INHERITANCE OF DILUTION. 79 



the hypothesis of triple allelomorphs, which we have found to agree 

 with the results of all the other crosses. The only possibility which 

 has not been eliminated is a linkage so close as to simulate a system of 

 triple allelomorphs. Unless exceptions occur which require it, such an 

 hypothesis need not be considered. 



Thus the data obtained by the writer are not only in harmony with 

 the theory that albinism, red-eye, and intensity form a series of triple 

 allelomorphs, but can be explained on no other basis, barring the pos- 

 sibility just noted. 



DILUTION. 



Such color varieties as agouti, black, brown, yellow, etc., are sharply 

 distinct from each other. They segregate from crosses without pro- 

 ducing intergrades and in unforced agreement with Mendelian expec- 

 tation. In contrast with these discontinuous variations are the con- 

 tinuous variations in the intensity of color of each main color variety. 

 Thus, among the yellows, there are all gradations from a pale cream to an 

 intense red. Among the agoutis, there are the pale silver agoutis, the 

 intense golden agoutis, and the intermediate yellow agoutis. There 

 are all grades of dilute blacks known to the fanciers as blues, for which 

 term, as has been explained, sepia is substituted in this paper. Finally, 

 there are all grades of dilute browns and cinnamons. (See plates 1 

 and 2.) 



The existence of these dilute types was noted by Castle (1905) and 

 Sollas (1909), both of whom also recognized that dilution could be 

 transferred from one series to another, e. g., from creams to blacks, 

 giving rise to sepias. They did not, however, suggest any factorial 

 explanation, finding the results of crosses highly irregular. Detlefsen 

 (1914) considered dilution to be recessive, but found the inheritance of 

 dilution very irregular among C. rufescens hybrids. He obtained 

 dilutes in F y after crossing dilute hybrids with a race of guinea-pigs 

 (brindle or 4-toe), among which dilution had never occurred and which 

 therefore should not carry it as a recessive. It may be remarked in 

 passing that the 4-toe race does contain albinism, which, with present 

 knowledge, satisfactorily accounts for these F x dilutes. 



Thus the difficulties in the way of an understanding of the heredity of 

 dilution have been due (1) to the intergrading of dilute with intense; 

 (2) to data which seemed to indicate that dilution could be due neither 

 to a recessive nor to a dominant unit factor, without complications. 

 Cross 39 gives many examples in which intense by intense has given 

 very dilute young, which seems to indicate that dilution must be 

 recessive if simple Mendelian at all. On the other hand, such cases 

 as that given by Detlefsen are difficult to interpret on this basis. 

 Further, dilute by dilute has often given young much more intense 

 than either parent. Thus, in cross 42-8, we have two medium sepias 

 producing a black. In cross 37 are many cases in which cream by cream 



