176 INHERITANCE IN RATS. 



The pink-eyed yellow and red-eyed yellow of rats are complementary 

 loss variations ; for when the two varieties are crossed with each other 

 they produce F : offspring which are either gray or black pigmented, 

 according as their yellow parents did or did not transmit the agouti 

 factor. These F! reversionary grays or blacks are paler in pigmenta- 

 tion than ordinary gray or black rats, indicating that neither character 

 in a heterozygous form is the full complement of the other. But it is 

 evident that in homozygous form each is the full complement of the 

 other, since in F 2 and later generations grays and blacks of full intensity 

 are obtained. 



The FI black-eyed animals (blacks or grays) obtained by crossing 

 pink-eyed yellows with red-eyed yellows, if mated with each other, pro- 

 duce an F 2 generation containing (1) black-eyed young (black or gray), 

 (2) red-eyed yellow young, and (3) pink-eyed yellow young. We have 

 obtained thus far 324 such F 2 young, of which 162 were of class (1), 90 

 of class (2), and 72 of class (3). 



If, as suggested, red-eyed yellow and pink-eyed yellow are due to 

 mutually independent Mendelian factors, then F 2 should contain four 

 classes instead of the apparent three; wherefore it seemed probable 

 that one of the three classes was really composite and that the three 

 should be as 9:3:4. On this basis the F 2 expectation would be 

 182 : 61 : 81 instead of the observed 162 : 72 : 90. Hence there appear 

 to be fewer black-eyed young than are expected. Further, when we 

 came to test the other F 2 classes to discover which of them was com- 

 posite, we found very few individuals which would fall in the hypotheti- 

 cal fourth class transmitting both pink-eyed and red-eyed yellow in the 

 same gamete. Instead of 1 in 16 as expected, we have been able to 

 discover a much smaller number of double recessives. Both a defi- 

 ciency in double recessives and a deficiency in double dominants (the 

 black-eyed class), which have been observed among the F 2 rats, would 

 be expected if pink-eyed yellow and red-eyed yellow are due to " linked 

 genes," i. e., to factors located near each other in the germ-plasm. For 

 in the cross under consideration each form of yellow enters the F! 

 zygote in a different gamete. Hence, in the gametes arising from such 

 zygotes we should expect the two forms of yellow to show mutual 

 repulsion. If they did so, then the gametes formed by F! zygotes, of 

 the four possible combinations, RP, Rp, rP, and rp, would not be 

 equally numerous, but Rp and rP should be more numerous than RP 

 and rp. That this is true is indicated by the facts presently to be 

 stated. To test the gametic composition of the F 2 yellows, those which 

 were red-eyed were mated with pink-eyed yellows of pure race, and 

 those which were pink eyed were mated with red-eyed yellows of pure 

 race. For it was known that, since red-eyed yellow is a recessive 

 variation, every red-eyed F 2 yellow must be homozygous for red-eye, 

 but conceivably it might be either heterozygous for pink-eye or might 

 lack it altogether. A cross with the pure pink-eyed yellow race would 



