In 1912, Sewall Wright was at the University of Illinois, studying a flatworm. The species in question, a bowfin parasite, was called Microphallus opacus, which as best as I can tell should be translated as "opaque tiny penis." Wright was studying its anatomy. He cut specimens into thin slices in order to create picture of the internal workings of the flatworm. According to William Provine's excellent biography Sewall Wright and Evolutionary Biology, 22-year-old Sewall "greatly enjoyed this research" (Provine 1986, p. 31). Even so, when Harvard professor William Ernest Castle came to Illinois and delivered a lecture on selection experiments in hooded rats and mammalian genetics, Wright was captivated and ready to leave Microphallus behind. He asked Castle if it was possible to do research with him, Castle agreed, and Wright quickly finished his flatworm project. That spring he left Illinois for the east coast.

For the next three years, Wright studied physiological genetics at Harvard's Bussey Institute. These years were critical to his development as a scientist, and to his growing interest in evolutionary theory.

Part of the reason Castle had so quickly agreed to take young Sewall on was that his graduate student John Detlefsen was graduating and leaving Harvard that spring. Detlefsen had been the main student working with Castle's experimental stock of guinea pigs. Towards the end of the summer in 1912, Wright spent some time with Detlefsen, learning as much as he could about the colony of small mammals. It was understood that Wright would be working with the guinea pig colony once Detlefsen left.

Wright's thesis research was on the genetics and inheritance of coat color in guinea pigs. He dove into this complicated world of physiological genetics happily, excited to use his knowledge of chemistry and genetics to understand the causes of the different coat colors present in the guinea pig population. Through this work, Wright made an important realization: that interactions between genetic elements are extremely common in the processes that link genetic factors (genotype) to observable traits (phenotype). In the end, he proposed a speculative system of several enzymes interacting. Provine describes Wright's hypothesis:

"The basic color producing enzyme (I) acting upon chromogen made yellow pigment. A second enzyme (II), or more likely a substance changing the first into a different enzyme (I-II), stabilized I and, acting upon chromogen, produced the dark colors. A third enzyme (III) or substance altering the others had a modifying effect upon the dark colors but not upon yellows, as in the pink-eye factor, or the reduction of blacks to browns" (p. 94).

This conclusion was steeped in the "reality, necessity, and ubiquitousness of gene interaction" - an indicator of Wright's future thinking on more lofty topics in the study of evolution. Provine stresses that this "conviction of the importance of gene interaction" pervaded Wright's thinking for the rest of his career (p. 94).

After Wright graduated, he took a job in the Animal Husbandry Division of the US Department of Agriculture in Washington. There he worked on understanding the effects of inbreeding in mammals. Again he used guinea pigs; he took samples from all of the stocks at Harvard up to the lab in Beltsville, Maryland where he started experimental populations.

Wright found that while most of the inbred families showed a general decline in vigor, several lines didn't seem to lose any functionality. More importantly, the lines clearly differentiated from one another, each fixing unique phenotypic characters like coat pattern, number of digits, etc. Wright's conclusion was that inbreeding could bring about novel gene interaction systems, something that selection in a large population could rarely do. His suggestion to breeders was to keep many inbred lines, and cross the ones with favorable characteristics: "Thus a crossbred stock can be developed which can be maintained at a higher level than the original stock, a level which could not have been reached by selection alone" (Wright 1922).

We should take two important notes from Wright's work on animal breeding. First, as would be expected given his previous work, he emphasized the importance of gene interaction systems. Wright's intuition from his work on guinea pig genetics was that some combinations of genes were favorable and others weren't, and that the result of a particular combination could not be described as the sum of its parts - interactions were important. Second, we should note Wright's practical motivation here to describe a scenario where breeders will have the most success at improving their livestock. When Wright turned to his theory of evolution in nature after spending years working on animal breeding, he brought over this notion of the most effective scenario for breeding and applied it as the most effective scenario for evolution in nature. This wasn't quite sound logic; the way evolution does happen in nature is not necessarily the most effective way it can happen (Pigliucci 2008). Even so, this idea seems to have subtly motivated Wright's work ever since his studies on animal breeding at the USDA.

Professor Michael Dietrich explains how Sewall Wright applied his ideas about efficient animal breeding to theories of long-term evolution.

While Wright was in Washington, about a decade after completing his thesis, he set to work on writing a serious manuscript about evolution in nature. Though he would continue his work on physiological genetics in guinea pigs, from this point forward Wright had his eyes set on the big questions of evolution, and along with other geneticists he would specifically try to find mathematical ways to describe evolution. I like to imagine this time in history like this:

Doctor Wright sits next to several cages in his USDA laboratory, diligently recording information in his leather-bound notebook about the shades of black and brown on the guinea pigs before him. He takes one out of its cage to get a closer look and sets it down on the table to record the pattern of its coat. As he is writing in his notebook, his mind drifts to the bigger question: how do genetic changes happen in natural populations? Turning slowly away from the cages, Sewall picks up a piece of chalk and starts scribbling out some thoughts on how an allele might change frequency in a population. He is writing for several minutes before he realizes he's made an error. He reaches down to grab the chalkboard eraser, quickly erases his mistake, and puts it down, never realizing that the "eraser" was the guinea pig he had been inspecting.
Did Wright actually use guinea pigs as erasers? The evidence is actually mixed, but you be the judge:

photo of sewall wright