26 BTLLETi:^ 1090. r. 5. DEPAET:MEXT or AGEICrLTrRE. 



which were derived in part from another source were at least as vig- 

 orous as the others, as is shown in Part II. It would thus be a 

 remarkable coincidence if the animals selected to found the control 

 stock happened to have so much more hereditary vigor in every re- 

 spect than those selected to found the inbred families, especially as 

 it will be shown later that vigor in one character is not correlated 

 with vigor in others. It thus seems improbable to the writer that the 

 differences can be explained on the grounds of an initial genetic infe- 

 riority in the foundation stack of the inbred families. 



Another possible explanation of the difference between inbreds and 

 controls is in the greater average age at which the latter were mated. 

 During 1916 and 1917 the differences in vigor were even more marked 

 than previously. During these years the average age of inbred dams 

 was 14.1 months and that of control dams 15.5 months. This differ- 

 ence is altogether too small, considering the slight eflFects of age on 

 the various characters, to give the controls any appreciable advantage. 



The average age of the inbreds at their first litter was 5.9 months. 

 About two-thirds of the controls were immature when mated. Their 

 average age was 5.2 months at the first litter. Most of the remaining 

 control matiugs were between more or less immature guinea pigs. 

 These considerations, therefore, merely indicate another character 

 in which the controls were more vigorous than inbreds, namely, age 

 of maturity. 



The most probable interpretation of the differences between the 

 inbreds and controls, is, therefore, that the inbreds started at about the 

 same level of vigor in all respects as the controls, but declined in the 

 course of time as a direct or indirect result of inbreeding. Here, 

 however, the objection arises that the controls have also declined since 

 1911, and at almost the same rate as the inbreds. If this decline 

 among the inbreds is genetic, so, it would seem, must be the decline 

 among the controls, while if the latter is due to environmental con- 

 ditions, so must be the former. The decline among the controls, how- 

 ever, was probably not genetic. It may be urged to the contrary 

 that the rigid system of mating by pedigree prevented selection of 

 animals for vigor as effectually as in the inbreeding experiments. 

 This is true, but the characteristics with which we are concerned 

 depend on heredity to such a slight extent that any selection of indi- 

 viduals should be wholly without appreciable effect, one way or the 

 other, in the short time in which the experiment has been in progress. 

 So far as we know at present, there seems to be no valid reason for 

 assuming any genetic change in the control stock. This being the case, 

 the decline since 1911 in both control and inbreds must be environ- 

 mental. The excellent record made by both stocks in 1919 and 

 1920 confirms this view. We thus reach the seemingly paradoxical 

 conclusions that the inbreds were falling off genetically from the 



4 



