Alan Walsh was the originator and developer of the atomic absorption method of chemical analysis, which revolutionized quantitative analysis in the 1950s and 1960s. Atomic absorption provided a quick, easy, accurate and highly sensitive method of determining the concentrations of more than sixty-five of the elements, rendering traditional wet-chemical methods obsolete. The method has found important application world-wide in areas as diverse as medicine, agriculture, mineral exploration, metallurgy, food analysis, biochemistry and environmental control, and has been described as 'the most significant advance in chemical analysis' in the twentieth century.
Alan Walsh was born on 19 December 1916 and brought up in Hoddlesden, a small moorland village in the borough of Darwen, in Lancashire, England, about twenty miles north of Manchester. He was the eldest son of Thomas Haworth Walsh, who managed a small family cotton mill in Hoddlesden, and Betsy Alice Walsh (née Robinson). He had an older sister, Evaline, and two younger brothers, Jack and Tom. In 1947 Alan emigrated to Australia and soon afterwards met an English-born nurse, Audrey Dale Hutchinson, whom he married in 1949. They had two sons, Thomas Haworth and David Alan.
The family cotton mill, Vale Rock Mill, Holden Haworth Ltd, was one of two mills that employed most of the people in Hoddlesden and many from the next town. Alan's father, Thomas, was an astute and remarkable man who managed the mill for 52 years, including the period during the late 1920s and 1930s when the cotton industry was badly hit by the depression. Thomas had the interests of his family at heart but laid down the rules and would not allow disobedience of any kind. Alan's mother, Betsy, was a charming, warm-hearted woman. Alan was very like her in some ways but his astuteness and determination came from his father. The village was quiet: just one pub, a school, a church and a village shop, which Alan's grandfather, Benjamin Walsh, ran and which sold everything from meat, vegetables and groceries to clothing. Benjamin was a friend and helper of the vicar, and had a big say in village affairs. Alan's grandmother, Mary (née Haworth), was a strong-minded woman of the old school who insisted on good manners at all times.
Alan's uncle and godfather, Marsden Walsh, described Alan as a delightful and serious boy. He was something of a loner and happy with his own company, qualities his uncle said would stand him in good stead. Even though he had quite a serious side to him, he had a great sense of fun. He had what was known as a lazy eye and for a period had to wear a patch. He made light of this and fooled about so that his friends thought nothing of it because he made them laugh. He had a great sense of humour, but was never unkind at someone else's expense. Whenever he talked about his early life, he would say he was brainwashed – too much religious teaching – and how narrow their life was. Once when asked what he wanted to do with his life he said 'I want to find things out and how they work'. So, when later he became a scientist his uncle was not surprised.
From the age of ten Alan attended the local grammar school in the nearby town of Darwen, where he passed the Northern Universities Matriculation examination in 1933 and the Higher School Certificate examination in 1935. Of his school days, Alan recalls:
I had no idea where I was headed during secondary school. In fact when I was 16 years old I was advised to do French, English and History and drop Science. At the time I was having trouble with eye strain and because I thought French, English and History would involve a lot of reading, I chose rather to study Mathematics, Chemistry and Physics in which I performed quite well…. 
At that stage I was very uncertain about my next step. I remember being attracted by a teacher training course, but my headmaster had other ideas. He finally spoke to my father and they persuaded me to apply for the honours course in physics at the University of Manchester. 
In October 1935 Alan entered the honours school of physics at the University of Manchester. In his second and third years he was a resident at St Anselm Hall, where the warden, the Reverend T.H. South, described him as 'a quiet, industrious student, with a sense of humour, and a sense of responsibility, who has been popular and respected in Hall. He has been a successful captain of our Association Football XI this season.' In his final year Alan took the joint course in physics and electrotechnics, an option available for honours students in physics.
Of his time at Manchester, Alan recalls: 
It was only after I went to university that I experienced the real joys of learning and research. I had been at university only two weeks when I attended a lecture by Professor (later Sir) Lawrence Bragg, who was head of the University Physics Department. In a lecture to the University Physical Society he told, in extremely simple language, the story of the pioneering work he and his father had done in their development of X-ray methods for determining the structure of crystals. The basic simplicity and beauty of their contribution greatly impressed us freshers: the drama was enhanced by the knowledge that their work had been of such outstanding merit that they were awarded the Nobel Prize.
Even after graduating from Manchester University, I still had no definite plans regarding my future employment. The problem was shelved when, much to my surprise, I was awarded a research scholarship to work on the determination of crystal structures by X-ray methods.
In August 1938 Alan undertook postgraduate research in the Physics Department at the Manchester College of Technology, which later became the University of Manchester Institute of Science and Technology. His supervisor was Dr William H. Taylor, Head of the Physics Department at that time and known for his X-ray work on mineral crystals. During this period Alan's research was influenced by Dr Henry Lipson, who proposed that he study the structure of ß-carotene, an important biological molecule that presented a considerable challenge to X-ray structure analysis at that time. He spent one year at the Manchester College of Technology and then continued the theoretical work on the analysis of ß-carotene for a further period after he had moved to the British Non-Ferrous Metals Research Association. He was awarded the degree of MSc (Tech) in 1944 for a thesis entitled An X-ray examination of ß-carotene. In 1960 he was awarded a DSc from the University of Manchester for his contributions to atomic and molecular spectroscopy.
In September 1939 Alan began duties as Investigator in the Physics Section of the British Non-Ferrous Metals Research Association (the 'BNF') in Euston Street, London, under the direction of the leading British spectrographer, D.M. Smith. Alan recalled:3
In the first place I was to work on the development and application of spectroscopic [emission] methods of metallurgical analysis, about which I virtually knew nothing, but with the intention of also, in due course, working on X-ray studies of metals. With the outbreak of war [on the day he was due to start], these plans were abandoned and for the duration I worked only on spectroscopy.
During World War II Alan was unable to join the Services because of his metallurgical occupation, but he undertook part-time service in the Home Guard, where 'he was put in the mobile cavalry (bicycle section), being of the athletic type. He had represented Manchester at tennis'. 
At the BNF Alan was given the task of determining which metals were being used in enemy bombers which had been shot down. The information was passed on to the war economists, who could then make deductions about how the German war effort was progressing. During this time Alan devised a number of methods for the rapid and accurate spectrographic analysis of aluminium, copper and zinc based alloys (1-4). These methods also had an important role in the production control of war materials and were widely used in industry. The usual procedure was to make the sample for analysis one electrode of an electric arc or spark and to examine the light emitted by means of a spectrograph. The presence of any element could be detected by noting the wavelengths of the spectral lines while their intensities were a measure of the concentrations.
During the course of this work Alan became aware that when following a method that worked well in one laboratory, difficulties arose when applying it in others. He then devised and built a prototype of the General Purpose Source Unit (8) from existing components and bits and pieces of government surplus. This unit was a highly versatile but simple electrical source unit, capable of generating a variety of electrical discharges, including arc-like and spark-like discharges, for use in spectrographic emission analysis. The 'Walsh Circuit' permitted a high degree of stability and reproducibility of the electrical characteristics of individual discharges. Alan assisted in developing the commercial form of the Source Unit, which was subsequently manufactured by Hilger and Watts Ltd, London as the BNF Spectrographic Source Unit FS 130. It appeared on the market in 1950 and was still being produced in the 1980s.
In January 1944 Alan was seconded to the post of Deputy Chief Chemist at the Metal and Produce Recovery Depot, Ministry of Aircraft Production, in Eaglescliffe, Durham. There he was in charge of the laboratory and technical operations concerning the preparation and inspection of aluminium ingots obtained by melting aircraft scrap.
In January 1945 Smith left the BNF to join Johnson Matthey, and Alan returned to the BNF as Chief Spectroscopist to take charge of the spectrographic research and to give assistance in other applications of physics to metallurgy. Later in 1945 Alan visited Germany as a member of the British Intelligence Objective Sub-Committee Team on Spectrochemical Analysis. Bill Ramsden, who worked at the BNF from 1950 to 1954 and later became a life-long friend of Alan's, writes: 
My own impressions are that Alan Walsh was a rather unique person, characterised by an original mind and an unusual ability to penetrate to the heart of a problem. He was also a 'character', possessed of a dazzling wit and a mischievous sense of humour, and one was very fortunate to be in his company…. As far as the BNF was concerned, I have no doubt that he created 'waves' in that establishment.
During the spring of 1945 the BNF was asked to explore the possibilities of developing a spectrographic technique for determining impurities in uranium metal, and Alan duly devised a method for doing this that was released for publication some years later (14). Around this period the BNF was involved in the 'Tube Alloys Project', which was a cover for the development of the British atom bomb. Former staff from the BNF were recently astonished to read in The Times  that a secretary at the BNF, Melita Sirnis (later Melita Norwood), had been recruited by the KGB and had been passing on 'highly sensitive' material to the Soviet Union for forty years under the code name of 'Hola'. Melita Sirnis had been personal secretary to the Director of the BNF, Dr G.L. Bailey, during the period 1939-1948 and would have had free access to the Association's work during that period. However, Bill Ramsden recalls that during his time at the BNF, from 1950-54, there was certainly no level of security or screening and doubts that 'highly sensitive' material would have been sent after 1950.  He writes 'Certainly, the thought of my early reports on photoelectric emission spectroscopy as well as the details of the Walsh BNF Source Unit being available to the Kremlin before being circulated to BNF members is slightly hilarious.'
Of his research at the BNF Alan wrote (83):
By the end of the war I think there was a general feeling of satisfaction, and perhaps even a state of euphoria, regarding the development of spectrochemistry. I believe few workers shared my strong conviction, which I frequently expressed, that further progress would require a completely new line of attack. I tried desperately hard to conceive totally different approaches but came to a total impasse…
I was particularly conscious of the fact that accurate analysis [by atomic emission] required standards of very similar composition to the sample for analysis. If one wanted to be cynical about this then one could claim that accurate spectrochemical analysis consisted in confirming that the composition of a sample was what it was supposed to be…
It was with a sigh of relief that I left these problems of spectrochemical analysis in 1946…
In 1945 Alan applied for an advertised position of Research Officer for Spectroscopic Investigations in the Chemical Physics Section of the Division of Industrial Chemistry, Council for Scientific and Industrial Research (CSIR), at Fisherman's Bend in Melbourne. The Chief of the Division, Dr I.W. (later Sir Ian) Wark, had proposed the establishment of a new Chemical Physics Section to apply modern physical techniques to the solution of chemical problems. The functions of the Section would include 'spectrographic work of a fundamental nature and general spectro-analysis for the Division'.  Dr A.L.G. (Lloyd) Rees, who took up the position of Section Leader of the new Chemical Physics Section in November 1944, convinced Wark to purchase several state-of-the-art spectroscopic instruments, including a large Hilger-Littrow quartz spectrograph, a Hilger-Müller double monochromator, a Beckman DU ultraviolet-visible spectrophotometer, and a Perkin-Elmer Model 12B infrared spectrometer. The Research Officer for Spectroscopic Investigations would be responsible for setting up a laboratory for emission spectrographic analysis and for undertaking research in the newly developing field of infrared spectroscopy.
In July 1945 a report of an interview at the Australian Scientific Liaison Office in London by a junior officer states 'Walsh does not give me the impression of one who could direct research, but I should imagine he would be a careful and painstaking worker'. The position was subsequently offered to a another applicant, who after a considerable period of time could not make up his mind. After Alan had contacted the Liaison Office to enquire about his application, Mr Lewis Lewis, who at that time was the Australian Scientific Liaison Officer in London, wrote to Wark about Walsh 'with whom I was quite well impressed'. In March 1946, in a letter to CSIR Head Office based on a recommendation from Rees, Wark wrote:
Walsh seems to have been in charge of a small team at the BNF and is held in sufficient regard to be sent to Germany… These facts, combined with Lewis' favourable impression, lead us to the conclusion that we have been unduly cautious regarding him. In any case, there is room for a man of his attainments for pure research work, even if we must ultimately seek another leader.
In May 1946 Alan was duly appointed to CSIR, but before leaving England, Wark and Rees arranged for him to spend a period of three to four months in the laboratory of G.B.B.M. (later Sir Gordon) Sutherland in Cambridge, obtaining experience in the new field of infrared molecular spectroscopy. During this period in Cambridge, Alan established life-long associations with two molecular spectroscopists, Donald Ramsay and Norman Sheppard, the latter of whom introduced Alan to the experimental and theoretical aspects of infrared spectroscopy. This work in Sutherland's group led to a paper in Nature on the infrared spectrum and molecular structure of phthiocerane (10). In a letter to Wark, Sutherland wrote 'in my opinion you have got hold of a very good man'.
Alan set sail for Australia via the USA, where he visited a number of companies and laboratories to see the various items of spectroscopic equipment that Rees had ordered for the new spectroscopy laboratory at CSIR. He arrived in Melbourne in April 1947 aboard the 'Dominion Monarch'. Upon arrival in the CSIR Division of Industrial Chemistry at Fisherman's Bend in Melbourne, Alan recounted: 
The main building of the laboratories was most impressive, almost posh. Behind was a motley collection of old army huts. But the scientific equipment was first class… The conditions for the 'men at the bench' were utopian. Individual freedom and initiative were not only permitted, they were actively encouraged; a bold failure was more highly regarded than a cautious advance. Red tape and bureaucratic nonsense were totally absent.
The working conditions bore no relationship whatsoever to the popular concept of a government-controlled organisation. The frequent arrival of new staff, many from overseas, and of magnificent new equipment contributed to the general feeling of excitement. It was as lively a place to work in as one could imagine. The high calibre of the leadership at that time is reflected by the subsequent careers of Sir Ian [Wark], his right-hand man Mr Lewis Lewis, and those who were section leaders…
We were a hard working bunch. Most of us worked on Tuesday and Thursday evenings, and there were usually several people in the laboratories on other evenings and at weekends. Even the Minister [in charge of CSIRO] R.G. Casey used to visit the laboratories at weekends. He took an earnest interest in all the research projects…
An amusing facet of life at that time was that many of us were operating instruments and techniques which were unique in Australia. We could therefore claim undisputed leadership within Australia of various areas of chemical research. We rather enjoyed being referred to as a 'pride of prima donnas'.
Whilst life was fun it was also earnest, and there was no escape from Wark's insistence on excellence. As an example, the inimitable R.G. (Dick) Thomas said, 'If you tell Wark in the morning you have discovered how to annihilate gravititational forces, he'll want to know what you're going to do in the afternoon'.
Upon commencing at CSIR Alan set about installing the new Perkin-Elmer Model 12B spectrometer, thereby establishing his first interaction with the Perkin-Elmer Corporation. This was the first operating infrared spectrometer in Australia, and there was a steady stream of requests for service and collaborative work from organic chemists both within the Division and outside it. Alan was particularly interested in understanding the mechanics of the technique and studying the structure of small molecules. Together with Arthur Pulford, an MSc student from the University of Sydney, he studied the vibrational spectrum of nitrosyl chloride (NOCl) and calculated its geometry and thermodynamic properties (15).
The Perkin-Elmer Model 12B, like all commercial infrared spectrometers at that time, used a direct current (DC) amplification system. To obtain good spectra with such systems was difficult because small changes in ambient temperature caused pronounced wandering of the baseline. Alan recalled (75):
The best spectra were obtained late at night or in the early hours of the morning. It was therefore a memorable occasion, which substantially improved my quality of life, when our Model 12B was converted to a Model 12C. This incorporated a fast thermocouple, which permitted the use of a modulated light source and a synchronous AC [alternating current] detection system and completely removed the problem of drifting base lines. Recording infrared spectra was transformed from a chore to a pleasure.
Alan soon realised that the resolution of the Perkin-Elmer (prism-based) spectrometer was quite inadequate for resolving the rotational lines of any but the lightest molecules, and even in these cases the full details of the spectrum were not revealed. To improve the resolution he devised a simple and elegant modification of the infrared prism monochromator in which radiation was passed two or more times through the same optical system (17). To do this he placed a pair of right-angle mirrors at the exit slit of the spectrometer to reflect the radiation back through the prism and, to isolate the desired multiple-pass beam from the other beams, he placed a rotating 'chopper' in front of the additional mirrors to modulate only the multiple-pass light and fed the output of the thermocouple detector to an amplifier tuned to the frequency of the chopper. An additional advantage of this modification was that the level of stray light, hitherto a major problem in infrared spectroscopy, was reduced dramatically.
Alan's 'double-pass monochromator' was patented in 1950, with coverage in Australia and eight overseas countries. Perkin-Elmer, the world's major manufacturer of infrared spectrometers, secured an exclusive licence and in 1953 began manufacturing a kit of 'Walsh Mirrors' to allow the conversion of their standard infrared spectrometer to a double-pass monochromator system. This experience with patenting and licensing and the interaction with Perkin-Elmer had significance for future events in that it involved Alan personally with the commercial aspects of scientific investigation.
Shortly after his arrival at CSIR Alan also initiated a project to investigate the fundamental processes occurring in spectroscopic atomic-emission sources, and in particular to attempt to correlate the emission of radiation from the discharge with the electrical phenomena occurring in circuits containing an arc or a spark gap. He had the instrument workshop construct a source unit to the Walsh BNF design (8), which was capable of producing sparks that were electrically identical. Although such source units were by then well established in laboratories in the UK, the CSIR unit proved rather unreliable and was not sufficiently stable for the research Alan was proposing. John Shelton, who worked with Alan on this project, writes: 
It is interesting to speculate whether Walsh would have invented atomic absorption analysis if the source unit had been successful and allowed the planned research on inter-element effects to proceed. The frustration of the planned emission work stimulated him to think more and more about the tremendous number of sample atoms in the ground state, compared with the few, sensitive to minor changes in electrical and other conditions, in excited states.
In an address to the Silver Anniversary Symposium on Great Moments in Analytical Chemistry at the Pittsburgh Conference in 1974, Alan recounted (68):
My initial interest in atomic absorption spectroscopy was a result of two interacting experiences; one of the spectrochemical analysis of metals over the period 1939-46; the other of molecular spectroscopy over the period from 1946-52. The interaction occurred in early 1952 when I began to wonder why, as in my experience, molecular spectra were usually obtained in absorption and atomic spectra in emission. The result of this musing was quite astonishing: there appeared to be no good reasons for neglecting [atomic] absorption spectra; on the contrary, they appeared to offer many vital advantages over atomic emission spectra as far as spectrochemical analysis was concerned. There was the attraction that absorption is, at least for atomic vapours produced thermally, virtually independent of the temperature of the atomic vapour and of excitation potential. In addition, atomic absorption methods offered the possibility of avoiding excitation interference, which at that time was thought by many to be responsible for some of the inter-element interference experienced in emission spectroscopy when using an electrical discharge as light source.
A number of journalists have written (and it is commonly believed) that Alan conceived the atomic absorption method of chemical analysis in 'a flash of inspiration' in early 1952. [11,13,14] However, a colleague of Alan's at the BNF during the war, Sidney Payne, writes: 
I think that he [Alan] envisaged atomic absorption far earlier than indicated by Karen Robinson.14 I clearly remember chatting to him whilst I was using a simple flame [emission] photometer and he commented upon the fact that only a small proportion of the atoms were excited by this technique and that it would be better if some way could be found to measure the much larger quantity of unexcited atoms. My reply was 'Well, why don't you go away and think about it'. History confirms that he did just that.
The following is a reconstruction of the events leading up to Alan's establishment of the principle of the atomic absorption method, based on articles by Alan (68, 82), a colleague John Shelton,10 and Andrew McKay,11 and a recorded interview with Alan: 
On a Sunday morning in March 1952 Walsh was working in the vegetable garden of his home in the Melbourne bayside suburb of Brighton when he suddenly had a revealing flash of thought, something that stemmed from his earlier work in related fields. He hurried inside, dirt still on his shoes, and phoned his colleague, John Shelton. 'Look John!' he exulted. 'We've been measuring the wrong bloody thing! We should be measuring absorption, not emission!' John reminded him: 'We've been through that before – you can't work out the concentration of a sample from the absorption because of the emitted light at the same wavelength'. Walsh replied: 'I've thought of that. We'll use a chopper on the source and a tuned amplifier, so the light emitted from the sample won't matter.'
Early next morning Walsh set up a simple experiment, using the element sodium. By morning tea he had a successful result. 'I was very excited and called in my colleague, Dr J.B. Willis, who at that time was working on infrared spectroscopy and later was to make important contributions to the atomic absorption method of chemical analysis. "Look", I shouted, "that's atomic absorption". His reply, which I have never let him forget, was "So what?" This was typical of the general reaction to my early work on atomic absorption'.
It would appear that the 'revealing flash of thought' on the Sunday morning in March 1952 alluded not to Alan's initial conception of the atomic absorption method of chemical analysis but rather to his sudden realisation, after a considerable period of mulling over the subject, that 'atomic absorption spectra appeared to offer many vital advantages over atomic emission spectra' (68) and that 'what we needed to do first was actually to measure absorption'. 
In his initial, simple demonstration of the atomic absorption method, Alan used a standard sodium vapour lamp operated from a 50 Hz mains supply and thus had an alternating output, so that it was not necessary to use a 'chopper'. The sodium D lines from this source were isolated, but not resolved from each other, by means of a simple direct-vision spectrometer and their combined intensity was measured by means of a photomultiplier tube, the output from which was recorded on a cathode ray oscillograph. Amplification of the signal was by the AC amplifier in the oscillograph. A simple air-coal gas flame was interposed between the sodium lamp and the entrance slit of the spectrometer. When a water solution containing a few milligrams of sodium chloride was sprayed into the air supply of the flame, the cathode spot on the oscillograph deflected to zero, thus establishing the principle of the atomic absorption method of chemical analysis.
In his chairman's address (71) to the Fifth International Conference on Atomic Spectroscopy in 1975, Alan conjectured why atomic absorption spectra had remained largely unexplored for almost one hundred years since Kirchhoff had first interpreted the Fraunhofer lines in the spectrum of the Sun as atomic absorption lines and used them to identify the elemental constituents of the solar atmosphere, and since Kirchhoff and Bunsen had founded qualititative spectrochemical analysis based on atomic spectra emitted by substances vaporized in a flame. Alan believed that one of the reasons atomic absorption spectra had been neglected for so long was a misunderstanding regarding the implications of Kirchhoff's law, which states that 'for radiation of the same wavelength at the same temperature the ratio of the emissive power to the absorptive power is the same for all bodies'. Alan pointed out that this law was often interpreted as 'good radiators are good absorbers and poor radiators are poor absorbers' (which holds only for radiation of a given wavelength and a given temperature) and that it had generally been assumed that methods based on emission spectra would be equally applicable to the same range of elements as those based on absorption spectra, which were generally much more difficult to measure, especially in a luminous flame. In early 1952 Alan began to realise this may not be the case and as a result of his two interacting experiences he began to wonder why it was that spectroscopists usually measured atomic spectra in emission and molecular spectra in absorption. As a result of considering this problem he concluded that atomic absorption spectra could prove much superior to conventional atomic emission spectra for many spectrochemical analyses.
Although his initial, simple demonstration of the atomic absorption method was performed using a sodium vapour lamp as the light source, Alan envisaged that the source would generally be a 'white-light' continuum source, such as a hydrogen or tungsten lamp, that would be capable of being used for the whole range of metals (68). However, when he attempted to determine copper and zinc using a continuum source and a high-resolution Hilger-Littrow spectrograph, he found the sensitivity to be disappointingly low (68). He realised that the resolution of the spectrograph was insufficient to accurately measure the profile of the extremely narrow absorption lines (about 0.003 nm) and that, even if a spectrograph or monochromator with much higher resolution became available, the energy transmitted over the small spectral bandpass would be much too low to provide adequate signal-to-noise ratios. Alan recounted (75):
I decided to abandon all attempts to produce a high-resolution dispersion system and [instead] to obtain high effective resolution by replacing the continuum light source by atomic spectral lamps which emitted lines which were considerably narrower than the absorption lines they measured. If the emission line is sufficiently narrow, the peak absorbance can be measured, and this can be correlated with atomic concentration. This concept of using a sharp-line source was the vital step in the development of atomic absorption spectrophotometers. It not only obviated the need for a high-resolution monochromator, it also gave atomic absorption methods one of their most attractive features. This is the ease and certainty with which one can isolate the required line, a characteristic that results from the fact that the line to be selected is usually one of the strongest emitted by the lamp, and it is only necessary to isolate this from other lines emitted by the lamp. This contrasts with emission methods, in which it is necessary to isolate the required line from all other lines emitted by the sample, many of which [from other elements] may be much more intense and at neighbouring wavelengths.
A CSIRO colleague, Alec Moodie, recalls that during the (North American) summer of 1952, when he was at Pennsylvania State University, he received an airletter from Alan with a sketch of his proposed atomic absorption scheme and a comment at the end, 'The sharp-line source doesn't yet exist!'
Apparently while reading Tolansky's book High Resolution Spectroscopy,  Alan learned that a hollow-cathode discharge can provide a source of very sharp spectral lines and quickly realised 'that could be a very robust and rugged source'. He also considered electrodeless discharge lamps of the type he and a colleague, Norman Ham, subsequently developed for Raman spectroscopy (32), but soon realised that hollow-cathode lamps offered a much wider coverage of elements. In January 1953 Alan, together with John Shelton and the glass instrument makers, George Jones and Frank Williams, set out to construct hollow-cathode lamps which used a closed gas-circulating system, of the type described by Tolansky, in which the rare gas was pumped continuously through traps to remove molecular impurities liberated by the discharge from the cathode and from the walls of the tube. This system involved a rack of elaborate gas handling and pumping gear and was not very convenient. During a visit to the USA in mid-1953, Alan reported back to John Shelton on the work of Dieke and Crosswhite,  who were using compact sealed-off hollow-cathode lamps in which the gaseous impurities were removed by a 'getter' of activated uranium. Alan and John Shelton then abandoned the gas circulating system and during August-September 1953 began the development of sealed-off hollow-cathode lamps for all the elements that could be determined by atomic absorption, using zirconium getters as suggested by Alan's section leader, Lloyd Rees. This was a daunting, exhaustive task, which took several years to accomplish. The first satisfactory sealed-off hollow-cathode lamps were constructed and tested during the period December 1953 to January 1954.
At this stage Alan had arrived at a satisfactory method for making the atomic absorption measurements, which was to become the generally accepted method, and at an experimental arrangement that had all the essential components of a modern commercial atomic absorption spectrophotometer: a sealed-off hollow-cathode lamp as source, a flame atomizer as absorber, and a 'chopper' and synchronously tuned amplifier. A critical factor in Alan's successful development of the atomic absorption method was his appreciation of the necessity for a modulated light source and a synchronously tuned amplifier system to discriminate between the emission of the source and that of the luminous flame absorber.
A provisional patent application was lodged on 17 November 1953. As soon as the final patent specification was filed, on 21 October 1954,  Alan submitted his landmark paper 'The application of atomic absorption spectra to chemical analysis' to Spectrochimica Acta (29). This was published in early 1955, virtually at the same time ** as a paper by C.T.J. Alkemade and J.M.W. Milatz,  who had arrived independently at the concept of analytical atomic absorption spectroscopy. The latter authors did not pursue their work further, possibly because they regarded the method merely as one for determining 'all metals that are usually to be determined in flame photometry'.
Alan's original paper on atomic absorption (29) is quite remarkable. In addition to proposing the atomic absorption method and discussing the various factors governing the relationship between atomic absorption and atomic concentration, he also proposed the details of the atomic absorption instrumentation that are essentially those in use today and he proposed or suggested several applications and developments of the atomic absorption method that were to keep teams of scientists, both at CSIRO and in other laboratories, occupied for the next twenty to thirty years. These included applications of atomic absorption to absolute chemical analysis, that is, analysis without the requirement of calibrating standards of known composition; applications to the determination of relative oscillator strengths of atomic resonance lines; applications to isotopic analysis; and the use of a furnace for vaporizing samples in atomic absorption spectroscopy.
During May to July 1953 Alan visited laboratories in England and the USA and discussed the possible commercial exploitation of atomic absorption with a number of instrument manufacturers. The only person to show any enthusiasm was Dr Alexander Menzies, a physicist and Director of Research for the leading British instrument manufacturer, Hilger and Watts Ltd, with whom Alan had previously had dealings through the manufacture of the BNF Spectrographic Source Unit. CSIRO arrived at a tentative exclusive licence agreement with Hilger and Watts, based on the provisional patent application.
The first public demonstration of a working atomic absorption instrument was in March 1954 at an exhibition of scientific instruments held by the Victorian Division of the (then British) Institute of Physics at the University of Melbourne. The exhibited instrument had all the essential components of a modern commercial atomic absorption instrument, including a sealed-off (copper) hollow-cathode lamp as source, a flame atomizer as absorber, and a 'chopper' and synchronously tuned amplifier to discriminate between the emission of the source and that of the luminous flame. There was also provision for a sodium vapour lamp and viewers were invited to 'dip their (salty) finger' into a beaker of water and this would register a deflection on the strip chart recorder. A photograph of the 'first atomic absorption spectrophotometer' is shown in Figure 2. Alan wrote (68):
The apparent complexity of the instrument was due largely to its being of the double-beam type, which in our early experiments, we regarded as essential because of the poor stability of many of our hollow-cathode lamps. The viewer was possibly further confused by the optical path being in opposite directions on the instrument and on the explanatory diagram. Whatever the reason, the instrument aroused no interest whatsoever during the three days it was on exhibition.
However, when Dr Menzies visited Melbourne shortly afterward to assess its performance, he was sufficiently impressed for his firm to decide to produce, under licence to CSIRO, the first commercial atomic absorption spectrophotometer.
In July 1954 CSIRO entered into an exclusive licence agreement with Hilger and Watts on terms which provided for a 5% royalty on each instrument sold. The exception to this exclusivity was for the case of a potential Australian manufacturer or if Hilger and Watts failed to produce satisfactory instrumentation, in which case the licence would be withdrawn.
In July 1954 Alan's colleague, John Shelton, left on a three-year secondment to the Australian Scientific Liaison Office in London, where it was agreed he would spend one-third of his time spreading the word about atomic absorption, visiting laboratories in England, and keeping in touch with the production of the atomic absorption equipment by Hilger and Watts. Barbara Russell was employed on a fixed-term appointment to replace John, who had expected to see a steady stream of papers from Alan and Barbara, showing the effectiveness of the atomic absorption method in hitherto difficult analyses. However, this did not happen. John writes:10 'Dr Wark had commented to Walsh that as the principle of the method had been established and was to be published, he should leave the "hack work" to others and get back to research'. Although Alan certainly did not regard the practical analytical work as 'hack work', he returned to his fundamental research on the assessment of the possibilities of absolute chemical analysis by atomic absorption and on the testing of the peak atomic absorption method, by comparing measurements on aqueous solutions with the known oscillator strengths of elements over a wide range of oscillator strength values and excitation wavelengths.
In early 1956 Alan sent John Shelton a draft of a paper that included the latest results to test the peak atomic absorption method. Dr Menzies, from Hilger and Watts, arranged for John to lecture on this work at a meeting of the Spectroscopy Group of the Institute of Physics in London in March 1956. The results aroused some interest and led to requests for the lecture to be repeated at the Chemical Inspectorate and Atomic Energy Laboratories at Woolwich and at the research laboratories of the British Aluminium Company. In a letter to Alan in March 1956, John reported: 
Apparently some people got the idea from your paper in Spectrochimica Acta that the method was a scientific curiosity rather than a practical analytical method. Several people have mentioned to me that they had not thought seriously about using atomic absorption until they had heard the lecture, so I feel pleased that some tangible result has come from the lecture.
John later wrote10 that he liked to think that this letter might have had some influence on the change in direction of the atomic absorption project that was soon to take place at CSIRO.
During the period Hilger and Watts held an exclusive license (1953-57), their interaction with Alan and CSIRO was sparse and progress was slowed by technical difficulties, including the development of satisfactory hollow-cathode lamps. A Hilger and Watts progress report for July 1956 stated that enquiries from potential users had 'not so far revealed any demand for a comprehensive instrument capable of dealing with many elements'.  It had therefore been decided to manufacture the atomic absorption equipment as an attachment to the existing Hilger and Watts Uvispek spectrophotometer.  However, there was no provision in the attachment for modulating the light from the source, probably because this would have necessitated major alterations to the Uvispek itself, which had a DC amplifier/detection system, whereas Alan's work had shown that a modulated light source and synchronously tuned amplifier were essential for discriminating between the emission of the source and that of the luminous flame absorber.
In early 1956 John Shelton learned that Dr Menzies was planning to present a paper on the Hilger and Watts unmodulated DC atomic absorption system to the Fifteenth Congress of the International Union of Pure and Applied Chemistry in Lisbon later that year. This triggered an immediate reaction from Lloyd Rees and Alan. John was sent to the Lisbon conference with the intention of making it clear that the Hilger and Watts system was not a genuine atomic absorption instrument and that the full benefits of atomic absorption necessitated a modulated source and synchronously tuned amplifier/detector. The CSIRO paper included some results of the first tests of the peak atomic absorption method by comparing measurements on aqueous solutions with the known oscillator strengths of elements over a wide range of oscillator strength values and excitation wavelengths and also some results on limits of detection. However, the paper did not contain any real analytical results, and John Shelton recalls that it 'deservedly went down like a lead balloon'. In a letter to Alan in September 1956, reporting on the Congress, John wrote: 
Menzies followed immediately afterwards and is now really plugging atomic absorption. His results on brass have given him a really improved outlook…. He took a Uvispek with atomic absorption attachment to Lisbon (some of it by air which no doubt hurt his Scots soul) and gave demonstrations. As a toy it goes quite well, and it's to be hoped that more people will get into doing the actual analytical tests – i.e. on real problems – that, it seems to me, are needed most now. Hack work it may be, but until somebody does some appreciable amount of first class analytical work with the method then the analyst will be shy of it. And they won't be convinced with synthetic samples. They'll want actual samples and a statistical analysis of results.
The Lisbon paper appeared in the congress proceedings (30) and a full version was published in 1957 (31). Neither paper attracted any interest. Alan told how 'the most interest shown at that stage in the technique had been by school children at a working display of the instrument at a Chemex exhibition in Melbourne in May 1956'. 
In mid-1957 Alan was planning an extended visit to Europe and he arranged to visit Hilger and Watts in London with the specific intention of pointing out the deficiencies in their atomic absorption equipment and having them rectified. Alec Moodie recalls that Alan decided to concentrate on just two objectives: to convince Hilger and Watts that they should incorporate a modulated light source with synchronous detection and that they should use a stainless steel, rather than a brass, burner for the flame to avoid obtaining spurious zinc atomic absorption signals. When Alan visited Hilger and Watts, he discovered that the company was experiencing economic difficulties and that two of its leading optical designers, F. Twyman and A. Green, had both left the company. The meeting, comprising a board of executives, went on well into the evening, and Alan still had made no progress, later confessing he had 'failed on both objectives'. After the meeting, when leaving the building, it was dark and raining quite heavily, and the executives drove off, leaving Alan standing in the rain, despondent and looking for the nearest public transport. After a while a limousine pulled up, with Alan dripping with water and expecting to be offered a lift to a station. One of the executives in the back of the limousine wound down the window, called 'Hey Walsh, do you realise there is a tube station down the road?', wound up the window, and drove off. At a gathering soon after, Alan was apparently asked whether it was true that the Hilger and Watts atomic absorption instrument was useless. 'No', he said, 'It is not useless, it would make a great pie warmer'.
In June 1957, Lewis Lewis, who was now at CSIRO Head Office in Melbourne, visited Hilger and Watts at the request of the (then) Deputy Chairman of CSIRO, Sir Frederick White, to put it to them that their efforts so far had not done justice to the potential of the technology. Hilger and Watts agreed to their exclusive licence being revoked and replaced with a non-exclusive licence, provided no other licences were granted on more favourable terms.
Although Hilger and Watts recognised the limitations of the Uvispek attachment, they were obliged to accept that it would take several years to bring out an improved version. Under the circumstances it was decided to continue with the manufacture of the simple attachment until such time as they could offer a sophisticated integrated instrument. In February 1958, Hilger and Watts sold their first atomic absorption instrument, the Model H 909, and from 1961 sales continued at about 30 to 60 instruments a year for several years. 
In 1958, Eric Allan of the Ruakura Soil Research Station at Hamilton, New Zealand, who had assembled his own atomic absorption equipment following discussions with Alan, published the first analytical atomic absorption results, on the determination of trace amounts of magnesium in various agricultural materials.  Shortly afterwards, John David, of the CSIRO Division of Plant Industry in Canberra, using improvized equipment made with Alan's assistance, reported the determination of zinc and other elements in plant-digest solutions.  This was followed by analytical applications in clinical chemistry, by John Willis of the CSIRO Division of Chemical Physics,  and in the mineral processing industry, by Max Amos and co-workers from Conzinc Rio Tinto Australia.  These initial analytical results on the applications of atomic absorption stimulated a steady flow of requests to Alan, mostly from industry, for help in getting atomic absorption into wider use in Australia.
By 1958 there was still no sign of any instrument manufacturer prepared to produce the type of instrument Alan considered necessary. He then decided to embark on 'Operation Backyard' – the construction of equipment in Australia to apply the atomic absorption method – and gave instructions on how to put together a 'do-it-yourself' kit.  Fred Box of CSIRO designed and built the electronics, which included a broadband AC amplifier (commonly called the 'Working Man's Amplifier' or 'WMA') and a power supply to run the hollow-cathode lamps (34). George Jones and later John Sullivan developed and provided the expertise and 'hands-on' skills for producing the hollow-cathode lamps (36), while John Willis worked on the analytical methods for specific analyses. A simple commercially available monochromator, such as a small Zeiss quartz-prism monochromator, was recommended for isolating the atomic resonance lines.
Alan then had to find businesses that were prepared to co-operate in manufacturing components that were not available commercially. He recalled: 
The electronic part of our equipment was perfectly conventional electronics, nothing fancy, so we put out a tender for manufacturing six of our amplifiers and power packs, and a little firm called Techtron put in the lowest bid, so they got the business. They had a staff of five. Then I toured the backyards of Melbourne to find a little machine shop [Stuart R. Skinner] with a staff of eight. Then we tried various glass-blowing people for the lamps, and we found a little firm [Ransley Glass Instruments, later to become Atomic Spectral Lamps] that was willing to try. This was a pure glass-blowing firm, who knew nothing about vacuum technique or electrical discharge in gases, and they had no technical people on their staff at all.
By mid-1962, it was estimated that in excess of thirty of these 'do-it-yourself' kits had been supplied to Australian laboratories and about ten to other parts of the world, including New Zealand and South Africa. Alan recalled:25 'It was certainly enough for Karl Zeiss in Germany to wonder why so many of their monochromators were being sold in Australia'.
In July 1962 Alan and Lloyd Rees arranged a symposium on atomic absorption spectroscopy through the Victorian State Committee of CSIRO, of which John Shelton was secretary. Alan and colleagues John Willis, John Sullivan and John McNeill, and several users of atomic absorption equipment, made presentations. The meeting was attended by about eighty users, potential users and CSIRO staff, including the Chairman of CSIRO, Sir Frederick White, and Lewis Lewis. At the end of the symposium, Geoffrey Frew, Chairman of Techtron Appliances Pty Ltd, declared his intention to manufacture a 'complete' atomic absorption spectrophotometer. Alan recalled how all the main players were present at the same place at the same time and the whole deal was virtually signed and sealed that evening. This announcement by Frew was 'tantamount to announcing the impending birth of an Australian spectroscopic instrument industry' (82).
In early 1964, Techtron produced the first all-Australian atomic absorption instrument, the Model AA-3, which incorporated a 'Sirospec' grating monochromator designed by John McNeill  at CSIRO, diffraction gratings ruled on a ruling engine designed and constructed by Dai Davies and Geoffrey Stiff at CSIRO, and a 'WMA' AC amplifier unit. The AA-3 was exhibited publicly for the first time at the Pittsburgh Conference on Analytical Chemistry in March 1964. A detailed account of the Techtron atomic absorption story has recently been written by Max Amos. 
At this stage Alan and his colleagues at CSIRO had established atomic absorption as a widely used analytical method in Australia, with a small flow-over into New Zealand and South Africa. Alan wrote (68):
While knowledge of the technique spread rapidly throughout Australian industry, there was one memorable exception. I recall the technical director of one of our biggest mining companies phoning CSIRO Head Office in the early 1960s and stating that he had just returned from South Africa where they were using a brand new instrument called the atomic absorption spectrophotometer. He wanted to know whether there was anyone in CSIRO who knew anything about it. Our man in Head Office said he didn't know but he would make enquiries.
In 1965 Max Amos from Sulphide Corporation and John Willis from CSIRO published a joint paper on the use of the high-temperature nitrous oxide-acetylene flame,  which extended the applicability of the atomic absorption method to more than sixty-five elements, including previously recalcitrant refractory elements such as aluminium, vanadium, zirconium and beryllium. From that stage onwards there was a dramatic increase in interest in the atomic absorption method and it rapidly gained wide acceptance. Alan recalled:30
The real winner in Australia, of course, was the mining boom and its timing was a real fluke. At the very time when we suddenly wanted tens of millions of analyses there was a technique waiting to do it. It's incredible that it happened like that. I don't think you could name a single mining company that didn't come here [to our laboratory].
In August 1965, Techtron Appliances Pty Ltd merged with Atomic Spectral Lamps Pty Ltd to form Techtron Pty Ltd, which manufactured the Model AA-4 with a synchronously tuned amplifier and a nitrous oxide-acelylene burner. This was followed by a period of rapid growth, with staff increasing to around 200 in 1966. Geoffrey Frew was obliged to move premises and decided to build a new factory. He tells25 how Alan accompanied him to the Oakleigh branch of the Commonwealth Trading Bank 'to give weight to our plans to build a modern factory for the manufacture of scientific instruments for export'. Frew was granted the loan and in March 1967 the company moved into the new factory, in Mulgrave on the outskirts of Melbourne. In October 1967, Techtron Pty Ltd was approached by Varian Associates, a successful instrument manufacturing company in Palo Alto, California, with an offer of acquisition, first a 50.5% holding and progressing to 100% over five years.33 The merger, to form Varian Techtron Pty Ltd, brought 'great strengths to the company in the way of manufacturing techniques, financial support, and perhaps most importantly, a world-wide distribution network for its products'.  This was followed by further rapid growth, with sales increasing at an average of 30% a year for the next six years and staff growing to 630 by 1972. The company continues today as Varian Australia Pty Ltd, and is the second largest manufacturer of atomic absorption equipment, exporting more than two-thirds of its output.
In 1970, Geoffrey Frew donated a substantial sum to the Australian Academy of Science 'in recognition of the successful commercial development of atomic absorption spectrochemical analysis, which had been originated by Dr A. Walsh of the CSIRO Division of Chemical Physics in 1954'. The Geoffrey Frew Fellowships enable distinguished scientists from abroad to travel to Australia to participate in the Australian Spectroscopy Conferences and to visit scientific centres around the country. Recipients have included Nobel Laureates A.L. Schawlow, G. Porter, G. Herzberg, C. Cohen-Tannoudji and J. Polanyi.
From 1958 Alan made regular visits to the USA, conducting lecture tours and reporting recent atomic absorption results from Australia and New Zealand. The analytical spectroscopists and major American instrument manufacturers remained sceptical of the value of the method. After papers he had presented at the Louisiana State University Symposium on Analytical Chemistry in January 1958, Alan reported that one spectroscopist, Jim Robinson from Esso Research, Baton Rouge, was enthusiastic about the potential of the atomic absorption method. In mid-1958 Robinson obtained approval to start some atomic absorption work, using equipment based on a Perkin-Elmer Model 13 infrared-ultraviolet spectrometer.  During his 1958 visit, Alan also visited the Perkin-Elmer Corporation, with whom he had previously had dealings in regard to the double-pass monochromator. A Perkin-Elmer representative indicated to him that the company would be 'seriously interested in becoming a licensed manufacturer of atomic absorption equipment if it could be shown capable of determining calcium in blood serum' (82). Alan's colleague, Alec Moodie, recalls how Alan's laboratory 'soon became littered with hospital samples that were laden with pathogens and which had to be treated rather cautiously'.
Alan realised he needed the support of an experienced chemist and asked a colleague, John Willis, who had been working on infrared spectroscopy, if he could look into the calcium problem. The determination of calcium in blood serum turned out to be one of the most difficult first problems John could have tackled. After a while he decided to tackle magnesium in blood instead, which turned out to be relatively straight-forward at a time when the available (wet) methods were so difficult and laborious that such a determination was scarcely attempted. Shortly afterwards, John was able to extend the atomic absorption method to the rapid determination of sodium, potassium and calcium in body fluids. 
In March 1959 John Willis submitted an interim report of his work on calcium and magnesium in blood serum to Perkin-Elmer. Meanwhile, Perkin-Elmer had assembled a flame photometer-like atomic absorption instrument and had started some atomic absorption work of their own.  In November 1959 the company was granted a licence from CSIRO to manufacture atomic absorption equipment. In 1960, after extensive internal discussions, Perkin-Elmer established a group, headed by Walter Slavin, to develop an atomic absorption instrument. In 1961 the company began shipments of an improvized atomic absorption instrument, the Model 214, using components that had been developed earlier for the Model 13 infrared-ultraviolet spectrometer.
In May 1961 Walter Slavin and a colleague, Herbert Kahn, submitted a detailed report  to Perkin-Elmer management that included the design of a completely new atomic absorption instrument, the Model 303. The report met with 'massive management resistance', especially in the marketing department.38 In early 1962 it was clear that management would block, or at least continue to delay, the start of the Model 303. So Walter Slavin phoned Alan, who agreed to go to Perkin-Elmer in Norwalk, Connecticut to meet with senior management. Alan recounted:3
After I had described the widespread use of atomic absorption methods in Australia the chairman of the meeting, Chester Nimitz Jr *** (a former submarine commander and son of Chester Nimitz, who commanded the US Pacific Fleet in World War 2) asked rather tersely: 'If this goddamn technique is as good as you say it is, why isn't it being used right here in United States of America?' My reply, which my friends at Perkin-Elmer love to recall, was 'You'll have to face up to it, Chester, the United States is just an underdeveloped country'.
Alan tells how each Christmas thereafter he and his wife Audrey received a card from Chester Nimitz, which invariably included the message: 'Glad to report we are developing nicely!'
In March 1962 Perkin-Elmer began building the Model 303. It was released on the market in April 1963, about the same time as the Techtron AA-2, which used an imported monochromator. By 1965 the Model 303 had already overtaken infrared spectroscopy as Perkin-Elmer's largest product line and had captured the bulk of the atomic absorption market. This prompted Alan to remark (71):
Indeed, whereas previously it [atomic absorption] had been regarded by some reactionaries as the greatest confidence trick since a Sydney taxi-driver sold the Harbour Bridge to an American millionaire, it was now being hailed as the greatest invention since the bed! I presume the truth lies somewhere between these two extremes.
In 1966 Alan's Chief, Lloyd Rees, felt that as result of Perkin-Elmer's highly successful atomic absorption operations the company ought to consider making a serious investment in the Australian scientific instrument business. Walter Slavin and Alan arranged for Lloyd Rees to meet the head of the Perkin-Elmer instrument business in Norwalk. Walter Slavin tells how Alan and he waited outside the office for the whole afternoon for the deal to be negotiated and were then advised that it was felt that it would be monopolistic for Perkin-Elmer to buy Techtron. The decision was that Perkin-Elmer would provide commercial support to Australian science by setting up a factory to manufacture a helium quadrupole mass-spectrometer leak detector developed by Don Swingler at the CSIRO Division of Chemical Physics. Alan and Walter had been seeking support for the commercialization of some of the atomic absorption research originating in Alan's laboratory, such as the resonance detector (46), but this had been rejected because it would have meant Perkin-Elmer going into competition with Techtron 'on their own turf'.
In 1967 Perkin-Elmer purchased land immediately adjacent to the site of the new Techtron factory in Mulgrave, Melbourne, to build a factory to manufacture the helium leak detector. Geoffrey Frew, the Chairman of Techtron, recalled:25 'Although we knew they had no licence to manufacture atomic absorption instruments in Australia, I was very annoyed by the speculations that followed the announcement of their land purchase and setting up business next door'. At the opening of the Perkin-Elmer factory in October 1967, Perkin-Elmer's founder and President, Richard Perkin, was 'very optimistic and predicted that his company would expand steadily in Australia and, as well as making the helium leak detector, would tender for government contracts'.25
Coincidentally, in October 1967 Techtron was approached with the offer of acquisition by Varian Associates. After just six weeks of operation Perkin-Elmer surprisingly closed its plant in Melbourne and in May 1968 the building was purchased by Varian Techtron Pty Ltd as part of a massive expansion of its operations.
During the 1960s and 1970s Alan's research was directed toward the development of novel instruments and techniques to simplify and improve atomic absorption equipment. He was especially keen to develop instruments of the simplest possible design for use in industrial environments where the samples were actually being taken.
In particular, Alan felt it should be possible to replace the monochromator, which was rather fragile, bulky and expensive, with a simpler and more rugged 'non-dispersive' device. In 1965 he and John Sullivan developed the resonance detector (46, 49, 52, 56), which consisted of a vapour cell of the appropriate element to selectively absorb the resonance lines from the source and a photomultiplier to detect the atomic fluorescence emitted by the vapour cell. Alan was known to remark 'Even a chemist can't put it out of alignment'. Such resonance detectors were subsequently tested in an industrial environment for the determination of calcium and magnesium in brown coal by the State Electricity Commission of Victoria and for the determination of nickel and zinc in ore samples.  Later, Alan and a colleague, Peter Larkins, developed the separated (nitrogen-sheathed) flame as a versatile resonance detector for the isolation of atomic resonance lines (70, 72).
The resonance detector was followed by the development of the ingenious technique of 'selective modulation' for isolating atomic resonance lines (48, 56, 59). Radiation from a sharp-line source is passed through a pulsating vapour of absorbing atoms and the resonance lines are detected using a synchronous amplifier tuned to the frequency of modulation of the atomic vapour. Alan predicted that the selective modulation technique should also lead to appreciable sharpening of the profile of the atomic resonance line, giving rise to 'sub-Doppler' linewidths. This was subsequently demonstrated experimentally,  culminating in Alan proudly claiming a bottle of red wine as a result of a long-standing wager with me.
Alan realised that light intensities higher than those available from standard hollow-cathode lamps would be required in applications involving resonance detectors or atomic fluorescence detection. In 1965 he and John Sullivan developed the 'high-intensity' hollow-cathode lamp (45), which employs two discharges, a hollow-cathode discharge to generate an atomic vapour by cathodic sputtering and a high electron-current discharge, isolated from the first, to excite the atoms. The 'Sullivan-Walsh' high-intensity lamp allows the intensity of the resonance lines to be increased up to a hundred-fold without any increase in atom density and hence linewidth. It also has the advantage that most of the light output is usually concentrated in the strongest resonance line. Such lamps are manufactured by Varian Australia Pty Ltd and by the Perkin-Elmer Corporation. A modified version, based on that developed in Alan's laboratory by Martin Lowe,  is manufactured by Photron Pty Ltd in Melbourne.
Alan's original paper (29) also envisaged the possibility of using atomic absorption as a simple method of isotopic analysis. By employing a sharp-line source containing only one isotope of an element, analyses can be performed for that isotope if the 'isotope shift' of the resonance line is larger than or comparable with the width of the line. Successful isotopic analyses have since been realised for elements having relatively large isotope shifts, such as lithium, boron (which was investigated in Alan's laboratory by Hannaford and Lowe ), lead, mercury and uranium.
In his landmark paper (29) Alan proposed that the atomic absorption method should offer the possibility of absolute chemical analysis, that is, analysis without the need for standard samples of known composition. His next two papers (30, 31) went some of the way towards demonstrating absolute analysis, but Alan realised the inadequacy of the flame absorber, in which the atomization was usually far from complete. Professor Boris L'vov in Leningrad accepted the challenge and later established graphite-furnace atomic absorption spectroscopy, involving the complete vaporization of samples. L'vov, in collaboration with Walter Slavin and colleagues at Perkin-Elmer, has now successfully realised absolute atomic absorption analysis for a wide range of elements.  In addition, use of the graphite furnace as an atomizer has increased the sensitivity of atomic absorption analysis by one to two orders of magnitude.
From the time of his original atomic absorption paper (29), Alan was conscious of the limitations of flame methods of atomization, sometimes referring to the flame as 'a hotbed of chemical reactions'. The limitations include incomplete atomization of most elements from their compounds, causing low sensitivity and possible chemical interferences; the necessity of an oxidant, rendering the vacuum ultraviolet (and hence elements like carbon, sulphur and phosphorus) inaccessible; and the need for prior dissolution of the sample. In addition, the presence of various molecular species can introduce background absorption signals, and the need for an explosive gas such as acetylene is undesirable in certain locations such as hospitals.
In a paper published in 1959, Alan and Barbara Russell reported (33) that a hollow-cathode discharge can provide a simple and convenient means of generating an atomic vapour of essentially any solid element. Energetic rare-gas ions formed in the hollow-cathode discharge bombard the surface of the cathode and eject atoms to produce an atomic vapour. Thus this process of cathodic sputtering provides a method in which metals and alloys can be atomized directly without prior dissolution. Furthermore, the method should, in principle, not be subject to any of the above limitations of the flame. In June 1967 Alan employed me to look further into the cathodic sputtering method of atomization. In 1973, together with David Gough, we reported the first results, on the determination of a range of elements in solid samples of iron-base alloys (63, 64). Alan wrote (68):
I would not expect the scientific instrument manufacturers to be greatly interested in the simple sputtering cell… I would, however, like to think that some of them are musing on possible ways of embellishment to ensure that any commercial version will have an impressive price tag.
Indeed, it took more than another decade before an atomic absorption sputtering system was manufactured, by Analyte Corporation, USA, in 1988.
In his final Keynote Address,  to the Pittsburgh Conference in 1990, Alan stated that his own work on atomic absorption 'originated in a laboratory devoted primarily to basic, curiosity-oriented research and finished in applied research of tremendous economic value'. He went on to say that one of his colleagues had 'taken the return journey'. I had worked with Alan on attempts to develop methods of atomization based on cathodic sputtering that were largely unsuccessful for routine analysis. However, my colleagues and I then showed that the atomic vapours produced by cathodic sputtering can provide a surprisingly good environment for conducting a variety of fundamental laser spectroscopic experiments. These included time-resolved fluorescence measurements of atomic lifetimes and their application to the determination of solar elemental abundances from the Fraunhofer absorption lines; coherence spectroscopy including quantum beats in excited or ground atomic states; and high-resolution Doppler-free laser saturation spectroscopy. Thus cathodic sputtering has permitted high-resolution and time-resolved laser spectroscopic techniques to be readily extended to a very wide range of atomic systems (81). Alan seemed particularly excited by this work, not only because of its potential as a universal method for determining atomic lifetimes, and hence absolute oscillator strengths, for atomic absorption spectroscopy, but also because the Fraunhofer absorption lines are 'just atomic absorption'. Alan concluded his Pittsburgh address with 'my experiences over forty years with CSIRO have convinced me that the doors between fundamental and applied research should remain open'.
Atomic absorption has provided a quick, easy, accurate and highly sensitive means of determining the concentrations of over sixty-five of the elements. The method has found important application world-wide in areas as diverse as medicine, agriculture, mineral exploration, metallurgy, food analysis, biochemistry and environmental monitoring. It has been described as 'the most significant advance in chemical analysis' in the twentieth century. 
By the time the original patents of atomic absorption had expired around 1969, twenty licences had been issued, and there were also several manufacturers in countries such as Japan in which patents had not been sought. During 1963-67 sales of atomic absorption instruments experienced exponential-like growth. By 1969 there were more than 10,000 atomic absorption spectrophotometers in use in hospitals, factories and laboratories around the world, and by 1977 this number had grown to around 40,000. Alan tells how a slight decrease in the rate of world sales around 1968 came as a relief to his colleague, John Willis, who 'feared that if sales continued to increase at the same rate as 1963-68, then by the turn of the century the whole surface of the Earth would be covered by atomic absorption spectrophotometers'.  This did not come about!
The current world market for atomic absorption instruments is around $A300 million a year. Varian Australia Pty Ltd in Melbourne, with a staff of around 400 and a similar number outside the company engaged in contract work, has the second largest share of the market, after the Perkin-Elmer Corporation, while GBC Scientific Equipment Pty Ltd in Melbourne, with a staff of around 180, is the third largest. In addition, Photron Pty Ltd in Melbourne manufactures hollow-cathode lamps and high-intensity hollow-cathode lamps for atomic absorption. The commercialization of the atomic absorption spectrophotometer essentially led to the birth of the scientific instrument industry in Australia.
In 1968, A.W. Brown, a scientist with postgraduate qualifications in business administration, was recruited by John Shelton at CSIRO Head Office to conduct a detailed cost-benefit analysis of the atomic absorption project. This study  conservatively assessed the value of the net benefits to the Australian economy at around $22 million (in 1968 Australian dollars), compared with $1.3 million originally spent on the research. (Later estimates gave the accumulated benefit to Australia by the year 1977 as in excess of $200 million, including overseas royalties, the setting up of new industry, and the productivity increases in a wide range of enterprises.) Much to the surprise of many, Brown found that the major benefits to the economy were not through the manufacture of atomic absorption equipment in Australia but rather through benefits to the user, that is, benefits associated with productivity gains, especially the ability to perform large numbers of assays very rapidly and with a high order of accuracy. This component far outweighed the benefits of manufacture. Royalty income was miniscule by comparison.
Mr Barry Jones, a former Minister for Science in the Australian Government, recently remarked:  'I don't know there is a single significant laboratory anywhere in the world that doesn't have an atomic absorption spectrophotometer. The tragedy is, of course, as with so many other of our ideas with something that really began here, licensing rights were sold off to other countries and the result is that only a small proportion of the actual machines were manufactured in Australia after a while.' This is a popularly held view among journalists, politicians and academics. During the period 1954 to 1962, Australia did not have the scientific instrument manufacturing capability to handle the massive expansion that resulted first from the Australian mineral boom of the 1960s and later from the 'environmental boom' and the enormous demand from around the world. There was no company in Australia geared up to cope with such a demand. Moreover, the cost-benefit analysis of Brown showed that the major benefits to Australia's economy lay not in royalties or in manufacture of the instrument but through benefits to the user.
Alan regarded the benefits of atomic absorption to humanity – for example, through its use in hospitals throughout the world – as having 'given him more satisfaction than all the dollars it has earned'. One of his favourite stories ,  concerned a five-year-old boy, who in 1968 had suffered extensive burns while playing with a can of petrol and was undergoing treatment at the Children's Medical Research Foundation in Sydney. For weeks doctors fought to save his life, and finally violent convulsions started for no apparent reason and death appeared imminent. Tests with an atomic absorption spectrophotometer established that the boy had suffered a severe loss of magnesium as a result of the burns. Doctors replaced the lost magnesium, the convulsions ceased, and the boy eventually recovered. A photograph of the boy had a prominent place in Alan's office for the remainder of his time at CSIRO.
When asked about the appropriateness of the development of atomic absorption in Australia, Alan replied: 
Well, of course it was fortunate. We say it was good planning! I think it's a good example of how uncommitted research can finally be more significant than directly applied work. If somebody had said in 1950 that there was going to be a mineral boom in ten years' time which would need new methods of analysis, I'm sure we would have tried to elaborate existing methods, rather than follow a completely new line.
In his final scientific paper (83), written in 1991, Alan concluded:
There are two important lessons to be learned from this account of the development of atomic absorption methods and the difficulties encountered in convincing analysts and scientific instrument manufacturers of their potential.
First, it should be noted that this work originated in a laboratory where scientists were encouraged to study a subject at a basic level and were not expected to have a specific goal for every set of investigations. I think this is a tremendously important point. Increasingly we find young scientists being channelled into increasingly narrow areas of activities aimed only at targets with good prospects of success. They are being given less and less room to manoeuvre. Their work is being largely confined to answering questions, ignoring the many lessons that have shown that much successful research has its origin in asking the right question.
The second lesson is that it is a mistake for the scientist or the inventor to try to sell an invention by scientific and technical arguments rather than by a demonstration of how well it can fulfill the functions it claims to fulfill. The licensee is not interested in how clever the invention is; he or she merely wants to know what benefits the invention affords the designer, manufacturer, and user of the equipment in which it is incorporated.
In November 1976 Alan gave notice of his intention to retire from CSIRO on 5 January 1977, just after his sixtieth birthday and after thirty years of service and fifteen years as Assistant Chief of the Division of Chemical Physics. He had for many years said 'I will have run out of new ideas by the age of sixty and I should make way for a younger person'.
Two weeks after giving notice, Alan received a telex:
I have pleasure in informing you that Her Majesty The Queen has been graciously pleased to approve the recommendation that you be awarded a Royal Medal in recognition of your distinguished contributions to emission and infrared spectroscopy and your origination of the atomic absorption method of quantitative analysis.
Alan had the distinction of being only the fourth Australian scientist to have been awarded a Royal Medal, after Ferdinand von Mueller in 1888 and Nobel Laureates Sir Macfarlane Burnet and Sir John Eccles. In the Silver Jubilee Queen's Birthday Honours List in June 1977, Alan was created a Knight Bachelor for 'his distinguished service to science'. Alan's many other honours included election to Fellowship of the Royal Society of London in 1969 and Foreign Member of the Royal Swedish Academy of Sciences in 1969, being only the second Australian scientist (after Burnet) on whom the latter honour had been bestowed.
On the occasion of Alan's retirement, his Chief of Division, Dr Lloyd Rees, paid the following tribute:51
Alan Walsh did not only invent an analytical instrument called the atomic absorption spectrophotometer – he created a field of scientific work in atomic absorption spectroscopy and initiated and cultivated its application to elemental chemical analysis in areas as disparate as agriculture and chemical industry, and medicine and the mining and metallurgical industries. His contribution to science, industry and human welfare has been enormous. In spite of his great distinction Alan Walsh is a human being – he enjoys life and has never found it necessary to develop eccentricities or affectations.
Alan seemed overwhelmed by all the fuss being made of his retirement, and at a CSIRO dinner in his honour remarked 'I've been to so many farewell dinners recently that I'm beginning to acquire a taste for wine'.
Alan was intending to become a private consultant to industry and sadly had to vacate his office and laboratory at the CSIRO Division of Chemical Physics. He had also accepted an honorary fellowship at Monash University, adjacent to CSIRO, which had awarded him an honorary doctorate in 1970. But first he was going to take a long holiday 'to recharge my batteries and to have some time to renovate the house, do some gardening, swim a little, and improve my golf swing'. After a weekend's golf Alan was known to comment 'golf defies all theory' and 'it is a relief to get back to research where the problems were more amenable to rational analysis'.
Six years later, in June 1982, Alan was elated to learn that he had been invited back to CSIRO as a Senior Research Fellow. In December 1994 the Spectroscopy Wing at the CSIRO Division of Chemical Physics was named the 'Alan Walsh Spectroscopy Laboratory'.
After his visit in 1962, Alan began to visit Perkin-Elmer in Norwalk on a regular basis and participated in some major commercial decisions. These included the decisions for Perkin-Elmer to construct their own hollow-cathode lamps, to manufacture a Zeeman attachment to their atomic absorption equipment to correct for background absorption, and to manufacture the inductively coupled plasma source. 
About a year after his retirement from CSIRO, Alan became a formal consultant to Perkin-Elmer. During 1978-1982 he and his wife, Audrey, spent several Australian winters beside the Bodensee near Überlingen, Germany, where Alan made frequent visits to the Perkin-Elmer plant, the 'Bodensee-werk'. At that time the chief research interest there was the hydride generation technique for atomic absorption, and Alan suggested that it might be very interesting to combine the hydride work with a 'solar-blind' photomultiplier to provide a simple non-dispersive atomic fluorescence spectrometer for the determination of arsenic and selenium. Alan and Walter Slavin convinced the Bodenseewerk engineers to build a prototype and it worked nicely (79). However, Perkin-Elmer marketing decided the market was too small. Some years later the Chinese built a highly successful hydride instrument for a large market. 
In the early 1980s Alan initiated a project to investigate coherent forward scattering as a possible method of spectrochemical analysis (80). Coherent forward scattering had been pioneered in Oxford as a spectroscopic technique in the mid-1960s by George Series,  of whom Alan had long been an admirer. The technique relies on the fact that the light emitted from atoms in the forward direction is phase-coherent and so the intensity is proportional to the square of the number of atoms, thus offering the possibility of higher sensitivity than atomic absorption. The method has the additional advantage that any background radiation scattered from 'particles' in the atomic vapour is not detected. With Alan's help, Perkin-Elmer conducted an extensive development programme on coherent forward scattering and built a system derived from their Zeeman background correction instrument. The project was abandoned when it could not be proved to be commercially attractive.
It is instructive to consider some of the characteristics that may have contributed to Alan's success as a scientist and to his successful development and commercialization of the atomic absorption method of chemical analysis.
First and foremost, his work was characterized by a remarkable simplicity and elegance, a hallmark of many great scientists. Alan himself wrote (83):
My general attitude to research was greatly influenced by the fact that I studied physics at Manchester University. The Physics Department had an illustrious record of major achievements, including Rutherford's development of the nuclear theory of the atom, Bohr's first theory of the origin of atomic spectra, Moseley's law of X-ray spectra, and [Lawrence] Bragg's work on the determination of crystal structures by X-ray crystallography. A feature of all these advances was [that] whilst they were profound they were all very simple. I think by the time I had finished my course at Manchester I took it for granted that the very essence of a significant contribution to physics was a fundamental simplicity.
Sometimes Alan's ideas and schemes were so deceptively simple that they were not always appreciated at first, but as the years went by, they frequently had a habit of becoming important.
Alan was blessed with an extraordinarily creative and fertile mind, forever generating new ideas. One of his colleagues, Peter Larkins, tells  how on one occasion a colleague was attempting to improve the performance of a sputtering system as an atomizer for atomic absorption. Alan made a suggestion which it was estimated would improve the absorption sensitivity by a factor of about two. He came back a while later with another suggestion to give another factor of two. By lunchtime it was estimated he had been back fifteen times! He also displayed great zest and enthusiasm in whatever he tackled, and this seemed to infect those around him. He was a great inspiration to work with. John Willis tells  how in the early days he had been working on the Littrow spectrograph with a modification which Alan and he had hoped would vastly improve its performance. Alan came into the laboratory and asked John how he was getting on. John replied with an air of disappointment that he couldn't do much better than a factor of two. 'Don't despise a factor two, John', he said. 'Three factors of two make a factor of ten!'
Alan had a rare combination of vivid imagination and experimental practicality. He was a wonderfully intuitive scientist, with an enormous grasp of the numerical, carrying numbers and orders of magnitude in his head. He was known to say  'Kelvin could take no pleasure in an equation unless he could feel its weight'. He could make things work so well that some people felt he could 'work magic'. He never pretended to have any unreal powers; he always had his feet firmly on the ground. On one occasion his Chief, Lloyd Rees, wanted to construct a reflecting infrared microscope, which was to be used by John Willis to study the infrared spectra of small protein samples. It was clear to Alan that the numerical aperture of the infrared spectrometer was ill-matched to the microscope. The instrument workshop launched into building the microscope, and the carpenters made a magnificent timber box lined with felt. No detectable signal came through. Alan was heard to say  'Lloyd seems to hold me personally responsible for the laws of optics being what they are'.
Alan grew up in a small family business where he acquired an acute business sense. He once remarked:  'My family was steeped in the traditions of the Lancashire textile industry. I was brought up in the real world – where one went out to make a quid'. Unlike many scientists of his generation, he enjoyed mixing with the captains of industry and spoke their language. It was fascinating to observe him in action at a meeting or in some business negotiation. He would rarely take the lead, preferring to listen politely and assimilate what others had to say. At the end of the meeting he would invariably be asked his opinion, and then would eloquently deliver a definitive statement, which would often end the discussion. He had little time for 'virtuoso performances'. He also had the tenacity and perseverance to see a long and difficult project or business negotiation through to completion. A colleague once said of him:13 'If he had a problem he would gnaw away at the damn thing until it surrendered'.
Many people expected a distinguished scientist like Alan to be a stereotype academic, totally devoted and consumed in his research and with little time for the ‘ordinary' things in life. His Chief, Lloyd Rees, once wrote: 
Throughout his life he [Alan] has been interested in sport, football, cricket, squash and latterly golf. His other sport was drinking red wine. At one stage of his career he indulged in all-year-round early morning swimming in Port Phillip Bay, which, during the winter months, can be sustained by the hardiest individuals only. He has, however, recovered from this aberration and now spends much of his spare time cultivating camellias.
Alan was an avid follower of cricket, especially English cricket, and a great admirer of the inimitable English cricket commentator John Arlott. Alec Moodie recalls how on one occasion Alan proudly showed him a newspaper clipping about his younger brother, Tom, who had opened the batting for the village cricket team in Hoddlesden and on this occasion 'carried his bat' through the innings but failed to 'open his account'. Alan described his brother as 'always cautious, typical of a Lancashire batsman'.
Alan had a remarkable ability to mix with people from all walks of life. His colleague, John Willis, states: 'I have never met anyone who was so immediately at home with people, whether it was politicians, businessmen, distinguished scientists, the younger members of the Division, the tradesmen, or the canteen ladies'. I myself can hardly recall an instance when Alan tried to put another person down or any scientific paper or lecture where he criticized the work of others. It was as though life were too short not to spend it being constructive at all times. He was a wonderful mentor and role model for any young scientist to emulate, forever striving for excellence and providing encouragement and words of wisdom.
Finally, Alan had a wonderful English North Country humour, which was legendary world-wide. During discussions or during gloomier moments in the laboratory he could defuse a situation by coming out with a characteristic 'one-liner' followed by loud, raucous laughter that echoed the length of the passage. Some of his sayings and traditions still live on in the Alan Walsh Spectroscopy Laboratory.
Alan Trumble, a close friend and neighbour of Audrey and Alan, commented, in a eulogy to Alan:57
Alan loved life and people, always showing a very genuine interest in others' activities and concerns for their problems. He was a devoted family man and extremely proud of his boys, Tom and David, absolutely delighted with his daughters-in-law, a doting grandfather to Chevaun, Miriam, Emily and Jack. You could not be close to the Walsh family without appreciating the wonderful affection of a husband for a wife.
In Alan's final years his memory sadly began to fade but he still retained his mischievous sense of humour and good nature to the end. In this memoir we have attempted to capture some of that humour for posterity.
Alan died in Melbourne on 3 August 1998, aged 81, survived by his widow, Audrey, and sons, Tom and David, and their families.
This memoir was originally published in Historical Records of Australian Science, vol.13, no.2, 2000. It was also published in Biographical Memoirs of Fellows of the Royal Society of London, 2000. It was written by Peter Hannaford, Alan Walsh Spectroscopy Laboratory, CSIRO Manufacturing Science and Technology, Clayton, Victoria 3169.
I express my deep gratitude to Sir Alan Walsh for his inspiration, encouragement and friendship over a period extending more than thirty years. I am especially indebted to Lady Walsh for providing background material and numerous anecdotes about Alan; to Alan's cousin, Mrs Kathleen Hoyle, for generously providing the material on Alan's early life in Hoddlesden; to Professor Alec Moodie for providing many of the stories and anecdotes concerning Alan's life at CSIRO; to Mr John Shelton for providing background material on Alan's development of the atomic absorption method; to Mr Walter Slavin for providing material about Alan's association with Perkin-Elmer; and to Dr John Willis for providing numerous sources of information, including the background on Alan's work in molecular spectroscopy, and for preparing the bibliography.
I am especially grateful to Professor Sandy Mathieson, John Shelton, Walter Slavin and John Willis for constructive comments on the manuscript. I also gratefully acknowledge contributions from Mr Max Amos, Mr Peter Beale, Mr David Gough, Dr Norman Ham, Mr Bill Ramsden, Professor Norman Sheppard, Mr John Sullivan, Mr Rodney Teakle and Dr Harold Whitfield. Finally, I wish to thank my colleagues from the Alan Walsh Spectroscopy Laboratory at CSIRO and my wife Kay for permitting me three months of pleasure to indulge in writing this biographical memoir.
The photograph of Alan Walsh was taken in June 1979 by the CSIRO Division of Chemical Physics.
Numbers in brackets refer to the bibliography.
Numbers in square brackets refer to the references.
* In 1949 the (Australian) Council for Scientific and Industrial Research (CSIR) became the Commonwealth Scientific and Industrial Research Organization (CSIRO). In October 1958 the Chemical Physics Section of the CSIRO Division of Industrial Chemistry became the Division of Chemical Physics, with Dr A.L.G. Rees as its foundation chief.
** The receipt date of Walsh's paper (29) is stated as 18 January 1955, which was subsequently amended to 19 November 1954 in an erratum (29). The receipt date of the paper by Alkemande and Milatz was 29 December 1954, and 27 December 1954 for a short Letter to the Editor.
*** Chester Nimitz Jr was Vice-President of the Instrument Division of the Perkin-Elmer Corporation and soon became President and later Chairman of the Board of Perkin-Elmer.
© 2018 Australian Academy of Science