86 MR. W. BATESON OX COLOUR-HEREDITY [May 26, 



Variegated black-and-white X chocolate-and- white gave on one 

 occasion {Steer) 5 black-eyed whites and 1 chocolate-and-white. 



As mentioned above, blue may be recessive to black and breed 

 true from its first appearance, and will doubtless prove to be a 

 homozygous colour. 



We may now pass to a consideration of the crosses made with 

 " Japanese " waltzing mice. The exact physiological nature of 

 the waltzing habit seems to be still uncertain. Reference to the 

 work especially of Cyon, Rawitz, and Zoth shows that, though 

 malformation of the labyrinth is not infrequently associated with 

 this condition, at least the degree of the structural malformation 

 varies considerably. 



The origin of the variation is still more obscure. Mouse-fanciers 

 have assui'ed me that something like it may appear in strains inbred 

 from the normal type, though I cannot find an indubitable case. 

 Such an occurrence may also be nothing but the appearance of a 

 rare recessive form. Certainly it is not a necessary consequence 

 of in-breeding, witness von Guaita's long series of inbred albinos. 

 From analogy with other cases, we should be prepared to find 

 that the existence of such a structural feature in one of the gametes 

 had an efiect on the colour of the heterozygote ; but the evidence, 

 as we shall see, is on the whole unfavourable to this view. 



As to crossing of waltzers and albinos, the earliest evidence is 

 that of Haacke, whose records are qualitative only. Crossing 

 waltzers, blue-grey with white marks, and albinos, he obtained 

 mice generally self-gTey (? agouti), more rarely self-black. Their 

 offspring occasionallv h9.d a small white mark on the ventral 

 surface. 



The next large body of evidence is that of von Guaita (19), who 

 used black-and-white waltzers with dark eyes (von Guaita inlitt.) 

 and an inbred strain of ordinary albinos. From this cross, F^ was 

 always (fi-om 4 paii-s) a self-coloured house-mouse, and was also 

 like that wild type in size (being larger than the waltzer and 

 smaller than the albino) and in wild disposition. F^, raised from 

 F^ bred inter se, consisted of albinos and 4 coloured types — black, 

 grey, black-and-white, grey-and-white. The totals were 30 

 coloured, 14 albinos. On the expectation of 3 : 1 there should 

 have been 33 and 11, so that the excess of albinos is distinct, 

 though the numbers are small ; but when all certain cases of 

 DRxDR (taking albino as R) are included, the numbers are 

 117 coloured and 43 albinos, coming very near indeed to the 

 expectation 120 to 40. There can therefore be no doubt that 

 the heterozygotes produced on an average equal numbers of 

 albino gametes, and of gametes bearing the various colour-types. 

 There are only two matings certainly in the form DR X R. 

 These gave 23 + 1 coloured, 20 + 1 albinos, closely approaching 

 the expected equality. 



In (20) Table I., from 1st and 5th pairs, we have families of 17 

 coloured and 13 coloured respectively, showing pretty clearly that 



