size. 43 



dimensions of the F x and F 2 ducks were also studied by Phillips, viz, 

 length of bill, tarsus, neck, and total length. Length of bill and length 

 of neck were slightly more variable in F 2 than in F : ; length of tarsus 

 was slightly less variable in F 2 than in F 1( while total length was more 

 variable in males but less variable in females in F 2 than in F x . It thus 

 appears that F 2 is not even uniformly more variable than F x in size char- 

 acters in this the purest material that had thus far been investigated as 

 to size inheritance among animals. Yet this supposed increase of vari- 

 ability is the only criterion of segregation in size crosses which has been 

 discovered or even suggested. Surely this is a wholly inadequate basis 

 on which to rest a theory that all inheritance is based on discontinuous 

 Mendelian factors. 



While the several investigations of size inheritance in rabbits, guinea- 

 pigs, and ducks were in progress, but before their outcome had become 

 apparent, the Peruvian expedition brought to the laboratory material 

 which seemed very favorable for such studies, and I have constantly 

 kept in mind its use in this way. Cavia cutleri from Peru gave us a 

 small race of undoubted purity, less than half the size of the guinea-pig, 

 but which has been found to produce fertile hybrids with it, which per- 

 mits obtaining an F 2 generation, a thing impossible with the rufescens 

 hybrids. The lea race and the Arequipa race have also afforded valu- 

 able material for size crosses with our own long-inbred and standardized 

 races of guinea-pigs. 



The results which have been obtained, so far as the demonstration 

 of mendelizing size-factors is concerned, are negative, like those previ- 

 ously obtained, though in some respects the material is more satis- 

 factory. But from their bearing on the question whether or not size 

 inheritance depends upon discontinuous Mendelian factors, these ob- 

 servations have, it is believed, several interesting features which will 

 become apparent as the description progresses. 



WEIGHTS AND GROWTH CURVES OF CAVIA CUTLERI, OF VARIOUS 

 GUINEA-PIG RACES, AND OF THEIR HYBRIDS. 



It will be recalled, from the description of the color inheritance 

 crosses, that cutleri males were crossed with females of two inbred races 

 of guinea-pigs, which we have designated races B and C respectively. 

 Many observations have been made on the weight of race B covering 

 the period from birth to old age. These afford an accurate knowledge 

 of the variability in weight of race B, and of the normal growth rate of 

 animals of this race. Our knowledge of the weight of race C is less 

 complete, though this race is equally inbred and appears not to be more 

 variable in size than race B. Its average size is probably a little greater 

 than that of race B, but the difference is negligible in comparison with 

 the difference of both from the size of C. cutleri. In the hybridization 

 experiments, race C hybrids were bred inter se, as were also the race 



