42 INHERITANCE IN GUINEA-PIGS. 



SIZE INHERITANCE IN GUINEA-PIG CROSSES. 

 PREVIOUS WORK ON SIZE INHERITANCE. 



For several years my pupils and I have been engaged in studying the 

 inheritance of size among tame or domesticated animals, a subject 

 deserving of careful investigation both because of its economic impor- 

 tance and because of the light which it may throw on general theories 

 of heredity. A preliminary study based on skeletal measurements of 

 rabbits was published in 1909 (Castle et aL), which seemed to show that 

 size inheritance is "blending" and does not involve the segregation and 

 recombination of distinct Mendelian size-factors. MacDowell (1914), 

 at my suggestion, repeated this work on a larger scale and with similar 

 observational results, establishing, however, the additional fact that 

 an F 2 , or a back-cross generation, usually shows greater variability in 

 size than an F x generation, following a cross between animals of unlike 

 sizes, though the general result in both cases is the production of inter- 

 mediates. On theoretical grounds MacDowell favored the Nilsson- 

 Ehle view that all variation, even when continuous, is caused by genetic 

 factors themselves discontinuous, and that blending inheritance in- 

 volves multiple segregating factors. But MacDowell points out that 

 this interpretation is not the only one of which his observations are 

 capable. The genetic purity of his material also, while sufficient to 

 establish the general blending character of the inheritance, is not suffi- 

 cient to meet the extreme demands of the multiple-factor hypothesis. 



At the same time that MacDowell was making his observations on 

 size inheritance in rabbits, Detlefsen (also in my laboratory) made 

 observations on size inheritance in crosses between Cavia rufescens and 

 the guinea-pig. He was unable to rear an F 2 generation, because of 

 the complete sterility of the male hybrids, but from a study of repeated 

 back-crosses concluded that "there were no great differences in vari- 

 ability in the back-crosses of hybrids to guinea-pigs which would indi- 

 cate segregation and recombination of factors for size." This conclu- 

 sion he reached without theoretical bias, for he adds: "The results in 

 no way controvert the possibility that size may be due to factors which 

 are inherited in Mendelian fashion; but segregation was not apparent 

 in these classes of matings in this species cross." 



My colleague, Dr. John C. Phillips, at about the same time (1912, 

 1914), undertook crosses of very pure races of ducks, which differed 

 widely in size, viz, Rouens and Mallards. The two parent races did 

 not overlap in variability in size. The F! offspring were of intermediate 

 weight, as were also the F 2 offspring. The F 2 generation of 63 indi- 

 viduals included only 1 individual which fell outside the range of the 

 70 F! individuals, though the standard deviation of the ]^ 2 generation 

 was somewhat larger than that of the F! generation. Four body- 



