From 1961 to 1975 the mathematician Jack Howlett, who has died aged 86, was director of the Atlas computer laboratory at Harwell. Jack, a numerical analyst, was one of that select group who early recognised the significance of computing, and influenced the development of mechanical and later electronic computing.
He was educated at Stand grammar school and read mathematics at Manchester University. He then worked, pre-war, for London Midland Scottish Railways - introducing to engineers the benefits of computation and analysis.
From 1940 until 1946 he was a member of Professor Hartree's Manchester University group which built a mechanical differential analyser; an analogue machine which was probably then Europe's most powerful calculating engine. The group's calculations made a valuable contribution to the atomic bomb project and led Jack to take charge in 1948 of what became the Harwell Atomic Energy Research Establishment's computing section.
In 1958 Harwell acquired a Ferranti Mercury computer and with Manchester University and Ferranti planned the Atlas project. The Atlas 1 was installed at Harwell, under the control of the National Institute for Research in Nuclear Sciences and outside the security fence to provide access for university users. Thus it was in 1961 that Jack became laboratory director.
The laboratory provided computing power and services, and had a programming group and a small group of research fellows. The laboratory was in the vanguard of numerical work. Jack set up a small group concerned with computing and X-ray crystallography, and took the first steps towards establishing what are now known as databases. Collaboration with the Meteorological Office led to a weather forecasting model which forms the basis of current computer-based forecasts. Under Jack's direction, the laboratory attracted a stream of British and overseas scientists and thus were computer techniques disseminated during the 1960s and early 1970s. He created an atmosphere of commitment in the laboratory. Loved and respected, he knew everyone by name and to have been a member of Jack's Atlas laboratory is recognised as a privilege.
After he relinquished the directorship he was for two years chairman of the national committee on computer networks and continued to work as a consultant to the computer company ICL, editing its technical journal. From 1977 until 1983 Jack and Lancaster University's Kenneth Beauchamp set up and ran Nato seminars at Bonas, near the French Pyrenees, on information technology and the computer network.
The Atlas Laboratory's visitors' book bears ample witness to the array of famous scientists and engineers who came to see him. Jack had international contacts across many scientific disciplines and a wide circle of friends. At Bonas, as a lover of good food and wine, he was in his element. He exercised his talents as a linguist by translating books on computing from the French.
When he took time oft to walk in the fells and Yorkshire moors and to indulge his love of music and the arts. To share some of this enthusiasm was a great joy to many, as he spoke of the unaccompanied violin works of Bach or the delights of Renoir. He lived and worked through a remarkable period and has left many with a sense of gratitude for making life that much more interesting.
He is survived by his wife Joan, who he married in 1939, four sons and a daughter.
