Admiralty Computing Service
Mathematics in Government Service and Industry: Some Deductions from the War-Time Experience of the Admiralty Computing Service (1946), by D H Sadler and John Todd.
Note. The John Todd referred to here is in our Archive as Jack Todd.
In 1942, Prof J A Carroll, shortly after joining the headquarters staff of the Director of Scientific Research in the Admiralty as assistant director, saw that scientific effort in the Admiralty establishments could not be fully employed in its proper field because scientific men were having to undertake their own computational work. Often, for lack of training in computational mathematics, they spent even more time on this than it needed intrinsically and, more serious still, for lack of internal or external facilities, some investigations requiring extensive computations were foregone. An investigation of the computational needs of establishments by Mr D H Sadler, superintendent of H M Nautical Almanac Office, was accordingly arranged. Earlier, the Nautical Almanac Office had been doing large-scale computational work for other Government departments, notably the Ministries of Aircraft Production and Supply, in addition to its navigational work for the Air Ministry; proposals for an inter-departmental computing organisation which had been made earlier by the Office were overtaken by the joint proposals, referred to later. Following Mr Sadler's report that a considerable amount of computational work could with advantage be centralised, a mathematical section was set up in the Department of Scientific Research and Experiment to deal with this work and with the more difficult mathematical problems which were known to arise in naval establishments and which were being either set aside or tackled by totally inadequate methods. This section became known as the Admiralty Computing Service. Mr John Todd was responsible for organisation and supervision of this Service. Staff were attached to the Nautical Almanac Office to carry out the actual computations under the direction of Mr Sadler, and arrangements were made whereby scientific workers in the universities and elsewhere could be employed as consultants.
3. Admiralty Computing Service (1947), by John Todd and D H Sadler.
Note. The John Todd referred to here is in our Archive as Jack Todd.
In 1942, in order to use more efficiently the scientific staff available in the Admiralty, the Director of Scientific Research set up, within the branch directed by Dr J A Carroll, an Admiralty Computing Service to centralise, where possible. the computational and mathematical work arising in Admiralty Experimental Establishments.
Mr John Todd undertook the organisation and supervision of the Service. By agreement with the Astronomer Royal additional staff were attached to H M Nautical Almanac Office to carry out the computational work under the direction of the Superintendent, Mr D H Sadler. In addition, arrangements were made to permit the employment of experts from the Universities and elsewhere as consultants.
The work undertaken by Admiralty Computing Service was in general of one of two classes: heavy computation, or difficult mathematics. Altogether more than one hundred separate investigations were carried out ranging from projects involving several thousand hours' computing to small problems for which a solution could be obtained in a few hours. In addition, a considerable amount of advisory work has been undertaken, usually informally. For instance the Five-figure Logarithm Tables were designed by Admiralty Computing Service for the Ministry of Supply, as one item in a comprehensive program for providing the optical industry with the tables they required. For various reasons, mainly owing to the increased use being made of machines and to the availability of the U.S. reprint of Peters' seven-figure table of natural trigonometric functions, other tables were never published.
Shortly after the formation of Admiralty Computing Service it became apparent that research work in Admiralty (and other) Establishments would be greatly facilitated if their members were informed in certain mathematical and computational techniques not usually covered in undergraduate courses, and of which no adequate account was available in easily accessible literature. Accordingly the preparation of a series of monographs of an expository nature was begun.
It was soon realised that while centralisation within a Department was an improvement, nothing less than centralisation on a national scale could be really efficient. At the end of 1943, an approach was therefore made to Sir Edward V Appleton, Secretary of the Department of Scientific and Industrial Research, asking for consideration of the formation of a National Mathematical Laboratory. Discussions, in which the experience gained by Admiralty Computing Service played an important part, have now resulted in the formation of a Mathematics Division of the National Physical Laboratory. Staff have been released from Admiralty Computing Service to form a nucleus for the computational sub-division of the new organisation. It is anticipated that the computational needs of the Admiralty will be met by outside organisations and the mathematical needs by an even larger use of the service of consultants, working under the general direction of the Director of Physical Research, Admiralty.