Fourth part of T.R. Thompson's report on the American Tour, May and June 1947.

See the catalogue record for the full report.

Research Comments: This is the fourth part of an in-depth report by Thomas Raymond Thompson (TRT) and Oliver Standingford on their visit to the USA in May and June 1947. It was during this visit that they learned of the development of EDSAC in Cambridge, England!

This fourth section is dedicated to 'electronic machines in the office' or 'computors'. The authors define these as machines which count and record "by means of electrons". 

The report first covers ENIAC, and the authors report that "sensational journalism" called this machine "The Electronic Brain", describing it in terms that implied powers of reasoning and judgment that it did not, and never could, possess. The intention of this report is to inquire "into the nature and possibilities of this machine... to find out whether it, or any adaptation of it, was capable of being put to use in commercial offices, and, if this was not the case, to try to stimulate the development of such a machine".

The functions of a machine like ENIAC are:

1. to receive and store information (in code). It is this 'memory' function that offers the "greatest possibilities", according to the authors.
2. to receive and store in its memory operational instructions (in code), e.g. a series of orders.
3. to select from its memory information so as to record it in some way or to use it in further operations.
4. to act on orders to bring together associated words or figures.
5. to compare words or figures and react to any differences found.
6. to select figures from its memory and add, subtract, multiply or divide them and store the results of the calculation so they can be used for further calculations or recorded as a result.
7. information in the memory can be extracted and made available in written form.

The authors state that the real achievement in all this is not that a machine can do these things, but that it can do so "at electronic speed". They correctly ascertain that input to such a machine cannot be undertaken by clerks, but rather "by mechanical and electrical means".

The report then lists the ways in which such functions can be carried out by a machine, covering:

1. the concept of 'coding' using electrical impulses and binary,
2. input at speeds required by the computer using punched cards, punched tape, magnetic wire or magnetic tape,
3. forms of memory such as 'acoustic delay lines', Selestron (cathode ray tube), magnetic drum and 'sound track recording'.

Page 13 of this section of the report has a "diagrammatic representation of an electronic computor", illustrating the need for 'synchronisers' to coordinate the low sped circuits of the input and output channels and the high speed circuits of the computer.

Part V of this section of the report outlines the possible applications of such a machine to office work, covering sales invoicing, letter writing and payroll.

Part VI lists the important contacts the authors have made during the visit. They refer, amongst others, to:

• Hermann Goldstine of the Institute for Advanced Study (IAS) who "was originally in charge of research for building ENIAC at the Moore School of Electrical Engineering". The authors state that "Prior to our meeting him he had not realized the scope for the use of electronic machinery in a commercial office" but that he "was very enthusiastic" about the idea and wished to help. It was Goldstine who then introduced the Lyons team to Douglas Hartree and Maurice Wilkes.
• the Moore School of Electrical Engineering, "where ENIAC was built by the original team of mathematical physicists, electrical engineers etc., which has now been disbanded". The authors report that there was interest here but no enthusiasm for the Lyons ideas.
• Mr. J.C. [Presper] Eckert, of Electronic Control Company, who was "the electrical engineer who actually designed the parts and was in charge of the building of the ENIAC"., but who had "broken with the Moore School over a question of patents and because the school would not contemplate the question of commercial development".
• Radio Corporation of America Laboratories who were working on cathode ray tube memory and who were interested in Lyons' ideas but "showed no desire to take part in this development".
• IBM who the authors suggest were only interested in "electronic calculation as an adjunct to punched cards".
• Prudential Insurance Company of America, who the authors identify as "the only commercial concern actively interested in the application of electronic machinery in its offices". 
• Howard Aiken of the Computation Laboratory at Harvard University, who the authors suggest "in the course of three years work appears to have achieved far more than the Moore School at Philadelphia" and has "two machines working [Harvard Mark I and II] and a third under construction". Aiken's Harvard Mark I is not "strictly an electronic calculator" but "operates by high speed relays". It is slow compared with ENIAC but more reliable, working 80% of the time compared with ENIAC's 20%. When TRT and Standingford visited it was working on mathematical tables, "doing work that would require 300 clerks using ordinary calculating machinery". It made no errors at all. The Harvard Mark II works "at 12 times the speed of Mark I". Harvard Mark III will be much smaller and will make use of magnetic drum memory and just 1,000 valves. It will be "20 times as fast as Mark II". Aiken apparently was "enthusiastic about the commercial possibilities and would co-operate in any research in this direction". Aiken has already thought beyond office work and looked into process control and indeed told the authors of this report that "we are on the threshold of a second industrial revolution".

Part VII of the report covers a visit to the Mathematical Laboratory at Cambridge, after the authors' return to England, where they met both Maurice Wilkes and Douglas Hartree. Both men were "keenly interested in our [Lyons'] proposals for a commercial machine and prepared to make their knowledge and advice available. They had not previously realised the commercial possibilities of the machine, being preoccupied with its use in scientific research".

The authors report that Cambridge is currently working on building a pilot machine with an 'acoustic delay line memory' to demonstrate what can be done so that research funds could be raised. The memory will be fed slowly by teleprinter. The report states that Lyons consider Cambridge to be "already far in advance of the Moore School, where the research is on similar lines." The Moore School are experimenting with 10" delay line tubes, but Cambridge is incorporating tubes 4-5' in length. Development is slow however, due to lack of funds: "We were told that given £2000 they cold complete much more rapidly."

Part VIII of the report looks forward to steps Lyons could take "to advance the development of Electronic Machines". The authors suggest that development of such a machine "may well be a prime factor in relieving the present [post-war] economic distress of the country... Lyons occupies a key position; no one else here, as far as we can learn, has realised the far-reaching possibilities of electronic machines." They go on to state that "It is possible for us [Lyons] to play a passive role by merely keeping in touch with developments, and in due course buying machines as they become available, probably from American sources. But such a role would not enable us to have any influence on the kind of machines built, and without commercial influence they may well be built in a form more suited to handling mathematical and census calculations".

A footnote to Part VIII refers to developments at the National Physical Laboratory under Dr. [Alan] Turing and Dr. [John] Womersley, who are "working on projects of the Services and the Board of Trade, but do not expect to show results for some years".

The final page of the report, before the Appendices, suggests five ways in which Lyons could proceed:

1. encourage Hartree and Wilkes and provide them with clerical procedures for experimenting with and possibly providing financial assistance or influencing other bodies such as the Nuffield Trust to support them.
2. trust large electrical companies such as EMI or GEC to develop and exploit a machine.
3. work with Electronic Controls Inc of Philadelphia, who would likely require financial assistance as well.
4. approach the British government, asking them to co-ordinate research in this country and provide resources "to get Britain first in the field".
5. "build a machine in our own workshops drawing information and advice from Cambridge and Harvard Universities".

In the event, the Lyons board chose a combination of options 1 and 5. (LM)

Provenance :
From David Caminer's papers.