Sewall Wright 91percent of the variance of hog production and prices by fluctuations in the corn
crop and various correlations. In several ways this was ahead of its time, especially
by introducing time lags, later used extensively by economists.
This work was done in the early 1920s, but, incredibly, Wright was not permit-
ted to publish it. An animal husbandman had no business working on economic
problems. The manuscript lay unpublished for several years. It required the inter-
vention of Henry Wallace, later Vice President of the United States, to persuade
his father, then Secretary of Agriculture, to allow the paper to be published. It
has long been regarded as a landmark, especially by economists.
In the early years after Wright’s discovery of the method, path analysis was
very popular among animal breeders. This was largely due to the strong influence
of Jay L. Lush, an animal breeder at Iowa State College, who was a great admirer
of Wright and a great popularizer of his work. Path analysis is best adapted to an-
alyzing data from natural populations rather than planned experiments. Later, as
animal breeding became more regularized, other methods that were more adapted
to computer analysis and to tests of statistical significance took over, so path
analysis is now something of a rarity among animal breeders. Meanwhile, it has
been taken up by social scientists and is now used by sociologists, economists, and
philosophers.
4 PHYSIOLOGICAL GENETICSFor most of his working life, Wright studied guinea pigs. His major contributions
were to gene action and interaction, what was then called physiological genetics.
While at USDA and the University of Chicago, a large part of his working day was
spent with his guinea pigs. He did the husbandry and record-keeping himself. His
extensive records still exist and have recently been transferred from Wisconsin to
the custody of William Provine, a historian of genetics, who wrote the definitive
biography of Wright[Provine, 1986], including the work of this section. Wright’s
physiological interest comes as a surprise to most geneticists, since his inbreeding,
population structure, and evolutionary work are much better known.
Wright’s initial work on guinea pigs was an accident of timing. William E. Cas-
tle, his mentor at Harvard, worked on various rodents and assigned each incoming
graduate student to a particular species. C. C. Little worked on mice, Castle him-
self studied rats, E. C. MacDowell studied rabbits, and John Detlefson, guinea
pigs. Wright arrived just as Detlefson was leaving so naturally he began working
on guinea pigs. He never willingly stopped. When he moved from Chicago to
the University of Wisconsin he had to give up the guinea pig colony for lack of
facilities. He regretted this, but I am sure that this was for the best, since he was
then free to write. His four-volume treatise on evolution and population genetics,
written in his eighties, might never have been completed if he had had his beloved
guinea pigs to work with. As it was, Wright spent the first five years in Wisconsin
completing analysis of his guinea pig records.
Soon after joining the USDA Wright wrote a series of articles in the Journal