W. F. 1?. Wkldon 295 



darkness of the hybrid eye inay be oxhibited as a " property of heterozygotes," and 

 so neglected, with the othcr inconveniont properties of first Grosses, in applyiug 

 the "great fact of ganiotic piirity." Mr Bateson accordingly writes {Nature, 

 No. 1742, March 19, 1903) "The first cross shows that when in this case an albino 

 " (pink-eye) gamete, G, meets a colour-boaring (pink-eye) gameto, G', in fortilisation 

 "we niust expect the rcsulting heterozygote, GG', to bc coloured in coat, with a 

 "dark eye." Now if this iiieans anything morc tlian a niere rc-statement of 

 Mr ]")ai'bishire's resuit, it meaiis tliat so far as coat-colour and eyc-eolour are 

 concerned the elemcnts in each gamete are to be treated as single units; they are 

 either actually singh-, or they are " eoni|i(Miiiil allclomorphs." 'l'hr atteniptcd 

 explanation, even on this view, only succeeds in piacing the phenoniena in the 

 sanie category with many otbers avowediy inexplicable, but it does avoid the 

 difficulty of the origin of a new or atavistic character from the uniou of two siniiiar 

 eyc-eolour elenu'nts. 



The coat-colour of the hybrids is still uiioxplained by Mr Bateson's statement; 

 for since every hybrid by hypothesis results froui the union of siniiiar pairs of 

 gametes, the hybrids themselves should be similar. Mr Bateson has hitherto 

 assumed in all his discussions that whatever the characters of similar hetero- 

 zygotes may be, at least they are similar ! In this case, he passes over the difficulty 

 in silenee. To me (and I am glad to know that Mr Darbishire agrees with me) 

 this variability of the hybrids is highly significant, when these results are compared 

 with those obtained by others. Using a race of albinos, known to be absolutely 

 pure because they had been inbrcd for 29 generations, and crossing these with 

 pure-bred Japanese waltzing mice with dark eyes and black-aiid-white für, von 

 Guaita teils us that he obtained hybrids which uniformly resembled wild mice ; 

 making a similar cross, but using albinos of which no definite history is recorded, 

 Haacke obtained hybrids sometimes like wild mice, and sometimes black. 

 Mr Darbishire has deliberately used albinos which were sometimes pure-bred, 

 sometimes known to have near ancestors of various colours, and he obtains a 

 highly variable group of hybrids. Now the gametic purity of an extracted re- 

 cessive form which breeds true cannot be questioned without abandoning the whole 

 Mendelian position, or any modification of it, and therefore all these albino mice 

 must be treated as equally pure, on Mr Bateson's view ; but I submit that the 

 three sets of facts aheady published by Mr Darbishire, by von Guaita, and by 

 Haacke, show a correlation between the anceatry of the pure recessive albino and 

 the character of the hybrids, sufficiently strong to disprove the "great fact of 

 gametic purity " in this case, at least. The correlation between the colour of the 

 coat in dominant hybrids, and in their hybrid grandchildren, which Mr Darbishire 

 has shown to be established by transmission through a pure recessive albino, is 

 still more conclusive evidence against the gametic purity of recessives, {siipra, 

 pp. 282-.5). 



So much for the resuit of the first crcss. What happens when the hybrids 

 of this generation are paired together? Mr Bateson represents the resuit as 



