Born Pukekoke, New Zealand, 15 August 1883
Leslie Comrie was an early leader in the modernization of computing techniques for astronomy, pioneering the use of machines to both speed production and improve the accuracy of astronomical tables. Comrie was the elder son of John Alexander and Lois Helen Comrie. His paternal grandparents immigrated to New Zealand from Scotland in the 1850s.
Comrie majored in chemistry and graduated with a BA from University College, Auckland, in 1915. He earned an MA in chemistry in 1916. Comrie developed an interest in astronomy while at University College and joined the British Astronomical Association [BAA] while still a student. After teaching for a short time at Auckland Grammar School, Comrie joined the New Zealand Expeditionary Forces and was in active service in France where he lost a leg during World War I. In 1919, Comrie was awarded a New Zealand Expeditionary Force Scholarship, which allowed him to study astronomy at Cambridge University under Arthur Eddington . He was awarded a Ph.D. in 1923 for his thesis on the occultation of stars by planets.
Comrie’s principal astronomical interest was positional astronomy. As a graduate student at Cambridge University, Comrie made a detailed study of methods of predicting the occultation of stars by planets. This work had a twofold significance. First, it led to a revival of interest in the observation of planetary occultations; and, second, it helped to stimulate Comrie’s interest in scientific computation and mathematical table making, an interest on which the rest of his career was based. In 1920, the BAA established a Computing Section to predict phenomena involving Saturn’s satellites but later expanded the section’s work to the prediction of other phenomena. Comrie was appointed as the Computing Section’s first director and undertook to coordinate a group of 24 volunteers computing the necessary data. Comrie produced a “Computing Memoir” in 1921 and in 1922 issued the first British Astronomical Association Handbook. Comrie resigned as director of the Computing Section in 1922 when his career took him to the United States, but he continued to maintain an interest in the work of the section. Importantly, this work gave him valuable experience in organizing computations and seeing the results through press.
Comrie spent nearly 3 years in the United States, teaching astronomy and numerical computation at Swarthmore College and Northwestern University. While teaching in the United States, Comrie began to publish widely on mathematical tables and computing. His main concern at this time was the outdated computing methods being used by astronomers. He encouraged astronomers to adopt calculating machines for their work.
In October 1925, Comrie returned to England to take up a post as an assistant at the British Nautical Almanac Office. Comrie’s appointment to the Nautical Almanac Office gave him the opportunity to implement his ideas on astronomical computation on a large scale. When he joined the office, computing was done by hand using logarithm tables with very few exceptions. Retired ex-employees were performing much of the work in their own homes. There was no mechanism for training new staff for the future. While the system worked well at that time, it was clear that this situation could not be sustained. As soon as he arrived, Comrie began to introduce commercial calculating machines and younger staff into the office. The first machines to be introduced were Brunsvigas, a Monroe, and a Comptometer. He then installed carriage-controlled adding and listing accounting machines and applied them to interpolation and differencing, devising new computing methods as he did so. Comrie’s most spectacular use of machine computation was his application of Hollerith punched card machines to the Fourier synthesis needed to produce tables of the position of the Moon. This work had kept two members of staff fully occupied all year round. However, with the punched card technique and machinery, Comrie was able to produce tables for the next 15 years in only 7 months.
In March 1926, Comrie was promoted to deputy superintendent and in August 1930 took over as superintendent. Comrie made two major contributions as superintendent of the Nautical Almanac Office. His first achievement was that he completely revolutionized the computing methods used to prepare the predications given in the Nautical Almanac. Comrie’s second achievement at the office was a complete revision of the structure of the Nautical Almanac. Aside from the additional publication of the Nautical Almanac, abridged for the use of seamen introduced in 1914, the form of the Nautical Almanac had remained largely unchanged since 1834. Comrie completely revised the Nautical Almanac to take account of advances in navigation, astronomy, computing methods, and typography. He also produced tables using the standard equinox of 1950.0 based on a suggestion he had made in a paper in the Monthly Notices of the Royal Astronomical Society in 1926. The publication of Planetary Co-ordinates Referred to the Equinox of 1950.0 in 1933 led to a simplification of the calculation of special perturbations.
In parallel to his work at the Nautical Almanac Office, Comrie was also building up an international reputation as a mathematical table maker. He was responsible for the start of the British Association Mathematical Tables Committee series of mathematical tables, and personally published several tables. His desire to help others with their computational problems, and the level of outside computing work he carried out, led to his dismissal from the Nautical Almanac Office in 1936. However, he continued as a table maker and, while he did not hold another astronomical appointment, did maintain a lifelong interest in positional astronomy. In August 1937, Comrie set up the Scientific Computing Service as a commercial scientific computing bureau – one of the first of its kind. Comrie’s new company provided invaluable support to British military operations early in World War II.
Comrie was elected fellow of the Royal Society in March 1950 for his contribution to computing and mathematical table making. He was also a fellow of the Royal Astronomical Society (serving on Council 1929–1933). Comrie was a member of the British Astronomical Association, the American Astronomical Society, the Astronomical Society of the Pacific, Sigma Xi, the New Zealand Astronomical Society, and the Astronomischen Gesellschaft. From 1928 onward, he was an active member of the International Astronomical Union, serving as president of Commission 4 (Ephemerides) from 1932 to 1938. Comrie was also secretary of the British Association for the Advancement of Science Mathematical Tables Committee from 1929 to 1936. A crater on the farside of the Moon is named Comrie.
During his lifetime, Comrie was well-known for his computational abilities, his energy, and his kindness and generosity, but he was also a blunt and forthright man, fanatical about his work. Although he would gladly offer help and advice, he expected his advice to be taken up with alacrity. His high standards of work and his emphasis on precision and accuracy meant that he did not suffer fools gladly, and he said so. While he will be remembered for his singular contributions to astronomy in the form of vastly improved ephemerides and working tables, Comrie was also responsible for the widespread adoption of commercial calculating machines into many branches of scientific computation and many improvements in mathematical table making and table typography outside astronomy. Comrie married twice: first in 1920 to Noeline Dagger (whom he later divorced) and second in 1933 to Phyllis Betty Kitto. Comrie had two sons, John and Julian, one from each marriage.