University College, London 1933-1943
In 1933, Karl Pearson retired and Fisher accepted the position of Galton Professor of Eugenics at University College, London. The position reflected Fisher’s unique qualifications – it was the only British University to teach statistics and the only post of Professor of Eugenics charged with research into human heredity. The Professor of Eugenics also managed the Galton Laboratory for National Eugenics, curated the Galton Collections and edited the Annals of Eugenics.
He did not inherit Pearson’s whole empire as a Department of Applied Statistics was split off headed by Pearson’s son, Egon. This structure did not make for harmony and relations between Fisher and members of Pearson’s department, especially its leading theorist, Jerzy Neyman.
Fisher encouraged his staff to develop their own research interests and hosted open tea-time discussions, which would in turn initiate new investigations. For example, the collecting of genealogies for Huntington’s chorea by one of his staff led Fisher to investigate linkages with the possibility of improving prognosis of the disease. Students would be invited to his home to see the living experimental material not only at the Galton Laboratories but at Rothamsted and in the field. However, his lecture courses were usually beyond most of his students. Only the smartest students could immediately grasp his brilliant but challenging ideas.
Experiments in Genetics
Notebook: Mice litters
The Galton Laboratory was a small department with four staff, plus Kenneth Mather as a genetical assistant. He assisted Fisher with experiments on linkages in mice and the purple loosestrife Lythrum salicaria. Fisher also collected data on moths, snails, nesting birds, grouse locusts and fruit-flies. He even performed taste tests on apes at the London Zoo and took part in the breeding of dogs with the Genetical Society.
However, as Professor of Eugenics (which would now be termed Human Genetics), most of his work involved the development of models for analysing human data and the estimation of gene frequencies, the effects of ascertainment and the detection and estimation of linkage, all supported by Fisher’s ethos of maximum likelihood. He published more than 30 papers in this field during his ten years in London.
On display:
- Notebook: Mice litters. Book no. 1 (3 April 1924)-no. 89 (5 April 1925) showing data for litters 1 and 2. With a ms story in Eileen’s hand at the beginning. Fisher Papers. Series 15. Data Records and notes
Lythrum salicaria
Lythrum Salicaria
Fisher began work on the purple loosestrife, Lythrum salicaria, in 1934 with a plot at his house in Harpenden and observations of plants in the wild. The wild plants grew in wet places like swampy fields and he would often return from his expeditions with trousers wet to his knees. He would also invite friends to his sister’s house in Argyllshire where he had found several colonies growing wild. Like many plants, Lythrum has polyploid cells with more than two paired sets of chromosomes which play an important role in evolution.
On display:
- Photograph: Lythrum salicaria
- Notebook: Lythrum data volume 1, 1942. The first of 24 volumes extending to 1968. Fisher Papers. Series 14
Blood Group Serum Unit
Ms list of blood types
Fisher suffered constricted budgets and struggled with administration to get the staff and equipment he wanted, although he was able to win external sponsorship. In 1935 the Rockefeller Foundation funded the establishment of a serological unit and the Eugenics Society provided an annual grant to keep the college journal The Annals of Eugenics published until well into the Second World War.
The Blood-Group Serum Unit aimed to collect data relevant to accessible pedigrees of medical interest under the now famous team of G.L. Taylor, A.M. Prior, E.W. Ikin and Dr R.R. Race
Blood groups had been little studied in Britain except for the needs of blood transfusion. The unit collected and typed thousands of specimens against ABO phenotypes to determine gene frequencies in the population and the inheritance of new blood groups. At the outbreak of war in 1939 the serological unit was instructed to move to Cambridge and become part of the blood transfusion unit. This unit was able, for the first time, to collect country-wide data on the frequencies of blood groups. It also collected blood samples from RAF personnel in exchange for soliciting volunteers for blood donations and providing blood group data to be stamped on the identity discs of each flyer. They also matched blood data with Scandinavia and Iceland which detected ancestral influences and, most importantly, elucidated the Rh factor.
On display:
- Newspaper clipping: ‘Deciding disputes by Science …’ [undated] Fisher Papers. Series 24: Cuttings book re [early] career
- Newspaper clipping: ‘Scientists uphold blood test warning …’ [1938] Fisher Papers. Series 24: Cuttings book re [early] career
- [Ms list of blood types of Fisher’s family listing Rh and genotype] Fisher Papers. Series 24: Cuttings book re [early] career
World War II
Following the bombing of London in 1940, and disagreements with administrators of University College regarding the value of research during the war and the destruction of the genetic animal stock, Fisher moved the Galton research unit to Rothamsted. Fisher’s scientific work was well known by 1939, but he was not asked to undertake war work, possibly because he was suspected, despite little evidence, of fascist leanings.
His two works Statistical Methods of Research Workers (1925) and The Design of Experiments (1935) greatly influenced many who ended up doing statistical war work, including a young sergeant from Gravesend, George Box, who studied the effects of poisons such as mustard gas and who eventually became Fisher’s son-in-law.
On display:
- Newspaper clipping: ‘London University. Plight of the Galton Laboratory’ / R.A. Fisher. Letter to The Times, October 3rd 1939
Scientific Advances
The Design of Experiments
Fisher’s book The Design of Experiments (1935) expounded the principles of experimental design which he had been working on since the mid-1920s. It is considered a foundational work in experimental design and introduced the concept of the null hypothesis in the context of the famous lady tasting tea experiment. It also included a chapter on the Latin Square. Eight editions were published between 1935 to 1966 and it has been translated into three languages.
In 1938 Fisher and Frank Yates released Statistical Tables for Biological, Agricultural and Medical Research, an expanded version of the tables provided in Statistical Methods for Research Workers prefaced with an introduction on how to use them. Five editions were published between in Fisher’s lifetime: the 6th posthumous edition was completed by Fisher in hospital during his last illness.
On display:
- Statistical Tables for Biological, Agricultural and Medical Research / R.A. Fisher and Frank Yates. 1st edition (London: Oliver and Boyd, 1938)
- The Design of Experiments. 1st edition (Edinburgh: Oliver and Boyd, 1935)
