Obituaries Australia

One of the University of Sydney's most distinguished scholars and long-serving academics, Emeritus Professor Henry Oliver Lancaster, has died in Sydney at the age of 88. He played a prominent role in the diverse but related fields of medical research, theoretical and applied probability and statistics, international research of the history of statistics, educating statisticians, and helping to found national professional organisations.

Two of his epidemiological population studies had important implications for better understanding of the causes of disease, namely, deafness as a result of congenital rubella and the relationship between sunlight and malignant melanoma.

Oliver Lancaster was born in Sydney and spent his childhood in Kempsey, where his father was a busy general practitioner. When his father died prematurely, his mother returned to her nursing profession to support Oliver, only nine years old, and his siblings.

Lancaster showed exceptional mathematical ability early in life, entertaining his father's professional guests with his skills in mental arithmetic. This aptitude led to an accelerated school education.

He was an enthusiastic cricketer at school and played rugby union for Manly in his student days before taking up billiards, bridge and chess after medical graduation. In later life, he played pennant bowls with referee status at several Sydney clubs and was a keen supporter of the Manly Sea Eagles. His love of body surfing, first acquired at South West Rocks and fine-tuned at Manly, was passed on to his sons and grandchildren.

At the University of Sydney, after one year in economics and arts, he switched to medicine, graduating in 1937. Initially unable to decide how best to apply his mathematical skills in his medical career, he trained in pathology for several years at Sydney Hospital before enlisting as a medical officer in the AIF in 1940, rising to the rank of major. He served as a pathologist with the Australian General Hospital in the Middle East, in Townsville, and finally in New Guinea, commemorated in Nora Heysen's 1944 pencil portrait for the Australian War Memorial's collection.

While still on active service in New Guinea, he enrolled in an external course in mathematics at his alma mater, studying under a kerosene lamp on tropical nights. On returning to civilian life, Lancaster secured a position with the university's then School of Public Health and Tropical Medicine (SPH&TM) where he continued his mathematical studies which were now slanted towards medical statistics.

Later he gained further training in medical statistics through a Rockefeller Fellowship in Medicine at the London School of Hygiene and Tropical Medicine, where the renowned statistician Austin Bradford Hill was professor of medical statistics.

There he was also influenced by J.O. Irwin and formed a lifelong friendship with another young statistician, Peter Armitage, later professor of medical statistics at that school, and Oxford, and president of the Royal Statistical Society.

At the University of Sydney Lancaster progressed from lecturer to associate professor of medical statistics. Because any promotion to full professor in medical statistics seemed unlikely because of internal politics at the SPH&TM, he decided to focus his interest on mathematical statistics and was appointed as the inaugural professor of mathematical statistics at the University of Sydney in 1959.

As a teacher, he was initially hampered by a significant stutter which he gradually overcame.

He was a strong believer in using Australian material, whenever possible, as examples in his lectures. Armitage has noted that Lancaster had the unusual ability to turn from his very applied body of work to quite difficult and advanced areas of mathematical statistics.

Wartime studies of hookworm infection had led to his first publications in 1944 (with T.E. Lowe) in the Medical Journal of Australia, followed by more than 50 articles and editorials in that journal spread over more than 30 years, mainly analysing trends in death rates from all causes. These studies laid the groundwork for future analyses by other epidemiologists in Australia.

When the Sydney eye specialist Sir Norman Gregg first described the devastating effects of rubella, or German measles, in pregnancy on the developing foetus, there was speculation that the epidemic of rubella in 1940 was due to an unusually virulent strain.

It was soon recognised that deafness was sometimes an important clinical feature of congenital rubella. This led Lancaster in the early 1950s to examine records giving the date of birth of deaf children in Australia and Iceland. He was able to show that births occurring in the late 19th and early 20th centuries were preceded by known epidemics of rubella, thereby dispelling the hypothesis that this was a new phenomenon.

Soon after, in his study of deaths from malignant melanoma in Australia, he provided strong support for the clinical hypothesis that sunlight might be causally related to melanoma by showing higher death rates for people living nearer the equator in sunny Queensland than in the more southern states.

He was appalled that his sons and their families didn't make more stringent efforts to avoid the hazards of sunbaking and general exposure to sunlight, a concern that now resonates nationwide.

Having learnt Latin and French at school in Kempsey, Lancaster later taught himself German and Russian by reading bilingual Bibles. This personal discipline laid the foundation for his extraordinary knowledge of people prominent in statistics and science all over the world and greatly broadened his access to valuable sources for the history of statistics, culminating in major international publications on these themes.

He was assisted in his rapid access to information on scientific personalities and their work by creating and maintaining a system with more than 20,000 (pre-computer) reference cards.

Stimulated by his wider contact with notable statisticians in London, the first of his more than 100 papers on mathematical and statistical topics was published in Biometrika in 1949, followed eventually by his book The Chi-squared distribution (1969).

Following his bachelor degrees in medicine and surgery (1937) and arts (1947) from the University of Sydney, he was awarded doctorates in philosophy (1953), medicine (1967), and science (1971), a rarely achieved trifecta.

Lancaster was involved in founding the Statistical Society of Australia and also the Australian Mathematical Society, holding key positions in both. He was the founding editor (1959-1971) of the Australian Journal of Statistics.

Nothing pleased him more than recognition by his academic and professional colleagues. He gained huge satisfaction from his fellowship of the Australian Academy of Science and membership of the International Statistical Institute.

He felt deeply honoured to be awarded the academy's Lyle Medal in Mathematics or Physics in the year (1961) he was admitted to the academy, and later to be made an Officer of the Order of Australia (1992). For his part, he returned the compliment by dedicating his books to other scientists whose work he admired.

Like many academics, he held strong personal views on his favourite topics. He was often uncompromising in his opinion of people whom he regarded as doing careless work.

Lancaster continued to read voraciously and publish actively after his formal retirement. From his den beneath the Fisher Library stacks, in his eighth and "magisterial'' in Roy Porter's review for Nature, followed by Quantitative Methods in Biological and Medical Sciences: A Historical Essay (1994).

His funeral was held at St Barnabas's Anglican Church, Broadway, a short walk from the University of Sydney and the world of academia, his spiritual home.

He is survived by his first wife, Joy, and their five sons (Paul, Peter, Llewellyn, Andrew and Jonathan), 15 grandchildren and five great-grandchildren, by his second wife, Nancy, and by his brother, Rick.