, pretty fruitless. But good relations with Walter Boas, then Chief of the adjacent CSIRO Division of Tribophysics, were a great help. D.F. Klemperer, a post-doctoral fellow with Boas, was seconded to work with JS; they studied the chemisorption of oxygen on nickel metal films by examining its influence on the photoelectric effect. This was JS's first acquaintance with UHV (ultra-high vacuum) techniques: before long he had a UHV system built in his personal laboratory and, alone, started research on field-emission microscopy (FEM) – then a rather new technique, but clearly a powerful probe for studying the chemistry of simple reactions on solid surfaces. People in the School were very impressed to see the Head of School doing his own research with his own hands, but this was entirely characteristic of JS until he was over 80.

An intriguing aspect of this period in Melbourne is the relationship, personal and professional, between JS and David Wadsley: so far it remains enigmatic. Wadsley was a crystallographer in the CSIRO Division of Mineral Chemistry, also in Melbourne, who in the 1950s was developing ideas on non-stoichiometry that were radically different from those espoused by JS in 'The conditions of equilibrium of 'non-stoichiometric' chemical compounds' [Proc. Roy. Soc., A, 185 (1946), 69-89]. Controversy did not surface explicitly and publicly until 1962, but their difference in outlook must have been obvious to them both in the late 1950s. It concerned the structures of grossly nonstoichiometric crystals – especially whether they were random solid solutions of classical 'point defects', as the thermodynamicists believed, or, as Wadsley maintained, sequences of ordered structures with closely-spaced stoichiometrics, such as the crystallographic shear structures. To those around him in Melbourne in the 1950s, JS certainly did not appear to be overtly enthusiastic about the new ideas, although it is now obvious that he was well aware of the facts on which they were based (he documented them in the first three editions of 'Emeleus and Anderson'). Indeed, somewhat ironically, they included Bevan's Harwell results on the cerium oxides – although it also seems likely that this was not fully appreciated at the time. It is certainly a pity that there is no contemporary account of this controversial situation. Reminiscing years later (in 1982) JS recalled that:

Wadsley had been excited by [crystallographic shear structures] when I was in Melbourne, in the 1950s, and had perspicaciously seen their importance. As a crystallographer, he was impressed by the increasing number of instances in which reportedly nonstoichiometric, thermodynamically bivariant systems were later being shown to embrace successions of ordered, intermediate compounds; well defined in composition and based on some common structural feature that generated a homologous series. The extreme view was that the concept of nonstoichiometry was illusory. I could not accept that extreme view as tenable; statistical mechanical considerations, the existence of order-disorder transitions, reversible transitions between ordered intermediate successions [of phases] and bivariant behaviour – these could not be explained away. There was ground for controversy, in which he became prominent on the one side and I, to some extent, on the other.

But, just as this situation should have become interesting, other events intervened. In 1951, after Sir Patrick Linstead's retirement from the position, JS had been urged to accept the Directorship of the NCL (National Chemical Laboratory, or Chemical Research Laboratory of the Department of Scientific and Industrial Research) in England, but had declined because among other things it seemed to be in the shadow of the NPL (National Physical Laboratory). In the middle of 1959, after Sir Harry Melville had overhauled the DSIR, he again received a pressing invitation to take up the position. This time he accepted, and arrived at the NCL in October of that year. As it turned out, this may well have been triply unfortunate: first, as intimated above, the tide was turning for research funding in Australian universities; secondly, two eminent solid-state chemists had arranged to spend the following year (1960) in JS's Melbourne School, namely LeRoy Eyring, from the State University of Iowa, and Sten Andersson, from the National Defence Research Laboratories in Stockholm, who was a close collaborator of Wadsley's; thirdly, the NCL eventually turned out to be as disappointing as he had thought it might be some years before, since many people, even at DSIR headquarters, still seemed to regard it as no more than an appendage of the NPL. It is therefore fortunate that he did not stay there very long. (Of his peregrinations he used to say, 'Human nature being what it is, it is a good thing to move on every few years.')

The National Chemical Laboratory 1959-1963

In fact, in 1959, a large proportion of the staff, resources and vigour at the NCL was in the Mineral Chemistry Group. This group – more or less independent of the rest of the NCL – worked on industrial problems under contracts with the UKAEA (UK Atomic Energy Authority) with little concern for background, fundamental science. And there were many additional problems and difficulties, on only some of which JS felt he was able to have any ameliorating effect. His complaint was that 'there was too much dead wood and too little turnover of staff'. And, as had been the case in Melbourne, there was a paucity of funds for modern instruments. His notes make it clear that he was very frustrated by the difference between his assessment of the purpose and possibilities for NCL and those of the bureaucracy.

However, he did set up a small personal research unit in which J.P. Jones took up and developed the FEM work JS had started in Melbourne, studying the structure and work function of metals (Cu, Au) adsorbed on tungsten; and D.W. Bassett set up Britain's first FIM (field ion microscope) and studied surface self-diffusion on tungsten. B.E.F. Fender also joined JS there. In addition, and more importantly, the hallmarks of the revolution in understanding the nature of non-stoichiometric compounds, referred to above, were beginning to appear and, although JS did little or no experimental work in this area at NCL, he was deeply involved in these important developments, as we shall discuss in the next section.

Within a year of his arrival at NCL, JS had been invited to take up the new Chair of Inorganic Chemistry at Oxford but, attractive though this was, he felt committed to the NCL and regretfully explained to Sir Cyril Hinshelwood why he felt compelled to decline. He thought that was the end of the matter, but about two years later the invitation was renewed and he now felt able to accept. After four years the NCL was in better shape, and perhaps a change of leadership there would improve matters still further. (It did not!) So he accepted the renewed invitation, resigned from DSIR in December 1962 and took up his new appointment in October 1963.

Oxford 1963-1976

At once JS involved himself in reorganizing the Inorganic Chemistry Laboratory (ICL), including provision of accommodation suitable for large, sophisticated, modern instruments; part of this later (1965-) housed an electron microscope, an acquisition undreamed of a few years earlier, but now seen to be highly relevant to the structural problems of solid state inorganic chemistry. He found the ICL to be deficient in equipment, and what did exist was very much 'private property'. But he 'stirred the possum', encouraged corporate rather than individual thinking, and received treatment from the University Grants Commission that he regarded as generous – about £180,000 in two years. He also encouraged staff to apply for grants from the DSIR/SRC. These efforts, plus the provision of a common room and the introduction of a morning coffee session, expanded personal contacts and improved morale.

B.E.F. Fender, later Director of the Institute Laue-Langevin and now Vice-Chancellor at Keele, joined JS from the NCL as his first appointee: he studied the thermodynamics of grossly nonstoichiometric oxides such as 'MnO' and 'FeO', using solid-state, electro-chemical cells. (Later Fender moved on to crystallographic work, especially neutron diffraction.) There was also some work on matrix-isolation spectroscopy of transient species and, with G.K.L. Cranstoun, a post-doctoral fellow on DSIR/SRC funds, more FIM work. Together JS and Cranstoun set out to see if reaction on a crystallographically perfect metal surface could be examined and followed in atomic detail: they studied the chemisorption of a small fraction of a monolayer of oxygen on tungsten, a body-centred-cubic structure, and its field desorption. Later, after the project was relinquished to Cranstoun, similar experiments were done on Ir (a face-centred-cubic structure), Rh (a hexagonal-close-packed structure), and even Fe. This was the last work JS did in the area of surface chemistry; it is clear that he thought highly of it, especially of Cranstoun's experimental ability. But we must now resume the discussion of his abiding interest in nonstoichiometry.

Much later, in 1982, JS wrote:

My chief intention at Oxford was to pursue the dominating interest in nonstoichiometric crystals, particularly the relation between gross deviations from stoichiometry and the ideas about crystallographic shear structures etc. that Wadsley and I had debated at Washington (1962) [cf. below]. Although Magnéli and his followers had discovered these and had determined the structures of the Mo, W, Ti and V homologous series [of oxides] (6), it was Wadsley who recognised and preached their significance (7). However, only their structures were known – their stability ranges, thermodynamics and other properties were quite unstudied. The contention was whether genuine nonstoichiometric phases existed at all, or whether, at true equilibrium, the supposedly nonstoichiometric state would be resolved into homologous series of CS phases, superstructure phases or other fully ordered species. My aim was therefore to start both exploratory experimental work and work particularly aimed at that basic, philosophical problem.

Quite early in his time at NCL, JS had been invited to contribute a chapter on non-stoichiometric inorganic compounds to a book being organised by L. Mandelcorn of the Westinghouse Research Laboratories (8). But having enough on his mind and shoulders at the NCL, and as the editor's approach was clearly designed to emphasise structure rather than thermodynamics, JS declined. However, he suggested to Mandelcorn that Wadsley might be the best possible contributor and the outcome was, in JS's view, 'an invaluable and masterly chapter, with a structural survey of the field that set out Wadsley's point of view. It remains an important contribution 20 years later.'

Then Roland Ward of the University of Connecticut arranged a symposium on non-stoichiometric compounds for the Annual Meeting of the American Chemical Society, to be held in Washington, D.C., in March 1962. Deliberately looking for a confrontation of views, he invited Wadsley and JS as plenary speakers. The symposium (9) was very timely: the experimental evidence had accumulated and opinions sharpened to the point where 'a confrontation' could be useful. JS's paper still emphasised the statistical thermodynamic approach but pointed out that interaction between 'point defects' could produce 'defect clusters' which, by further ordering, could produce 'superstructures' (as in wüstite, 'FeO', in the fluorite-related oxides e.g. CeO_2-x, or in the crystallographic shear [CS] structures). Wadsley's article discussed crystal structures, with emphasis on those involving CS [the oxides of W, Ti and (Ti,Nb)] and similar 'bronze' structures, and suggested that non-stoichiometry could arise by disordering the characteristic (ordered) structure elements of line phases. Thus, in a sense, and as JS later recorded, their two papers at this symposium did show some 'convergence of views: a common distinction between local, short-range order in grossly nonstoichiometric compounds, and long-range order in line phases of sometimes 'grotesque stoichiometry' (10), for example Na₁₀Ti₁₈O₄₁ or Ti₂Nb₁₀O₂₉.

>From this distinction between short-range and long-range order, JS started to think about the thermodynamics of microdomains of one structure and stoichiometry statistically distributed within a continuous matrix of another as a model (already propounded by Ariya (11)) for 'statistically disordered and defective crystals'. Such a model appeared to be ideal for interpreting the recent neutron-diffraction results of W.L. Roth on wüstite Fe_1-xO [Acta Cryst., 13 (1960), 140-149] and of B.T.M. Willis on UO_2+x [Nature, 197 (1963), 755-756]. Both these sets of data had been interpreted as indicating small defect clusters rather than a random distribution of point defects. As there were ample thermodynamic data for these two systems, JS attempted to apply Terrell Hill's 'small systems thermodynamics' (12). For several years he wrestled with the problem of trying to frame a satisfactory statistical thermodynamics treatment of the stability conditions of such 'nonstoichiometric' crystals, but he was never entirely satisfied with the results. His conclusions seem to have been:

(i) 'relaxation of the system [point defects in crystal] may go far beyond a simple rebalancing of coulomb forces, and may effect a structural transformation of the "solvent" crystal lattice';

(ii) 'the statistical thermodynamics treatment holds if point defects are replaced by extended defects or defect clusters';

(iii) 'one must conclude that thermodynamic measurements do not suffice to discriminate between alternative structural models'.

These are in accord with the best recent thinking on the subject, but they leave a complex theoretical problem; and no-one has yet improved upon JS's attempts to solve it.

Before starting experimental work at Oxford, JS invented a model, atomic mechanism for the generation of one family of CS structures from a parent ReO₃-type structure – by simply eliminating oxygen-only planes at regular intervals, and closing up adjacent remnants to regenerate octahedral coordination of the cations. The (100) projection of the corner-connected array of {Re}O₆ coordination polyhedra being a chequerboard pattern, he used red-and-white check shelf-paper to represent the ReO₃ structure, made scissors cuts along the planes of missing oxygens, and then pasted adjacent pieces together to make a correct representation of the derivative CS structure. In fact, the collapse procedure had already been explicitly described by Wadsley; JS's contribution was to identify the necessity to remove a specific plane of oxygen atoms to make the process feasible. He discussed this and, with that well known twinkle in his eye that always showed that he was rather pleased with himself, displayed his simple, elegant, shelf-paper model at a Gordon Research Conference on High-Temperature Chemistry in the following August (1964). With a little polishing via trans-Atlantic telephone calls, the original notion, and its extension to CS planes other than his original, were written up as a paper and presented (by LeRoy Eyring) at a meeting in Bordeaux the following month. It proposed a dislocation mechanism (novel in chemistry) for generating ReO₃-derived CS structures by reducing the parent type. Later, a similar trick was used to describe CS in the more complicated rutile-type structure.

Reacting to a perhaps excessive emphasis on oxides, JS also started research at the ICL on other types of compounds – particularly phase analysis and structural studies of rare earth carbides and oxy- and nitride-carbides with N.J. Clarke, now Reader at the Flinders University, Adelaide, and I. McColm, now Reader at Bradford. Thermodynamic measurements, using solid state electrochemical cells, were later made on the same systems (with A.N. Bagshaw, now Deputy Director of the Chemistry Centre, W.A.). T.B. Reed, an inventive experimentalist from the MIT Lincoln Laboratories, who spent a year in the laboratory on a DSIR/SRC Senior Visiting Fellowship, developed the required very-high-temperature techniques, and built the necessary equipment. Studies of manganese carbides followed.

In 1966, JS spent a six-month sabbatical period with D.J.M Bevan and B.G. Hyde in the School of Chemistry at the University of Western Australia. In honour of his visit a small, informal, two-day meeting on solid state chemistry was organized; the participants included the Western Australian groups and L.A. Bursill, J.M. Cowley, A.K. Head, S.W. Kennedy, A.C. McLaren, A.L.G. Rees, T.M. Sabine, A.E. Spargo, A.D. Wadsley and several others from eastern Australia. At this meeting, relations between Wadsley and Anderson visibly improved: some of us recall the almost audible cracking of the ice separating the two during a relaxed conversation at University House. Later that year, Wadsley and his wife were entertained at the Andersons' Islip home.

This was also the time of the first systematic application by chemists of electron microscopy (EM) to the problems of solid state chemistry, and it is appropriate to give brief summary of the immediately preceding developments. Following the Swedish discoveries, already referred to, of large numbers of CS structures, in the middle and late 1960s Wadsley and his co-workers had laid a massive groundwork on 'block' structures. The stimulus for chemists to apply the EM technique to their problems was an international conference, 'I – Electron Diffraction and II – The Nature of Defects in Crystals' (13) held in Melbourne under the auspices of the Australian Academy of Science and the International Union of Crystallography in August 1965. (JS was not present.) The first relevant EM paper by chemists on 'block' structures followed in 1968 (14). JS's first EM publications were in 1970. The 'revolution' in solid state chemistry therefore occupied the 1960s, coinciding with the early part of JS's period at Oxford.

JS had acquired an electron microscope in the middle '60s but, initially, was handicapped by not having a goniometer stage. However, with co-workers he carried out experimental work on CS structures in reduced MoO₃, rutile, and so on. With R.M. Gibb he discovered a quite new type of CS plane in the system TiO₂ + Ga₂O₃. Other work on rutile-related CS structures included theoretical studies, EM studies, and thermodynamic measurements an both vanadium and titanium oxides. And there were also EM studies of ferrites, bronzes and others. The phenomena of 'swinging CS' in reduced rutile and the fluorite-related 'vernier' structures (15) undoubtedly led him to his important reification of the idea of 'infinitely adaptive structures'.

His experimental pièce de résistance was the study of structure and reaction (oxidation and reduction) in the niobium oxide 'block' structures, with Browne, Hutchison and Nimmo. These became even more convincing – dramatically so – later in the 1970s when the resolution of commercially-available electron microscopes improved to better than ~3 A. Again he used a simple, homely device (cf. his previous FIM work), identifying blocks of different sizes by superimposing characteristic colours. This aided the detection of changes in block size during reaction, and the tracing of Wadsley defects in imperfect structures, a procedure which, for financial or other reasons, it was not always possible to reproduce in published papers. The principle reaction mechanism for 'zipping' or 'unzipping' blocks turned out to be the one proposed by Andersson and Wadsley (16) as an alternative to that of JS.

Thus JS's Oxford years were at the centre of an exciting period of intense development in understanding of many of the structural problems of nonstoichiometry, right down to the atomic level. Many visitors spent time in his laboratory, including several Australians (D.J.M. Bevan, F.J. Lincoln, A.C. McLaren, H. Rossell),as well as C.N.R. Rao who later became Director of the Indian Institute of Science and FRS – and JS, in turn, spent time at other laboratories. As well as visiting Perth, he spent another sabbatical period with LeRoy Eyring and others at Arizona State University in 1969, and he also visited India, principally to work with C.N.R. Rao. He retired from the Oxford chair in 1975.

Aberystwyth 1975-1982

JS's activity did not abate when he 'retired'. He aimed to continue work with another active solid state research group, free of all administrative chores, and this he achieved. Unable to remain at the Inorganic Chemistry Laboratory at Oxford, he considered several other options such as joining Peter Hirsch's department there, or John Thomas's department at Aberystwyth. There were also possibilities in Australia: he was pressed to join Lloyd Rees' CSIRO Division of Chemical Physics, in Melbourne, and there was correspondence about his going to the ANU in Canberra, with R.L. Martin, or to the Flinders University in Adelaide, with D.J.M. Bevan.

In the event, and with strong SRC support, he (and J.L. Hutchison) joined J.M. Thomas (now Sir John Meurig Thomas, FRS) at the University College of Wales, Aberystwyth. In the early '70s, their common interests in solid state chemistry and electron microscopy had rapidly led to a strong friendship. Now, he was generously provided with accommodation, excellent facilities, personal assistance and, in addition to the two already existing machines, his own electron microscope from Oxford. (There was no question about his simply settling down to write: he wanted to do 'real', that is experimental, research.) He took part in the life of the department, teaching in the MSc course in solid state chemistry and stimulating common-room discussion; and with J.L. Hutchison and a few students (notably Sian Crawford) he continued his EM work on reactions in block structures, and active publication. Young members of Thomas's group regarded him as something of a guru. Professor Mansel Davies was a close colleague.

Domestically, he organized and supervised the renovation of a cottage for Joan and himself on a smallholding he purchased at Abermagwr, in the country a few miles inland from Aberystwyth. In addition to the tribulations that might normally be expected from such a plan, while the house was being renovated he lived alone through an unusually cold and appallingly miserable winter in a caravan on the property. He seems to have been unaware that (as the story goes) his introduction of an Oxford architect to plan the rebuilding of a cottage in the heart of Wales did nothing to further relations with the local inhabitants. Nevertheless, he made friends amongst them, no doubt aided by the unusually cussed and peripatetic habits of Joan's small flock of living, Welsh-Nationalist mutton (17).

Early in 1978, John Thomas left Aberystwyth with his group for the Chair of Physical Chemistry at Cambridge. While JS strongly supported Thomas's move, his letters from this period show that it and the subsequent departures of his old friend Mansel Davies and then J.O. Williams (for UMIST) left him feeling isolated scientifically and topographically: he gave graphic descriptions of the difficulty in getting from West Wales to London, especially by public transport. Eventually, in 1979, perhaps foreseeing the imminent demise of the chemistry department at 'Aber', the Andersons decided to return once again to Australia, where three of their four children (and their grand-children) lived. This time they went to Canberra, where B.G. Hyde had recently taken up a chair at the Research School of Chemistry (RC) in the ANU. This fifth and final antipodean translation was achieved in 1982.

Canberra 1982-1989

More than most, JS thrived on frequent interaction with young scientists, and there were plenty of those at the RSC. As a Visiting Fellow, he was greatly valued and enjoyed by everyone, especially perhaps the young ones who were as impressed with his character, style and apparently unlimited knowledge as others had been in all his previous appointments. His knowledge and expertise were also valued elsewhere: he gave a number of seminars around the country, and was on a board that reviewed one of the CSIRO Divisions during the upheaval inflicted on that Organization by the Australian government. As ever, his continued pleasure in and insistence upon personal work 'at the bench', even as a senior academic, amazed everyone. He was there virtually every day, mainly continuing the work on alkali-metal uranates that Jill Kepert had started at Melbourne almost thirty years before, studying phase equilibria and structures by Guinier, powder X-ray methods since it was impossible to get sufficiently large single crystals. The problem turned out to be of extreme difficulty, indeed intractable. To some this was doubly unfortunate, for it kept him from the writing which, given his long, varied and rich experience – including 60 years of science and a great deal of travel, some in unusual places – would surely have been of unique interest.

When the new, 'high-temperature', super-conducting oxides hit the headlines in 1987, JS's enthusiasm and energy at the age of 79 were as great as anyone's, and especially effective because of his vast knowledge of solid state chemistry. He insisted on joining in the local effort, and over a short period was co-author of eight publications in the field. After the initial surge, he became immersed in a relevant but broader study of the still imperfectly understood ternary cuprates.

His drive decreased only when, in 1989, he was diagnosed as having cancer of his throat. Quietly amused that in spite of early exposure to more than a fair share of laboratory hazards (CO, nickel carbonyl, X-rays, asbestos fibres, mercury, and explosions) his health had been unusually good until that time, he now entered a difficult and traumatic period. Radiation therapy was delayed for a time because the 'machine' was out of action: his colleagues felt that he would have liked to fix it himself When he did receive radiation, he suffered the usual unpleasantness, losing some of his customary resilience and, for the first time, interrupting his regular work habits. For the last two years of his life, his wife Joan was confined to a nursing home, which was an added strain – but his children, his young colleagues and his secretary Caroline guarded his welfare and, in the laboratory, celebrated his 80th, 81st and 82nd birthdays. His final illness was prolonged and painful, necessitating hospital and irksome restriction. It was borne with characteristic stoicism, and he refused further radiation treatment. His phenomenal memory, puckish humour and sharpness of intellect persisted to the last days, relief came on Christmas Day, 1990. He was cremated at a family ceremony on 28 December and a memorial service, attended by many friends and colleagues, was held at the Church of St John the Baptist, Canberra, on 29 January 1991.

The man and the scientist

By nature JS was an unusually private man, so evaluation of his character is particularly difficult. Different people often see him differently, the various recollections falling into almost distinct groups with only minor overlap between them. He seldom seemed relaxed and open, even with close colleagues whom he regarded as old friends. The first reaction of most people was aptly summed by the description, 'in repose, his features are forbidding', a trait exemplified in several formal photographs.(We have utilised one that captures him more naturally.) One could say that, to the casual observer, 'He was the very model of a modern [i.e. 1880s] major-general'. He appeared to be unapproachable but this was a mask, for he loved a good argument, especially on broad scientific matters, and would then display his puckish sense of humour. One considered assessment is, 'To those he liked and valued JSA was usually considerate and generous. The rest of the human race was not so lucky...JSA's world of people fell neatly into two categories: there were those he dealt with reasonably, befriended, communed with; and all the rest were persona non grata. All administrators were in the latter category.' (Administration did not rate highly in JS's scheme of things, and this undoubtedly led to some unfavourable assessments of his character.) For those who knew him only after he was 30, it was difficult to imagine him as a young man: one correspondent recalls about the period 1946-7 (i.e. when JS was still in his thirties), 'the astonishing air of remote infallibility which surrounded him in our eyes'

The word invariably used – especially by his students or young collaborators – is 'awe', but there were exceptions. One that amused a few of his old friends and colleagues occurred in 1963, near Supai village in Havasu Canyon, Arizona. In order to cook outdoors one evening, JS and his companions spread out along the floor of this deep, narrow and isolated canyon, searching for fire-wood in the brush. While thus occupied, JS was caught red-handed by a very old Indian lady who, in her own language, upbraided him in no uncertain, and very shrill, terms. The burden of her remarks was quite clear – what he was collecting was in short supply and regarded as the exclusive property of the Havasupai, and was certainly not available to visiting academics. To his companions it was quite incredible that anyone dared to speak to him in this way. (But, during the washing-up afterwards he readily revealed what lay beneath the veneer.) On the same trip it was almost as incongruous to see him, beneath a large Stetson, riding a mule the several miles from the canyon rim down to the village – as was proved by the audience reaction when a slide of this was shown at an Oxford seminar not long afterwards. Most of his longstanding colleagues understood and enjoyed this up-beat side of his personality. It was also clearly apparent on many occasions at the dinner table, especially when he was visiting colleagues abroad. Then he was frequently 'the life and soul of the party', taking particular delight in charming the ladies and, as they sometimes dared to tease him (especially in Australia), playing along with the 'Oxford Professor' image.

He was famous for his skill at turning the apt phrase: 'Extrapolation is an inexhaustible source of fallacious reasoning', 'Infinitely adaptive structures', and so on. A non-scientific aphorism that he relished came from his 1957 family trip by car to the Flinders Ranges in South Australia, which were then rather inaccessible (he was delighted that 'one had to construct one's own road to cross a creekbed'). It is very dry, sheep-station country, and one station manager opined to him that 'In this country a man can make a living off a thousand square miles'. His sangfroid was equally well known: one example will suffice. One afternoon in Melbourne in the 1940s, there was a violent thunderstorm during which a lightning bolt struck somewhere within the University, just as JS and two students were about to enter a glass-panelled research laboratory. There was a blinding flash of light through the glass and, almost simultaneously, a tremendous clap of thunder. The two students recoiled in fright; JS simply commented, 'Hm; about 500 feet, I would guess.'

Another aspect of his somewhat legendary, even mythological status, is the stories about his various cars. In his first Melbourne period he had an A-model Ford, in which he travelled in the bush. K.B. Alberman recalls that when JS moved to Harwell, 'He had been allowed to bring personal effects with him and he interpreted this to include an ancient car with a canvas top in whose uniqueness he took great pride' R.W.M. D'Eye was also impressed, and recalls: 'Then there was his car – a model-T-Ford, which he drove with a degree of abandon amid the smoke screen from his pipe. One winter's night, driving back from a meeting of the Chemical Society my colleague whispered, as we only just negotiated a bend, "If he had not hooked his pipe round that tree we would not have made it this time" ' Model A or T, there is no doubt of the impressions created. During his second Melbourne period he had a Rover, built along the lines of a tank. It was said to have been the only car that 'took on' a Melbourne tram and won. In it, one evening, he also accepted the challenge from one of his junior colleagues, mounted on a bicycle, to a race down Johnston Street. Later, in Oxford, he progressed to a Daimler.

He also impressed by his first-rate intellect and his fabulous memory and erudition. In the Australian vernacular – which he would never have used – 'he knew bloody everything, twice over'. (Australian colleagues would have been amused to have heard him referred to, by some ignoramuses in his Oxford days, as 'that Australian professor'!) One of his students in Melbourne recalls typically: 'I knew him only as an undergraduate, 2nd-year, Chem IIA student in 1959...I enjoyed it – not so much because JSA was an easy lecturer to take "good notes" from but rather because his lectures were stimulating. He "lectured" rather than taught, addressing the unsolved as well as the worked-over topics he chose to present to us...It was for the insight JSA had that I best remember him.' Indeed, JS's lectures, though always full of 'meat', were not always easy to assimilate. A sidelight on this emerged during his visit to Perth in 1966, when he confided that he could no longer walk into a lecture theatre and give his lecture 'cold': old age was catching up with him (he was 58), and he now had to prepare his lectures before he gave them! On the other hand, in Melbourne during the 1960s, Derek Klemperer noted that lecturing was a strain for JS: 'Once I grabbed him as he came out of the [lecture] theatre...We got to his office and I suddenly realised he couldn't keep still and was breathing hard. Miss [Isobel] Rennie [JS's secretary] had left a note on his desk which he snatched up, screwed up and threw into the bin.'

He had a very fertile imagination and was passionate about his science. Klemperer also recalls an evening when JS walked the length of the corridor between their laboratories, and demanded that Klemperer immediately return with him to his own lab. 'There, with eyes aglow, he showed me a field emission microscope pattern ... He had made the tube himself... and everything else for running it. He was so proud and excited.' JS was loath to recognise experimental difficulties and limitations of any sort, either in his own work or in that of his young co-workers. These last often found such an approach forbidding (e.g. ignorance of the German language not being recognised as a handicap in reading the German literature), but invariably appreciated it in retrospect: despite their initial shock at Harwell in 1947, Alberman and D'Eye – typical of many others – write, 'It was a great privilege /an honour to have worked with him.' It is probably correct to describe him as a very 'British' chemist: he loved the simple string-and-sealing-wax solutions to practical problems, and practical work in general, especially glass-blowing. He also delighted in lightning, 'back of the envelope' (usually mental) calculations, which impressed and educated those graduate students who had not yet learned to remember Avogadro's number, the volume of a mole of ideal gas at NTP, the universal gas constant, and other handy quantities. His ability to concentrate deeply, excluding all surrounding activities and people, is also commonly remarked upon (but his exclusion, when someone else was similarly preoccupied, was not tolerated!).

When a question was put to him, the answer was sometimes preceded by a thoughtful silence: then out would come the well-phrased reply or, more rarely, an admission of ignorance. In his eulogy at the Memorial Service in Canberra, D.P. Craig (FRS, President of the Australian Academy of Science) described JS's memory as 'a storehouse that was marvellously well stocked and organized', and went on, 'He drew from it with magisterial ease. If you touched on something in his own research background you would reap a harvest. His answers were accurate and full. He was like Paul Dirac [who] was told by his father never to begin a sentence without knowing how it would end. In [Stuart's] case too what he had to say came out round and finished' This same quality appeared in his publications: though sometimes involved and not easy to read, they usually said what they had to say (and that was often a great deal) clearly and succinctly, and they were never polemical (18). The most difficult ones were those in which he was painting a broad canvas, especially his review papers on theoretical solid state chemistry. Over a period of forty years he published many of these, almost invariably without co-authors. The unusually large numbers of such invited papers was only partly due to the burgeoning of his field: mainly it was because JS was perceived as having a broad grasp of it, and an ability to organize a large mass of facts.

Perhaps the main contributions JS made to his science may be summarized as: (i) his quick perception of connections not previously noted, and (ii) his inspiration of many generations of students and young scientists in his areas of interest (and it must be remembered that these were very wide, extending well outside the generally recognized limits of solid state chemistry). The first of these may be exemplified by his application of Raman spectroscopy to valence problems, his union of the ideas of Schottky and Wagner with those of Fowler and Lacher to account for the composition ranges of nonstoichiometric compounds, his use of FEM and FIM to study surface reactions at the atomic level, and of the electron microscope to solve problems of reaction mechanisms in the solid state. The second strongly and broadly affected the laboratories in which, and the many individuals with whom, he worked, raising the reputations of the former (the University of Melbourne in the 1940s and in the '50s, the AERE at Harwell in between, and later the ICL at Oxford University) and the enthusiasms of the latter. Finally, of course, the influence of 'Emeleus and Anderson' was enormous and wide-spread (19). In 1986, when challenged to nominate his most important work, characteristically, he prevaricated: 'in retrospect it seems mighty little'. When pressed, he reluctantly suggested 'the years around 1944 in Melbourne', that is, the work summarized by 'The conditions of equilibrium of 'non-stoichiometric' chemical compounds', (Proc. Roy. Soc., A, 185 (1946), 69-89).

In retrospect, he seems to have been slow or reluctant to appreciate the significance of Wadsley's questioning, in the 1950s and early '60s, of the accepted ideas on the structures of nonstoichiometric compounds. Although aware of the new crystallographic developments as they occurred, he was perhaps subject to the 'tyranny of thermodynamics': one notes that it took him twenty years or so to progress from the primitive Schottky-Wagner model to more realistic pictures of 'defect structures'. (On the other hand, most people have still not made that transition!) It could also be argued that he placed undue emphasis on 'thermodynamic equilibrium' which is rarely achieved in solid state experiments. Similarly, when confronted with it towards the end of his career, he evinced no great enthusiasm for the important phenomenon of 'modulated structures'. Perhaps, in the field of crystal structures, there was some sort of unconscious 'mindset'? For example, his reaction to modulated structures and, earlier, to Bevan's infinite sequence of yttrium oxyfluoride structures (20) suggests that JS was uncomfortable away from the traditional concept of the unit cell. This is almost inconsistent with his own enunciation of the 'infinitely adaptive structure' concept; and so we are left with another enigma!

Because of his private nature, JS's extra-scientific accomplishments are more obscure. It is known that he was an amateur painter (at least until, at Harwell, he fell off his bike and broke his wrist). And – in spite of his early, endured but not enjoyed, experience as a young boy – he enjoyed music. His 1932 and 1934 (and other) walks in Eastern Europe and his later love of the Australian bush showed how much he enjoyed 'Nature'. His erudition extended to a considerable knowledge of the literature on the exploration of Australia, especially the explorers' journals, and of some of the territory involved; and this was readily and freely available to anyone interested. Other fields of expert knowledge sometimes emerged unexpectedly and surprised the listener, who had no idea that JS was likely to be informed at all, let alone well informed, on the subject concerned. It seems likely that there were other, hidden facets.

Appointments

• 1930-1938 Imperial College: Demonstrator, 1930-31; Assistant Lecturer, 1932-38.
• 1938-1947 University of Melbourne: Senior Lecturer in Inorganic Chemistry.
• 1947-1964 AERE, Harwell: Senior Principal, later Deputy Chief, Scientific Officer.
• 1954-1959 University of Melbourne: Professor of Chemistry.
• 1959-1963 National Chemical Laboratory, Teddington: Director.
• 1963-1975 University of Oxford: Professor of Inorganic Chemistry (invited to be first occupant of this new Chair). St. Catherine's College: Fellow.
• 1975-1981 University College, Aberystwyth: Honorary Professorial Fellow.
• 1981-1990 Research School of Chemistry, Australian National University: Visiting Fellow.

Honours

• 1943 Liversidge Lecture, Royal Society of New South Wales.
• 1944 H.G. Smith Medal, Royal Australian Chemical Institute.
• 1945 Syme Research Prize, University of Melbourne.
• 1953 Tilden Lecture, Chemical Society (of London).
• 1953 Elected FRS.
• 1954 Elected (Foundation) FAA. [Member of Council & Treasurer, 1956-1959]
• 1958 Masson Lecture, ANZAAS.
• 1963 Liversidge Lecture, Chemical Society.
• 1964 Fellow of the Imperial College.
• 1966 Matthew Flinders Lecture, Australian Academy of Science.
• 1973 Davy Medal, Royal Society of London.
• 1974-1976 President of the Dalton Division of the Chemical Society
• 1975 Award for Solid State Chemistry, Chemical Society.
• 1975 Longstaff Medal, Chemical Society.
• 1978 Honorary Fellow, Indian Academy of Science.
• 1979 Hon. DSc, University of Bath.
• 1980 Hugo Muller Medal/Lecture, Chemical Society.

Acknowledgements

A number of Stuart Anderson's erstwhile colleagues and others have provided personal recollections and other contributions which have been useful in the writing of this memoir: they include K.B. Alberman, N. Bartlett, D.J.M. Bevan, M.A. Bennett, D.P. Craig, H.J. Emeleus, R.W.M. D'Eye, Jane Figgis, N.N. Greenwood, R.D. Hoppe, D.F. Klemperer, T.A. O'Donnell, Joan T. Radford, Penny M. Richardson, L.E.J. Roberts, A.M. Sargeson, J.M. Thomas, R.J.D. Tilley and D.J. Whelan. We are grateful for these, and also for further assistance from his children, especially Jean Groves and Elizabeth Smith.

We are especially indebted to Jane Figgis and the Australian Academy of Science for the transcript of a partly autobiographical interview recorded, at the instigation of A.M. Sargeson, by JS in February 1985; and to the Academy and the Royal Society of London for his (written) autobiographical notes of December 1982 from which we have quoted extensively. We are also indebted to the librarians of the Academy and the Society Rosanne Clayton and Kate Douglas, for further assistance.

Notes

(1) Where not otherwise identified, verbatim quotations are from Anderson's autobiographical notes.

(2) On the other hand H.J. Emeleus says of Baker, 'He left his students to find their own research topics and, if they met with success, told them to publish in their own names. There was no team work, each having a separate topic of his own choosing. Baker was on his own and also worked with his own hands. He was always interested in a sudent's work but did not, as I recall, contribute very much. He did not read widely, as far as one could judge.'

(3) Much later (probably the early 1960s)! On the first page of 'Non-stoichiometric compounds'( Ann. Rep. Progress Chem. (Chem. Soc., London, 1946) 43 (1947), 104-120.) JS stated, 'Our present knowledge of crystal structure confers precise meaning on the term 'solid solution' as applied to crystals of atomic lattice types. In a crystal phase of ideal formula AB_n, a stoicheiometric excess of element B can be accommodated structurally in only three ways [our italics]: (I) Substitutional solid solution: B atoms replace A atoms on lattice sites proper to A. (ii) Interstitial solid solution: additional B atoms are located in inter-lattice positions. (iii) Subtractive solid solution: all B atoms occupy proper B lattice sites, but a number of A lattice sites is left untenanted'. (Cf. JS's remark, in the same notes, that 'It is difficult to retrace the sequence of mental processes'.)

(4) As paper by Bevan, Grey and Willis, J. Solid State Chem., 61 (1986), 1-7.

(5) Two Guinier cameras, for JS and for A.D. Wadsley, were built in a CSIRO workshop. Later, two more were built and, later still, the building of a similar camera, now slightly modified from Hägg's original design, became the final examination for students in the machine shops at the Hobart Technical School in Tasmania. The supervisor in this latter workshop, Mr Jack Leverett, was a friend of Wadsley's.

(6) Cf. A. Magnéli, Nova Acta Regiae Societatis Scientiarium Upsaliensis, Ser. IV, vol. 14, No.8, 19pp. (1950); G. Hägg and A. Magnéli, Rev. Pure Appl. Chem., 4 (1954) 235-250.

(7) Cf. A.D. Wadsley, Rev. Pure Appl. Chem., 5 (1955) 165-193.

(8) Published as Non-Stoichiometric Compounds, ed. L. Mandelcorn (New York: Academic Press, 1964).

(9) This is the Washington meeting referred to by JS in the quotation above. It was published as Nonstoichiometric Compounds (Advances in Chemistry Series, vol. 39), ed. R.F. Gould (Washington, D.C.: American Chemical Society, 1962).

(10) Wadsley's colourful term.

(11) S.M. Ariya and M.P. Morozova, Zhur. Obshch. Khim., 28 (1958), 2617; S.M. Ariya and Yu.G. Popov, ibid., 32 (1962), 2077.

(12) T.L. Hill, Thermodynamics of Small Systems, Parts I and II (New York: Benjamin, 1963-4).

(13) I – Electron Diffraction and II – The Nature of Defects in Crystals: Abstracts of Papers Presented at an International Conference, Melbourne, Australia, 16-21 August 1965 (Canberra: Australian Academy of Science, 1965).

(14) J.G. Allpress, J.V. Sanders and A.D. Wadsley, Phys. Stat. Sol., 25 (1968), 541-549. Tragically, Wadsley died in January, 1969.

(15) These provided perhaps the ultimate in new 'nonstoichiometric' phenomena: in both the composition range/stoichiometric variation is continuous, but all structures, at every composition, are perfectly ordered. (I) CS derivatives of the rutile type, TiO_x, 1.892 less than or equal to x less than or equal to 1.93 [L.A. Bursill, B.G. Hyde and D.K. Philp, Phil. Mag., 23 (1971), 1501]; (ii) fluorite-related yttrium oxyfluorides, Y(O,F)_x, 2.133 less than or equal to x less than or equal to 2.223 [D.J.M. Bevan, pp. 749-759 in Solid State Chemistry, N.B.S. Special Publication 364, ed. R.S. Roth and S.J. Schneider (Washington: U.S. Government Printing Office, 1972); cf. E. Makovicky and B.B. Hyde, Structure and Bonding, 46 (1981), 101-170]. The structural principles involved in the two cases are completely different, of course.

(16) S. Andersson and A.D. Wadsley, Nature, 211 (1966), 581.

(17) This was not their first attempt at being small-scale graziers: during the '50s their Melbourne house – in Balwyn, a salubrious mid-outer suburb – stood on two acres of land. Joan kept a few sheep there, in the middle of suburbia.

(18) Perhaps because another of his favourite sayings was, 'You should never fight with your enemies'?

(19) There were at least sixteen printings in English and four in other languages.

(20) T.L. Hill, Thermodynamics of Small Systems, Parts I and II (New York: Benjamin, 1963-4).

B.G. Hyde, Research School of Chemistry, Australian National University.

P. Day, Royal Institution of Great Britain.