(My name is Doug McCann, I'm interviewing Reginald Ryan in his home at 19A Gardyne Street, Waverley), a suburb of Sydney, NSW. The date is the 8 April, 1999, the topic is the CSIRAC computer)
McCann: Reg, when did you commence employment at the CSIR division of Radiophysics?
Ryan: In March, 1948.
McCann: Could you outline the circumstances that led to you becoming part the Mk I team?
Ryan: Well I had worked at Radiophysics Laboratory for my Engineering student work experience for six months. I worked with Harry Minnett, so I knew a lot of the people in the Radiophysics Division including the Chief, Dr Bowen, and Harry Minnett, Maston Beard and many other people. So when I graduated and was appointed to Radiophysics Division I had some idea what was going on in the Division and the number of projects there, because I had worked in Astronomy at Mount Stromlo, again during a vacation as a student. The first suggestion was that I should work in the Radio Astronomy section of Radiophysics Division and fellow graduate, Ron Styles, was to work on the computer in the CSIRAC team. But another project which was also running in Radiophysics was the Aircraft Navigational Aids and it turned out that they needed somebody there fairly urgently, so Ron Styles was allocated to Navigational Aids which left… needed somebody to go on the CSIRAC computer, so I was allocated there and I joined to work with Maston Beard and Trevor Pearcey, Geoff Hill on that project.
McCann: What aspect of the project did you work on?
Ryan: Maston Beard had been developing the Arithmetic Registers and other power supply parts of the CSIRAC computer but the other major part of the hardware requirements were for a memory and it had already been decided the best sort of memory system was mercury delay line, supersonic mercury delay line memory unit, which the information and all the bits were circulated through this delay line and kept on re-circulating so that the information was stored. So my job was to read the reports and that which we had obtained from overseas laboratories, develop my understanding of this device and the necessary electronics to keep the information circulating. So I did that initially, I had never worked on the mercury delay line before so it was necessary to become familiar and in doing that then, the next step of course was to fabricate the actual devices, deciding on the materials, mercury was the medium for delaying the supersonic sounds which were going along the delay line and getting the quartz crystals which generated the supersonic signals from electrical signals imposed on them and one of the factors was the necessary bandwidth or the timing resolution which was necessary for the mercury delay line so that you could maintain the pulse shapes in the information flow and the associated high gain, high bandwidth amplifiers that were developed for that. So this was my work over 1948 to improve my understanding and to fabricate the necessary devices, get the particular sorts of vacuum tubes. I remember 6AC7s were high gain tubes which were necessary to get the performance in the amplifiers and drivers of the mercury delay lines. So as I said we had to develop the first, I had to develop the first, delay line, get it built and operating, understand it, get the associated electronics.
McCann: How long did it take you to demonstrate the feasibility of a working delay line?
Ryan: I would estimate it must have been six months, so somewhere the second half of 1948 we would have had the understanding and the actual fabrication of the devices under control.
McCann: What materials did you use for the delay lines themselves?
Ryan: Mercury was the medium for the supersonic flow of information, but the mercury takes up contamination fairly easily so we couldn't use steel or conventional tubing and I settled on monel metal as being the best one to give us the lowest contamination. The other question that has just come back to me, was the temperature co-efficient, the delay time would have changed with the temperature and you could correct for this, but since you were going to end up with a bank of 16 tubes if there was any differential temperature going through the bank then you would have had different times so that the pulses from different parts of the memory would have been out of synchronization. As I recall, I think we ended up with an insulated box around the whole bank of tubes so that the temperature was maintained reasonably constant against outside fluctuations.
McCann: It became known as the coffin apparently.
Ryan: The coffin yeah, it looked like…it was a bank…it was about a couple of metres long I think wasn't it. The length of the tubes is given in the paper but it was something in that order. It was quite long and it must have been half a metre or so high.
McCann: Were there any special problems in having the construction…for instance when you were boring the holes in for…. in the metal and so on?
Ryan: I think we bought some tubes all ready. Just monel metal tubing, we couldn't have drilled them, they were too long to drill. I think there were some problems, the supersonic frequency was 10 megahertz and the crystals have to be resonant at that frequency, so they had to be very thin, and then they had to be backed so that they didn't by…a lead slug in behind them. So they were rather fragile so they needed a fair degree of care in assembling the crystals into their mounting and then onto the delay line and then filling with purified mercury. That's right, we must have had a mercury still in a fume cupboard to purify, to evaporate the mercury over to get rid of the contamination which would gradually accumulate. So we had to get very pure mercury and then get it into the tube and seal it off so it would stay pure for a reasonable period of time.
McCann: Were you conscious of the toxicity and did you have special precautions?
Ryan: As I recall, I remember being quite aware of this and I remember, maybe it was only psychological, but I remember sort of feeling skin on my teeth when I was working with the mercury, so I was pretty careful to do all the mercury operations where the mercury was… mostly it was sealed in jars and that, big demijohn jars, I think it would have came in, but we, I would work in this fume cupboard with proper ventilation and mop up or make sure that all the mercury (we couldn't help but spill some mercury around), try and make sure that was all mopped up so that you didn't get mercury vapour in the area. I think that was done in the one big room, but probably up one end there was a fume cupboard which we handled the mercury in the delay lines.
McCann: Were you the person who handled the mercury most of the time?
Ryan: As I recall. The machines, the tubings, was cut and the holders for the crystals were fabricated at the Radiophysics workshop by the turners and fitters down there. The crystals, were as far as I can recall, were fabricated by AWA. They would have been making… they would have had the expertise for grinding, cutting quartz crystals for making frequency control circuits. So they would have all that, but it was my job to design the mercury delay line, mechanical design. To then put them together and to handle the mercury and delay line. I don't remember having technical assistance, but maybe I did, I might have had some of the technician people may have helped me, in that I don't recall it. I certainly would have been doing some of it myself and then as time went by I might have handed it over to them. Same with the electronics associated with the… I would have designed the circuits and then they would have been fabricated in the workshop and I would have tested them and maybe the technicians would have also helped me in the testing of them.
McCann: After demonstrating the first delay line and showed that it worked, it was feasible, what was the next stage in the process of building a working memory unit?
Ryan: Well as I recall, it was to first of all make a bank of 8 tubes, 4x4, and get them all working in together. So, a matter of fabricating the 8 sets of the 8 mercury delay lines, the 8 sets of electronics and the associated gating so that they could be selected, so that you took the words that came out of one line and... That's right, I think yeah, a serial, what say we took…would have had to select the particular time when they came out, access time. The time it took to do that, was that the first one, getting the idea, the principle worked out would have taken about six months or so, so it would have been towards the end of 1948 and we must have got the bunch of 8 tubes going sometime in 1949, I think towards the middle end of 1949 probably by the time we got that all going. One thing which would have delayed us was the unreliability of equipment. I mentioned the oscilloscopes and then the electronics themselves. Valves were not nearly as reliable then as they became later and, of course, and as semi-conductors became eventually. Nobody has much problem with reliability in the 1990's, but in the 1950's, '40's, '50's, it was quite a problem of reliability.
McCann: You mentioned those oscilloscopes came from Britain?
Ryan: Yes, Cossor was the company, they had the major share of the market at that stage and they were valve operated and therefore they ran very hot and the valves and the heat probably made them unreliable. So I recall having, we did have a test equipment laboratory which was supposed to, did do a lot of work in keeping the test equipment going, but they would break down and the quickest way rather than take the oscilloscope down to the test laboratory to have it repaired, I would normally, or frequently anyway, do the repairs myself. Get the oscilloscope going, again so that I could then check out what the problems were in the circuitry of the mercury delay line.
McCann: So, how often would that happen?
Ryan: Well certainly on a weekly basis. Maybe more frequently than that even, they would be breaking down.
McCann: A major problem.
McCann: Can you remember things like the strikes, the power strikes and so on?
Ryan: Oh yes, another factor was that there was a lot of industrial unrest at that stage and we were having black-outs continually for periods of months, or maybe even over a period of a year or two, I can't remember exactly. You would have to look up the…. I remember it came to a big battle on the wharves when Prime Minister Chifley ordered the troops in to unload and load the coal, when was it? Anyway the coal miners were on strike and there were shortages of coal, therefore the power stations were having to stop power production and maybe they were rotating blackouts around the city, I can't remember. But I know power would fail and we couldn't do anything, we couldn't even test the oscilloscopes. I probably would go back, to first of all, keeping myself up to date on the literature in the library which was luckily just next door to the computer room and reading reports and doing drawings probably, doing design work. And then ok, when the power came on you would go back and start up the equipment and proceed to do more development.
McCann: This is a fairly crucial question because we haven't been able to pin it down. But do you remember when the various components of the computer integrated together and the first test program was run?
Ryan: No I don't have a specific date, I didn't keep diaries unfortunately. But from the publications I noticed my article "Mercury Delay Line Memory Unit" published in the proceedings of the IRE Australia, the manuscript was received by the Institution in November 1952 and was actually published, not until, April 1954. So, from reading that article in the publication, it would have taken me quite a while to write up the article and then submit in '52, so I certainly must have been working on the article and presumably had already finished and had the equipment operating, the Mercury Delay Line memory, operating by early '52, I would imagine. But that would have been the complete 16 bank unit. The 8 bank unit would have been earlier so just by extrapolation backwards, one would say that the 8 unit would have been working in certainly in '51, maybe '50/'51 period, is when that would have been coming on line. Whether we did any smaller units, I don't recall. I would be inclined to think that we probably didn't try do anything very much until we had the bank of 8 going because there wasn't a lot of memory anyway even with the 16 bank/16 lines. So the 8… but I can't really give you a…
McCann: Do you remember any visitors visiting your laboratory in those early stages, around about the time the first programs were run?
Ryan: I don't recall the people who would have been there myself.
McCann: It seems to me, according to what you are saying and the literature, it was about 18 months from the time you started to when you were ready to test program.
Ryan: It's hard…I can only go on the actual publication dates, so it may have been much earlier. I don't know. The main thing would have been to keep the equipment operating against this break downs and faults developing and the writing up would therefore possibly have been delayed, so that's consistent with maybe the equipment being working quite a bit earlier than the time the report was actually presented.
McCann: When did the Mk1 actually develop to the stage of carrying out routine operations?
Ryan: I would guess that it must have been in the '52-'53 period.
McCann: Do you remember the Conference - that '51 conference, August '51. The first Computer Conference at all?
Ryan: No I don't have any recollection of that. I remember attending lectures by Professor Jaeger, but I don't recall the conference. That doesn't mean that I didn't go. I might have gone.
McCann: When do you think you moved from the bank of 8 to the 16?
Ryan: I would imagine it would have been in 1952 that we would have been up and running substantially.
McCann: What modifications or additions were made to the original memory unit to increase its capacity?
Ryan: As I mentioned earlier, in getting to understand the operation of the mercury delay line, I realised that the band widths and the timing resolution capability of the mercury delay lines was higher than that of the electronic circuitry which was limited by the vacuum tubes etc. The computer itself was working at about 330kHz rate. The mercury delay lines could cope quite adequately with a 600-700kHz, so that we could then interleave two sets of pulses in the mercury delay line and separate them out so that we could effectively double the memory capacity which was quite an important thing since that was a big limitation on the capabilities of the computer, and once we realised that it only meant fairly simple electronic development. At the end the mercury delay line stayed the same of course, and the driving, receiving and transmitting circuits stayed the same. It was simply a matter of adding on fairly simple electronics to institute that and the original paper which I submitted in November 1952 does describe the interleaving system, and the footnote at the end of that paper when it was published in 1954, says that the inter-space system, in the text it says "such an inter-space system is presently undergoing tests". That would have been in the November 1952 period and the footnote says that the system has now been operating satisfactorily for some months. So presumably, the inter-spacing would have come into operation late 1953-54.
McCann: Now concurrently with the development of the mercury delay lines you also developed some test equipment. Can you tell us when did you begin developing the test equipment. What was it used for?
Ryan: I think it was probably fairly late in the piece. The sequence would have been, one mercury delay line, 8 mercury delay lines, 16 mercury delay lines, interleaving. So that happened in 1953-'54 and I think I would then have, probably in 1954, moved on to developing this Rack. One Standard 19" Rack with an oscilloscope, specialized (I think it was two oscilloscopes) specialized registers and connector switches so that this monitor rack could be wheeled along to the memory unit, plugged in to test, check in to see what the operation, whether there was any faults, and then to the other parts of the computer to check the operation of those registers. So that would have been developed and brought into operation I imagine round 1954, maybe it went to 1955, but I think that it probably would have been in 1954 more than 1955.
McCann: Yes more sort of early 1954.
Ryan: Yes probably I said 1953 or 1954 the inter-spacing operating, so I think it would have been brought into operation say mid 1954 would have been the timing.
McCann: How and when did you come up with the idea of a permanent high speed memory store?
Ryan: This was probably an area where we were operating as a team, the mathematical computer people, Trevor Pearcey and Geoff Hill and the electronics people, Maston Beard, and myself. And in this sort of interchange, I became cognisant of the problems of the limited memory which could be handled by the mercury delay line. The limitations on the number of, the amount of storage which you could cope with there, and the realisation that many of the, much of the information which was in the mercury delay line memory never changed, it stayed the same, sort of fixed sub-routines. And my immediate… what I came up with there was, well in that case you don't have to, you can come up with a much simpler memory, a much higher capacity memory, in the permanent high speed cathode ray tube type of memory. So the idea of that may have come a bit from the alternative to the mercury delay line memory which was being developed in the United Kingdom by people coming out of TRE, but rather than doing it as an erasable cathode ray tube memory, I said ok if we are not going to be changing the words, a lot of words wouldn't be changed very much, let's put it all into a fixed form. So I proposed this and this was published in the transactions of the Institution of Radio Engineers in September 1954. I think Trevor Pearcey and Geoff Hill were encouraging on that because they realised they were the ones who were stressing the need for much more memory capacity and they saw this as a way of getting memory capacity in a more convenient form.
McCann: Could I just clarify that bit - was that an original idea on your part or did you say that you had some influence from overseas ideas?
Ryan: The idea was mine. I think, I would have, I was aware of this erasable cathode ray storage which was being developed and I think was brought into operation, probably in their Mark II Digital Computer to effectively store, it's a bit like the television camera tubes to charge up little spots on the cathode ray tube and then to read them all. Now I said ok, you do that with the cathode ray tube, which has the advantage, you can scan all over, you can access the large number of spots, essentially bit storage points with the cathode ray tube. But ok, that's a much more complicated system to maintain the erasability. How about we do a much simpler one, where we just have it fixed, and it is not erasable, not changeable, but that was the original idea which I developed as a proposal. I don't think that was proposed anywhere else. As far as I know it was never actually tried out either.
McCann: When did you become aware of read only memory developing?
Ryan: Well not for twenty or so years afterwards. Not until the semi-conductors came along and personal computers, where there was a need for read only memories for example for the Bios systems, when you switch a computer on, it automatically goes to read only memory. Maybe they weren't even developed before that but again I suppose we go back, you have to look at the early digital computers, the main frames, how did they start up. Presumably, somebody at some stage may have come up with this idea. Then of course, when the PCs came along, that idea was taken over by the PC's. It was a very laborious procedure to start up a computer when Trevor and Geoff Hill were doing the business of reading tapes and things in. I don't imagine it would have been developed until the late '60's, maybe into the '70's, before it would have been applied even in the large computers, IBM and people who developed.
McCann: It's something we need to do a bit more research on, I think.
Ryan: Yes, I've had the idea. Maybe I should go and have a look in the literature, but it's such a volume of this literature, that I have never got around to it.
McCann: Now when did you actually leave the Mk1 Project?
Ryan: From the biography, in that IRE article which I mentioned on the permanent memory, it must have been around the…it says there "he's recently transferred to work on semi-conductor devices", so that would have been in, well I would have written that, it must be early 1954. Does that agree with what I was saying earlier. This monitor memory. Quite likely what happened was that I didn't drop the computer on one day and start the semi-conductor another day. It was probably a transitional period of time. So I imagine it was in the 1954 period, but maybe not till 1955.
McCann: Now this is a more general question. What were the circumstances that led Radiophysics Management to end the project. Do you have any knowledge of that?
Ryan: Doesn't come back to me particularly. As far as I am personally concerned, I recall that with the permanent high speed store which was published in 1954, I understood that this would never be built, that it was not a project that was going ahead. Because if it had been going ahead I probably wouldn't have published the paper then. I would have, as I did with the delay line, I would have developed the hardware and the electronics and got it operating and then I probably would have published the paper. But it became evident that the computer section was not going to develop, so certainly anyway this permanent high speed store was not going to be implemented, so I understood it to be finishing then. I don't recall being involved in, or even knowing much about the decision making. I think it was probably above my level. Maston Beard was not very communicative about that sort of thing, so he probably would not have told me about it, and Trevor Pearcey who was more involved, I don't recall him mentioning it particularly. Although he was particularly interested in this permanent high speed memory. He was working on how this would be used in the programming of the software side of it, him and Geoff Hill and we did publish that paper with our three names on it at the same time as I published this one on the principle of it.
McCann: Really the circumstances behind the actual axing of the project - or the termination of the project - it's been suggested, we were talking about perhaps the analogue, the competition with analogue computers?
Ryan: Well maybe in hindsight I'm saying this, but I think there was a contest which went on, not only in the '50's, but well into the '70's or even the '80's between analogue computers and digital computers. The analogue computers were, the hardware was much more available, the mechanical and that early on and the future of it could be seen more clearly. The digital computers were really very rudimentary and therefore it was hard for people in the '50's to realise how important, how much had been developed in computers. Yes, well, the digital computers were still rudimentary and were being developed so I think it was much easier, it was, it would have been a big gamble by Radiophysics Division to invest more resources into the digital computers and while Joe Pawsey, the Assistant Chief, I think would have been more favourably inclined to keeping it going, he was more likely to have foreseen the possibility of development of digital computers. He was orientated, his major interest was in the Radio Astronomy and Bowen on advice, or maybe from his own thinking about things, his own appreciation, and his overseas trips and that, he would have felt that he could see that the digital computing would have required a large amount of resources from Radiophysics to get it to a viable situation and he made the decision. Of course, I think he was more in favour, he was more interested in the Aircraft Navigational Aids, in the Radio Astronomy and in the Rain Making, which were the three other sort of competing projects and which had grown fairly naturally out of the war time work of Radiophysics Laboratory. The digital computers did grow out of it, because they used a lot of electronic techniques, but the concept of the digital computer itself was relatively unfamiliar, I think relatively new and not properly understood, so it would have been a…Trevor Pearcey would have had a difficult time convincing the people of the importance of it, which became evident in hindsight. I'm not sure that it.. it may not have even been so evident to Trevor, but at least he would have had a much better appreciation of the possibilities.
McCann: And I guess people like David Myers?
Ryan: There was quite a lot of local opposition, and overseas, David Myers was the local person who was interested in analogue computers and was developing them and he would have been in contact and advising, giving his opinion anyway to Bowen, and there would have been overseas people like Hartree in the UK that I seem to recall was also biased, probably was not working with analogue computers or knew the analogue computers and felt more comfortable with the prospects of their development rather than the digital computers.
McCann: Of course the differential analyzer was also a pretty important piece of equipment….
Ryan: When I say analogue computers, I sort of include the differential analyzers in that as well.
McCann: I just want to ask you a bit more some general questions about personalities involved. In general what were the qualities, attitudes, personalities and vices of the Radiophysics Management - I'm talking about Edward Bowen and Joe Pawsey.
Ryan: Yes, well, Taffy Bowen as we call him, he was Welsh I think originally. He was a very strong personality, he was a very effective Chief of Division and he had quite a big body of work under his belt when he came to Australia. He was quite important in the development of the Airborne Radar in the United Kingdom as I recall and with H2S radar which was very important in the night fighting over Germany and he also was a leading member of the Tizard Committee, or Tizard group of people, who went from the United Kingdom to the United States in September, 1940. One of the most important lots of information they took, this was when United States came into the war, was the Magnetron which Oliphant and people had developed at Manchester I think it was and Bowen was very well connected, he knew all the people in the UK scientific establishment, Tizard etc. and in the United States, Vanevar Bush and many people who would have been working or who would have later been working at the MIT Radiation Laboratory and also the financial foundations which he tapped on later. So he had the political nouse and connections and was also a very forceful personality. He was more applied scientist I think than Pawsey who was more the pure scientist and Pawsey was not so politically aware and not such a forceful personality, but I think he had a deeper understanding of the scientific aspects. As the whole of Radiophysics Laboratory I think CSIRO they made a very good team working together, they fitted together and Pawsey would have been the prime protagonist of the Radio Astronomy side and to a lesser extent of the Computing side whereas Bowen was strongly onto the Navigational aids and the rain making. The other couple of people who I think were important in the administration were Arthur Higgs, who was the sort of scientific secretary of the division and Sally Atkinson who was Taffy Bowen's personal private secretary. She knew, and between the two of them you could always find out everything about the division. Administration was generally very good, I don't remember any personnel officers, although I think Arthur Higgs acted as a personnel officer in records. Not the way we have our Human Resources people these days, but everybody sort of knew everybody and knew their strengths and their weaknesses and I think Taffy and his mentor Fred White have to be given considerable accolades for development and seeing them. It could have been that Radiophysics was just closed down and they saw the potential and recruited people like John Bolton who was recruited for Radio Astronomy, and he backed people like Ruby Payne-Scott and Ron Bracewell and Jack Warner, Brian Cooper and many other people, he recruited people from overseas, from the UK, as well as the core of Australians that were brought in during wartime to work on the radar systems.
McCann: How did Joe Pawsey fit into the scheme of things?
Ryan: Bowen was chief of division but he was overseas a lot. I think Pawsey was sort of Deputy Assistant, but I think all the important, the major decisions would have been taken by Bowen and Pawsey would have put his oar in and Bowen I'm sure listened to him and would have taken a lot of notice of his scientific aptitudes and interests.
McCann: What were their attitudes towards the Mk I project? Did it change much over time for instance - were they supportive and responsive to the project's needs?
Ryan: As far as I remember, they would have been very supportive in the early stages, in the '40s, before I arrived. In '46/'47 period, when Trevor Pearcey was developing the ideas and then Maston Beard I think was seconded. He was if I remember rightly, working on navigational aids and he would have been shifted by Bowen with Bowen's approval to work with Pearcey in the '47 period and then he obviously he recruited Geoff Hill in to help Trevor Pearcey and recruited me into working with Maston Beard on the electronics side and the fabrication as far as I remember of the whole computer… Maston Beard's responsibility, the registers, the arithmetic registers and power supplies all that and then mine on the mercury delay line memory . We got good support, workshop effort was allocated to it and the money was made available for buying materials and all the rest of it, it was probably spent for that time quite a lot of money on it, a lot of resources were put into it. It's difficult to remember, I know that by '54 when I wrote that high speed storage paper that the project was not going to go ahead, so I can only extrapolate backwards and say that I think possibly because of the reliability problems, it didn't show results like say Radio Astronomy was showing results; the Crab Nebula being picked up by Stanley and Bolton and Ruby Payne-Scott I think was doing things too and they would publish papers. Trevor Pearcey, myself and Maston were publishing papers but more on equipment, it wasn't anything startling and new, something which was attracting attention, and we didn't get papers into Nature and places like that. So maybe it was perceived, Bowen was perceiving it, as a bit of a disappointment by that stage. Whereas Pawsey I think might have backed him, can see that ok it wasn't the sort of area where you are going to get spectacular results quickly. I don't know, you would have to compare what people were doing overseas. Again as I recall it, I mentioned earlier, this cathode ray erasable memory being developed in the UK and there were others, but development of digital computers was a more, a longer period. With the Radio Astronomy, you could take a radar aerial, you can set it up in Dover Heights, you can pick up your signals from Crab Nebula and you can publish your paper and that meant you could see where to go to improve that work and you could come out with results fairly quickly year after year, whereas ours we would probably only publish, between the four of us, only a few papers in that period of three or four years, until Bowen made the decision. And there may have been strong opposition, the advice he was getting might have been adverse. The other thing was that perhaps, when he saw the size of the CSIRAC computer and the resources needed in it, he could see that if you were going to develop real number crunching capacity you were looking at very large mainframe massive computers that he could see that this would require large resources and he with Fred White would have seen that he probably would have had to cut back on these other projects, so he had to make a decision and he said no we're going to cut off that development.
McCann: One question I forgot to ask you before was the order in which people came onto the team. There was obviously Trevor first, then Maston and then probably yourself. Was that about a year into the planning?
Ryan: I couldn't be sure, Geoff Hill was there very early on, whether he was there before or after me, I couldn't say. The CSIRO records would show. It was Trevor first of all would start thinking about it and then Maston Beard would have come onto it in early 1947 or interacted development, and then I started March '48 and Geoff Hill may have been there before. I think he was a fairly new graduate, so it's quite likely that he came on about the same time, as we were both recruited as new graduates. A lot of the people at Radiophysics were all recruited during the war, early on in the war I think in London. Brian Cooper says when he came in about 1940 or something, when the Brits first started. Bowen came out and started developing it and then I think they would not have recruited all that much during the war, with maybe one or two, Don Yabsley probably came in a bit later in the war. They were recruiting more people I think at that time in the '48 period when he graduated.
McCann: You were a fairly young group of people at that stage I guess.
Ryan: Yes what was I, 23, and Geoff would have been quite young too. Maston was about 5 years older, he would have been in late 20s and I'm not sure about Trevor, he may have been a little bit older again, about 30.
McCann: A fairly unique situation really, where perhaps in an established institution, where you have older figures running it, whereas you tended to come in as a pretty young team. Maybe that had something to do with the progress you did make. Was there a lot of energy there did you feel?
Ryan: I think so, yes, it was a very enthusiastic division of people and I think in some ways people, even during the war, were looking at what are we going to do with all this know-how and this stuff we have developed when peace breaks out, an awful lot of things we can do and ok when the opportunity came available, they were raring to go and a lot of things actually did get done and I think it was an expanding, Radiophysics expanded, recruited people, did more and more work in radio astronomy and navigational aids and so on, worked on later on. So from the '47 period through till into the '60s, I left in '62, so I can't say about that, I think with it expanding so much it became obvious that we would need a bigger building and before I left in '62 there were plans to move the division to a new site, which they eventually did up at North Ryde, and I think that may have been in my thinking in leaving Radiophysics, so it was sort of outgrowing it, it was very much an expanding thing, a sort of golden age starting in this area.
McCann: Just talking about the project team itself, for instance, Trevor Pearcey, Maston Beard and others. Could you describe your relationship and interaction with those figures.
Ryan: It was a very friendly, very co-operative, very much a scientific, technical team approach, not only the professionals that you just mentioned, but the draughtsman, Frank Tonking was one that I think was quite important in the project in getting things done. And probably I think in handling a workshop, I would do the design and Frank would do the drawings, and then he would see it through the workshop, make sure that it was made to proper specifications and things like that and the workshop people themselves, in the fabrication, they would be coming up and talking to us up there and we would be going down to the workshop, so it was very much a team thing. I worked directly to Maston Beard who was a somewhat introverted character as many of us scientists were. He was a sort of a man of few words as we say, but I think he was quite fair in dealing with us. On the personnel basis I felt that I was looked after properly in getting backing and getting promotions as years went by and in helping me in the work. Although I think Trevor and Geoff Hill may have worked more closely than I worked with Maston Beard and that maybe reflected our personalities, maybe we were both a bit introverted and worked on our own, whereas Geoff was rather more extroverted type I think and Trevor would have been somewhere in between I imagine. But there was quite a bit of interaction between Maston and Trevor obviously and later on between Maston and myself and Trevor and Geoff on the hardware and software side, and again as I said earlier, with the draughtsman Frank Tonking and at the technician level.
McCann: You did mention as far as the programming side versus the engineering side when you were communicating with Trevor. Did you fully understand what Trevor was on about with the programming side of things and also Geoff Hill, was he helpful in that regard?
Ryan: No, I think I naturally had an interest in exactly how this hardware was going to be used, how do these commands operate in the computer, how did you develop the software and I remember trying to get to understand this, talking probably mainly to Geoff Hill and to some extent to Trevor Pearcey and as I recall it, rightly/wrongly, they tended to talk above the heads of non-mathematicians. (first side tape ended)
McCann: You were talking about the communication between you and Trevor Pearcey and Geoff Hill?
Ryan: Yes, as I said they tended to be a bit uncommunicative about exactly how the software operated and I did make an effort to try and understand what was going on. I think I would have liked to do a little bit myself, to be able to run it, do a program, but as far as I recall, I didn't find them very helpful in this way, at this point. It was certainly my lack of understanding of the mathematical software side but I think they also contributed to the problem by not making an effort to communicate the concepts and ideas of software, which was a pity, I think it would have helped a lot. Let's qualify that, I am not saying I didn't have any comprehension, I think there was considerable interaction between us and that was evidenced by the development of that high speed memory because we were involved. I knew enough to know there was a very severe limitation on what you could do with this sort of computer, in the amount of memory space that you had, and therefore I thought about how can we get more memory and I said ok. This is one way we can get more effective memory for the information, the sub-routines which just didn't change, so we had interaction. I think I was at that time a bit disappointed that we didn't get more interaction, that I didn't understand more, but maybe it worked both ways, maybe they wanted to know more about what could be done with the electronics side of it, and maybe it was part of the compartmentalization of science which has been going on ever since of course and it's very hard to get the two sides to talk.
McCann: And a disciplinary thing as well because mathematicians are notorious for this apparently, aren't they?
Ryan: Yes, well it was not just in the programming computer side, that mathematicians in general tended to be a very esoteric, exotic bunch. I did three years of mathematics at Sydney University, but I still never felt that I had any large capacity as a mathematician which was a pity. I think I felt it was a lack in my other work, for example, in understanding the mercury delay line, the physics in the mercury delay line, I could have done with a better mathematical capacity there. So I think it was an inter-disciplinary thing as you say, mathematicians against engineers or electronics people.
McCann: How about your interaction with other members of staff, for instance, Brian Cooper?
Ryan: Well I found Brian very friendly, easy going type. He was a bit laconic, but he had a lot more words say than Maston Beard had. He was very tall I remember and he was more or less, I don't know that he ever became, he wasn't so much a member of the team of CSIRAC, he was really one… I remember him physically, he was in a corner of the room working on his magnetic drum, it was coupled in. He was more senior than I and Geoff Hill were, so he probably had an office elsewhere, whereas we just had a desk, I had a desk in this large room. So I don't, he was again, while Trevor and Maston and myself and Geoff were on it from the period 1946-1947 through to 1954 or so, Brian came in, I'm not sure at what stage, but he came in a couple of years after, two or three years after, and he didn't stay that long as I recall. I'm not sure just what the timing was.
McCann: He came in about 1951 I think and stayed for about two years.
Ryan: I was in there 1948, and Trevor earlier on, so he was sort of an addition to an existing team.
McCann: So the team really was you four plus Brian plus the technicians etc?
Ryan: Yeah, well Frank Tonking was the draughtsman and the other technicians and the workshop people I think, all contributed. The whole division was very much an integrated thing. I said Radio Astronomy just took over Radar Aerials and put them at Dover Heights, but in fact that would have involved the workshop people doing this. But mainly I think, we mentioned earlier, will mention later, Johnny Algie, I think Johnny Algie worked his way up from the workshop. People would work their way up from the workshop into becoming…some of them ended up as full professional people, they would get university degrees and that, so it was very much an integrated team and people came up as their abilities allowed them to.
McCann: So your memory of the technicians that actually worked with you were…
Ryan: Frank Tonking, John Algie, Ron Bowles, I remember the name, I don't remember much about working…Harry Flood I think was another person and you mentioned the name earlier, Palmer wasn't it, there was a Bill Palmer, I forget his name, he was a very happy go lucky cheery sort of Father Christmas type, as I recall.
McCann: What about your interaction with other teams - you know the larger teams - say the Radio Astronomy group and the Rain Physics group?
Ryan: Yes, well the whole division, not only Radiophysics division but National Standards division, which shared the same building, we shared the common cafeteria and that and a lot of morning tea, lunchtime, afternoon tea discussions were very important in throwing around ideas and getting interactions and as far as Radiophysics was concerned there was quite a lot of, we knew about one and others work, about the Radio Astronomy and the Navigational Aids as going on and the Rain Making. I think the Rain Making was a bit out on its own, I don't think it was, I don't know whether I'm doing Taffy Bowen an injustice but I think he sort of imposed that on the Radiophysics. The other lines more or less grew naturally out of the wartime work using the same electronics techniques and the aerial techniques for the Radio Astronomy and Navigational Aids, using them too. Whereas Rain Physics, I'm not sure when it started, but I think it came in, probably, it may have been Fred White was it that was suggesting that there was a sort of great need for rain in Australia and thought that, I think the Americans were having a crack at it too, and Bowen seized on this idea and pushed it and maybe it was sort of imposed, so I would say there was strong friendliness/interaction, between the Radio Astronomy, the Navigational Aids, the CSIRAC computer and to a lesser extent the Rain Physics people.
McCann: In a sense the computer team lost out in a way to the Rain Physics group, I guess.
Ryan: Yes, certainly in one way of looking at it. As I said, Bowen was very strongly attached to the Rain Physics one. He put a lot of his own status or whatever is the proper word, on the line on Rain Physics and in fact he lost out I think, we can see in retrospect it wasn't a success. You can see in hindsight that he would have done much better to put his effort into Radio Astronomy and well certainly computing, you can see in hindsight, and I'm not saying that he could have seen it at the time. Navigational Aids lost out to some extent but that was because of political factors beyond even Taffy Bowen's political powers I think, and I don't think anybody could attach any blame to the effort he put into Navigational Aids was probably justified. The one that I worked on later, the transistor DME. Brian Cooper and people built the original valve distance measuring equipment, then I switched onto the transistor development and that whole thing went on for many many years and was in use in Australia and elsewhere for a long long while so that was all justified I think by Taffy's decision, so he got two out of three right. I think he did pretty well, the Rain Physics didn't come out well. McCann: It seems that Navigational Aids was a very successful research program area, wasn't it? Ryan: I think so, it started in the 1940s went through the '50s, '60s and I think it may well even have been operating the transistor DME and the production commercial model developed by AWA was probably still being used quite late and instead it was only overtaken by political factors that forced them out of the 200 megahertz band and therefore would lead to develop 1,000 megahertz Tacan which was developed by the British companies and the United States. And of course, Radiophysics ran later with Interscan project, which was mentioned in that article, which was really carried on from the Navigational Aids. And that was again the team, it wasn't just CSIRO and Radiophysics, we also had a very strong team relationship with AWA for the development of the commercial distance measuring equipment and also with the Department of Civil Aviation, Egon Stern, was the man I remember there, who liaised with Taffy Bowen and with Radiophysics to keep that going, so it was really a inter-departmental, you know with DCA, Department of Civil Aviation, was part of the public service really and CSIRO was a research place, and AWA was a fully private company, but we did work very well together.
McCann: I will just ask you some questions about the context of Australian science in general. Can you comment on the general scientific area in which you worked. I mean the context of Australian science in general.
Ryan: Yes, well it was I think, as I said earlier, the start of a golden age and it was in the middle of a political furores. That Chifley had trouble with the miners, there was the communist menace was supposed to be raging and we had blackouts which certainly slowed up our development of the computer and probably affected other parts of Radiophysics too. I remember, you mentioning teams earlier, Tommy Kaiser, was an important person, I'm not sure where he was working, probably in the Radio Astronomy I imagine, then he went on a studentship to the United Kingdom and because he, like a lot of people in the Radiophysics division, was very left wing and maybe even communist some of them. He took part in a demonstration in London, probably against nuclear weapons or something like that, and this was during the Chifley era and Chifley, even though he was Labor, brought pressure on Fred White probably and the CSIRO to stop Tommy Kaiser. I think he was sacked essentially because of that, so that was the sort of political milieu. But on the other hand when Menzies got in, subsequently when Chifley lost out at the next election, Menzies was very supportive of the research side I think. CSIRO funding went on, as I said, Radiophysics division kept on expanding, and he supported that, pure research was supported quite strongly, Radio Astronomy was absolutely pure, as well as the Applied one and the Navigational Aids and the CSIRAC would have been. Menzies would have been in there before the CSIRAC finished, so I don't think it was because Menzies said something. This was shown later on by him developing, or about the same time developing, ANU and I think we have mentioned before that one of the factors in the demise of the CSIRAC project was that Professor J.C. Jaeger, who was an important mathematician, and possibly might have taken some leadership role in the tech/digital computing area was given a position at ANU, therefore he wasn't available to take part in that, but obviously ANU was I think a very good development and was a big contributor to Australian science generally.
McCann: What about your place in world science as a whole. For example, did you feel isolated? Was there a strong British and American influence? And was Australian science equal to the rest of the world etc. would you say?
Ryan: I think it started behind the eight ball. Before the war the universities, the State universities were, some were quite moribund. There were two people that I remember when I was going through my course at Sydney University who were very good. One was R.E.B. Makinson, in the Physics School, I think he was a real scientist and he kept going for a long while, his wife Rachel Makinson too, was quite important to CSIRO in the wool research and secondly in the Electrical Engineering Department, Ron Aitchison, he was later professor at Macquarie University, was very good in the electronics area I remember. Generally it was not very good and this is why CSIR, as it was originally, came into being in the agricultural area to do scientific work which obviously had to be done in Australia with the particular agricultural problems and the fauna and flora in Australia and that organization existing was then the obvious place to put Radio National Standards laboratory, first of all, and then Radiophysics laboratory into the CSIR at that stage and when it came along during the war, so CSIR and CSIRO later, had the funding and did the majority of the research, the universities were very short on, were very small on the ground I think in the scientific area, scientific work, research work then. But of course, Menzies I think was the one who initiated more funding to the universities and they started becoming more aggressive and there was more. I think there hadn't always been this sort of whingeing about why did CSIR get all this funding and we don't get any, but nothing much came of it when Menzies came along and that stage I think the universities decided to get work and of course the role of the CSIRAC, I not sure how the timing was, whether that played any role in the demise of the CSIRAC project, but Harry Messel of course came on the scene at Sydney University, he livened it up considerably, and imported a sort of fully operating computer, the SILLIAC later on and I think CSIRAC ended and SILLIAC came along later, but the bias was switching from the… CSIRO still got quite considerable funding but it was recognised that there was a role which only the university can play. There was a need just to train research people, if universities weren't doing research work then they wouldn't be turning out graduates. I must say, in my case, even during my under-graduate degree, I did some research work in electrical engineering, a photomultiplier nuclear detector I was working on as my thesis project, and Ron Aitchison or someone who probably was the one who was supervising me on that. So there was a move to it earlier, but it was very, funding was very small and the effort was probably pretty marginal.
McCann: And you said you did some of your training within the Radiophysics division.
Ryan: Yes, the engineering course was a four year course and in your third year you did two terms and then you stopped doing any formal courses and you went and got a job in industry and I got a job in CSIRO Radiophysics Division and worked with Harry Minnett on, I don't remember the detail, I remember we worked on Klystrons and we were looking at some camera project, Harry Minnett recalls, so I worked there for six months and got to know how to do things, to get things operating.
McCann: Just thinking about the present day compared with the day when you were a young scientist, has the structure and organization of Australian science changed a lot since those early post War, WWII, days?
Ryan: Yes, I think so, I think it has changed very markedly and even since I have retired in 1990 I think...I can see…probably the universities went up on an upsurge, we were just saying, they were in the moribund, in the doldrums, with some areas of good work going on, very limited areas, and then there was the surge up in the '60s and '70s and I think they reached a peak and then possibly they over-expanded, there were too many universities, too many students coming in and then of course the expenditure, the economic rationalism came in and expenditure was cut, so I think the universities themselves are saying we are on a decline and the Physics School for example is one perhaps which comes first to mind, Sydney University Physics School was in a very moribund state and then I think it surged up when Messel came along and it bottomed, if I'm not mistaken. I think it is well much into decline at the moment like a lot of Physics Departments.
McCann: If you use that model like a curve of it coming up, you were really in at the beginning.
Ryan: I was very lucky, I got in on the ground floor and I went up to the surge and then by the time it was on the turn I was getting out into actually more interesting work in the safeguards in the '80s where I could sort of go on doing my role...First of all in AAEC I was doing radiation detector work with gallium arsenide which was very interesting and which we did get quite a lot of support for there. The responses were, well they were certainly getting limited it wasn't anything like it is now I don't think, even at Atomic Energy Commission there was quite generous support, I think CSIRO had the same sort of support, but it's become much tougher these days, certainly the universities, and CSIRO. A major difference of our time was this need to go out into industry and get contracts, get support, get money, funding directly from industry. In our time the money just came from the government budgets, no problem, you didn't have to go out and beg for funds.
McCann: It seems to be a very optimistic, idealistic period from the science point of view, there was the emphasis on pure science and so on.
Ryan: We had our own, we could do more or less our own thing. Quite largely what we were interested in, within limitations, it certainly wasn't completely… but I remember the Australian Atomic Energy Commission when we developed the Gallium Arsenide. I wanted to, I put a proposal we go into Gallium Arsenide solar cells and while the solar cells and the Gallium Arsenide probably wouldn't have been a good long term project, the Gallium Arsenide itself has been very commonly developed and CSIRO later on took that path and did work in Gallium Arsenide, so I didn't get the right to do that, but I went ahead into other areas quite happy in the work. So looking back on my career, I think I was very happy at Radiophysics. I went into a cul-de-sac in medical physics for a couple of years at the University of NSW Medical School. Again I think it was an inter-disciplinary problem. It was very difficult to get engineering, electronics people working together with the medical people. Attitudes and approaches to scientific problems were just so divergent that it just didn't work out and I left after a couple of years and went to Atomic Energy Commission where again I think we had quite a lot of, we had ups and downs there too, I think Atomic Energy Commission was not nearly as happy a place to work as CSIRO divisions were, it was perhaps too orientated to the Applied, to the Engineering side and when it was AAEC, not enough to the more pure scientific side. It certainly had a lot of support and a lot of good work was done there by many people.
McCann: Could I just ask you about that period in the late 1940s, early 1950s when there was a change in Government and all the sort of political turmoil that was going on. Did you notice a difference, you worked through that period in the Division, did you notice a difference in the way things were done or the funding happened and things like that. Not that you might have been so much aware of it, concentrating on your project, but did you notice a difference in between styles between say the Labor Government and the Liberal Government took over and so on?
Ryan: No not really, as I said, I think the funding still came through largely. It was more the McCarthy era thing like you know that went on in Hollywood that if you were… Tommy Kaiser was a very upfront communist in demonstrating that. I think most of us were left wing and this would come out in the CSIRO Officers' Association, which was a professional union and we were… I don't remember anybody going on demonstrations here, not like in the 1970s, '80s, '90s. I don't think the conservatives, Menzies, perceived science itself as being anti-conservative, they thought of individual scientists like Tommy Kaiser as going too far and therefore needing to be pulled back into line. The Australian Atomic Energy Commission of course had security things, but the CSIRO, I mentioned to you, that when the nuclear work was, when the Government decided that we should get into some sort of nuclear work, atomic, there was I think quite strong consideration that it should become a division of CSIRO. The same way as Radiophysics had become a division with radar coming in. And I think the CSIRO as an organisation would have been quite happy to do that and going back to my earlier comments, I think CSIRO would have put a better stamp of scientific excellence, pure scientific excellence in the nuclear area than the AAEC did, but mainly because of the security aspects imposed by the Americans that they would only release the information on nuclear matters if there was a, if the security service ASIO and all them had a hand in it. CSIRO's reaction to that was, no, we are, o.k. we kept the radar thing that was confidential during the war but once the war was over, the same way as MIT published all their work on the radar, CSIR Radiophysics, was, with all the other CSIR divisions committed to quite open and free access, so they said no, we are not having it, we won't take into this organisation, you can't have one division which is highly confidential and secret, working…, it was sort of an integrated organisation, so perhaps for that reason or because of Baxter's, and again personalities came into it, J.P. Baxter was the first head of the Australian Atomic Energy Commission, maybe he pushed the Government, it became a separate organisation, but there was no real political determination of how we…. I don't think we were penalised because we were left wing, not by Menzies, I don't recall that….
McCann: Just one final question. After you left the Mk1 project, did you keep an interest in computing in general. Did you work in that area or did you move away from that?
Ryan: I don't recall keeping much of a contact in it. I knew about it. I think it developed relatively slowly and it developed in the direction probably which didn't interest me. In the mainframe computers, the IBM and Cobol, Fortran thing which was more business orientated, not I think that was the way it went and only when it came back in the 1980s into the smaller ones, the pc and the predecessors, did I again become interested. Well, I probably used ordinary little calculators, but my first real re-entry into the computing era was not till the Hewlett Packard little scientific programmable calculator, HP41 came along and I started using that in doing my data reduction on non-destructive (NDA) assay measurements on nuclear fuel and then using the laptop computers, the same sort of thing. So I probably read about it, knew they were building the mainframes but I didn't know anything about it. I never learnt to program in Fortran which many of my compatriots did. I only did Basic when I needed to use the laptop computers. ..
McCann: So I guess you lost contact with people like Trevor Pearcey and Maston and so on.
Ryan: Yeah, as I recall the only contact with Maston Beard was really late in the day in the 1980s probably when SANA - Scientists Against Nuclear Armaments - Maston was active in that and I joined it and stayed in it for a while, but I don't recall. That's why you had some trouble in locating me, because I lost contact, I didn't really see these people. They moved up to North Ryde and I was still living in the eastern suburbs and we just didn't happen to meet socially. Technical meetings is one place I might have met them, I just met Don Yabsley the other night at an Institution of Engineers meeting.
McCann: You mentioned Don Yabsley - what position was he in?
Ryan: He was in the Radio Astronomy I think for the major part of his career in Radiophysics. I'm not sure how he ended up, he may have gone into something else later on, but he stayed at Radiophysics right through from.., he was recruited like Maston Beard and that, from graduation in university and stayed with them until he retired. Probably about the same time as I did I imagine.
McCann: Once the computer moved to Melbourne, did you hear any reports about the computer from that point onwards, from mid '50s onwards, sort of thing?
Ryan: No, not really. You mentioned earlier, Murray Allen working on that other computer, CIRRUS, I knew Murray Allen and Ross Blunden was another man over there, but they were working in the electrical engineering school, I think at Sydney University, so I would have had… That was probably while I was still in the CSIR, working in the university grounds up until the '60s, so I would have known through them, but I don't recall anything. I knew it was in Melbourne, but I didn't go down to Melbourne particularly, I might have gone through Melbourne, going on holidays to Tasmania, but I don't remember. Did they publish many papers? I don't remember seeing anything coming out of CSIRAC in that period. Again, mainly if they did it I wouldn't have seen it, I wouldn't have been looking at the computer journals, I would have been more interested….
McCann: No, I don't think there was much, it was more the projects/programs they were running on the computer itself, rather than any computer developments per se.
Ryan: No, I was in the transistor area, working on transistors at Radiophysics and then using transistors at AAEC as well… and using the computers and that and the radiation detectors at Atomic Energy Commission Safeguards Office.
McCann: Well, thanks very much for your time and the information you have given. I think we will finish the interview now.