INHERITANCE OF MINOR VARIATIONS IN INTENSITY. 87 



However, both of these tests fail. No single gamete stock of dilutes 

 has been found which will not give the entire range of variation when 

 tested. Thus, male D30 redo, an intense which carried dilution as a 

 recessive (CCd), was crossed with red-eyes. His dilute young must all 

 owe their dilution to the same single gamete. They ranged from D340 

 blacki to D 152 sepia 5 . Yellows which owe their dilution to this same 

 single gamete (derived from male 00 creame CaC a , the father of D30) 

 range from D391 yellow 2 to 00 creamy . Dilution from a single gamete 

 of A674 sepiae (CdC a ) has given rise to D652 black! and M306 sepia?, 

 D409 yellow 3 , and M199 cream 7 . This last case involves no admixture 

 of South American blood. Inspection of the tables will yield many 

 similar cases. Evidently dilution from a single gamete may appear hi 

 dilutes of any grade of intensity. The extreme variations may occur 

 within a single litter (offspring of D30). Again, many examples can be 

 given in which the offspring are much darker than either parent. D652 

 blacki was the offspring of D215 sepia 3 and D106 sepia 4 . In cross 37 

 there are 6 cases in which creame X cream^ has produced yellow 3 , with 

 other less extreme cases of this kind. These results do not demonstrate 

 that no more than four allelomorphs in the albino series are present in 

 our stock. They do show that there are other causes producing varia- 

 tion of much more importance than any other allelomorphs which may 

 be present. 



THE RELATIONS OF IMPERFECT DOMINANCE, STOCK, AND AGE TO 

 GRADES OF INTENSITY. 



In tables 37 and 38, and diagrammatically in figure 5 and figure 6, all 

 records of grades of dilution at birth are analyzed with respect to 

 genetic constitution and stock. All of those whose genetic constitu- 

 tion was known with complete or nearly complete certainty, either from 

 parentage or from offspring, are put after the proper formula, CdCd, 

 CaC r , etc. All from litters containing two classes are listed separately 

 with the numerical expectation of the classes as (20 CdCd : 32 CdC a ), etc. 

 Those hi the litters whose formulae were later determined by a test 

 mating are given below in parentheses. These tested individuals are 

 included both among those of certain constitution and hi the litters 

 containing two classes. No very close analysis of the influence of 

 stock was possible from the data obtained. However, the following 

 stocks were recognized: 



Dil., Dilute selection stock. 



Misc., Miscellaneous stocks with but little BW blood and no lea or Arequipa blood. 



These contained much dilute selection and 4-toe blood and some C. 



rufescens ancestry. 



$BW, FI from the cross of miscellaneous with BW stock, 

 f BW, Back-cross of $BW with BW stock. 

 S. Am., All animals with lea or Arequipa blood, in most cases about J South 



American, J BW, and i miscellaneous, but including pure lea, IcaXBW, 



etc. 



