1903.] IX FANCY MICE AND RATS. 87 



some of the coloured individuals contained no albino. As far as 

 the few observations went, the extracted albinos gave only albinos. 

 80 far therefore the Mendelian hypothesis harmonises well with 

 the phenomena. 



When, however, we begin to consider the relations of the 

 several colour-types to each other, we meet some important 

 problems. The original waltzers are described as black-and-white. 

 Of what pigments the black was made up we do not know. 

 Probably it contained both the black and the brown elements. 

 However this may have been, the I'eversionary heterozygote 

 clearly did, though whether it also contained the yelloio pigment 

 is not so clear. 



On examining the details as to the offspring of the several 

 pairs, it appeal's that though the self-greys may, as the first cross 

 proves, contain all the other 3 coloured types and the albino, the 

 grey-and-white contain the albino only. Similarly the black-and- 

 white can only contain the albino, so far as the evidence goes. 

 But black seems to be dominant over black-and-white. 



The facts are not sufficient to make these deductions quite 

 certain ; and, in particular, since the evidence in rats shows that 

 grey-and-white may dominate over black-and-white, it may be 

 mei'ely from accident in the choice of individuals that no black-and- 

 white was produced by any of the grey-and-white mice. 



The families from the 4th pair on Table I., and from the 3rd 

 pair on Table II., are especially interesting as giving indications as 

 to the gametic ratios in a complex case, though the evidence is 

 insufficient to determine these ratios. In the first case black- 

 and-white X black gave 10 black, 15 black-and-white, 7 albino. 

 Both parents were heterozygotes containing albino, being each 

 raised from self -greys x white. From the facts it is clear that one 

 parent at least was giving oflF gametes black, black-and-white, and 

 white ; and from the indication that black is dominant to black- 

 and-white, it is probable that this parent was the black. The 

 simplest supposition is, then, that the black-and-white gave oflT 

 blw. and w., and that the black gave off bl. and blw. in equal 

 numbers, and whites equal to their sum. This distribution would 

 give the ratio 



1 bl, : 2 blw. : 1 w., 

 and where experiment gave 



10 bl. : 15 blw. : 7 w. 

 we should expect 8 bl. : 16 blw. : 8 w., 



which fits well. But in the 3rd pair on Table II. we have a blk. x 

 albino giving 



7 bl. : 16 blw. : 20 w., 



where, on the hypothesis suggested, we should expect equality 

 between bl. and bl.w., and the discrepancy is considerable and 

 emphatic. 



Pending further experiment, the relations of bl. to blw. and to 

 white cannot be stated with any confidence. Another point 



