ALLEN. — THE HEREDITY OF COAT COLOR IN MICE. 117 



certaiu centres on tlie body, the appearance of albinism in certain body 

 regions would seem to be due to a weakening of certain of the centres 

 for pigment development. The writer inbred house mice for four genera- 

 tions in the hope of bringing about by this means some trace of albinism, 

 but none was produced. 



2. Heredity of partial albinism. Among mice and rats, partial albin- 

 ism tends to be inherited as a distinct character, so that when two pure- 

 bred spotted animals are mated, all the offspring are spotted, though 

 they may differ from the parents in respect to the precise degree of j)ar- 

 tial pigmentation. In a paper already mentioned, Castle and Allen (: 03) 

 have referred to these spotted mice as mosaics in which the characters, 

 pigmentation and albinism, are both visible side by side in distinct areas of 

 the animal's body, though when the two characters are brought together 

 by the mating of a totally pigmented and an albino animal, albinism is re- 

 cessive and does not appear in the soma. For this reason, such a spotted 

 animal might be regarded as a mosaic in which the two characters are so 

 united as not to be segregable in the formation of gametes. As Pro- 

 fessor Castle has suggested to the writer, it is conceivable that incom- 

 plete segregation may take place in a heterozygote containing both the 

 pigment and the albino characters. Such partial segregation would give 

 rise to gametes partaking of both these characteristics, and the union of two 

 such gametes would produce a partial albino. That spotted mice have 

 really had such an origin, however, remains to be proved. It is clear from 

 the experiments detailed in the preceding chapter that there is no evi- 

 dence for the general occurrence of any but the normal Mendelian 

 segregation of the two characters concerned. Moreover, the fact that 

 two pure-bred spotted animals breed true and produce no albinos, shows 

 that the albino character cannot become, as far as present experience 

 goes, again segregated into a separate germ. If this view be tenable, the 

 spotted or mosaic individual contains both the dominant and the recessive 

 elements side by side in its mature germ cells as well as its soma, and 

 may be symbolically expressed by the term DR. It may, however, 

 with equal propriety be regai'ded as simply a " mutation " (de Vries) 

 in which the production of pigment is for some reason more limited 

 than it is in the normal individual. Among plants, that variety of 

 Datura in which tlie fruits have prickles grouped in patches, instead of 

 being uniformly disposed, may perhaps be similarly regarded as exhibit- 

 ing a mosaic condition, since the thorny character is found to be dominant 

 in heredity over the smooth condition of fruit (Bateson, : 02, p. 23). 

 The fact that iu mice the spotted condition is one which cannot be pro- 



