96 MR. AV. BATESON ON COLOUR-HEREDITY [May 26, 



Mendelian hypothesis to special cases, as with the formulation of 

 the specific laws followed by the several characters of ^ various 

 animals and plants in gameto-genesis and in union by fertilisation. 

 As in chemistry, these laws must be worked out separately case 

 by case, and each as it is determined has the value rather of fact 

 than of hypothesis. 



In dealing with this class of fact, special precautions are 

 necessary in order to establish the identity and purity of any 

 variety chosen for experiment. From the description of the 

 varieties of mice given at p. 73 it will be seen that some colours 

 may be easily confounded in description, though the pigments on 

 which they depend have a different chemical behaviour. This is 

 especially the case with regard to " fawn," " fawn-yellow," and 

 " yellow." In such cases it is absolutely necessary that the presence 

 or absence of cZar^ pigment should be noted, and that some attempt 

 should be made to distinguish the two dark pigments from each 

 other. 



In all attempts to trace laAvs of colour-heredity, colours of the 

 various parts will usually have to be reckoned with. In horses, 

 for example, the general bocly-coloui-, without that of the mane, 

 tail, and fetlocks is likely to be an insufficient guide to the 

 heredity. In man the heredity of eye-colour cannot be sufiiciently 

 investigated if it be separated from the colour of the hair, and so 

 forth. For the present, therefore, Pi-ofessor Pearson's conclusion 

 that the Mendelian system does not apply to coat-colour of horses 

 or to eye-colour in man should not be received without resei've. 

 By neglect of the precautions named above many results may be 

 described as conflicting with each other, which fui-ther analysis 

 would show to be harmonious. 



[A^ote added Juh/, 1903. 



When this communication was read I had not seen the im- 

 portant paper of Cuenot (27) dated March 1903. He states that 

 grey mice of his F^ when crossed with albino gave several blacks. 

 These, when bred with certain albinos produced black hetero- 

 zygotes, which bred inter se gave the Mendelian 3 black to 1 albino. 

 Some of these latter blacks were then homozygous, and from them 

 a pure strain of blacks was raised. This strain crossed with wild 

 mitiSGulus behaved as a simple recessive, giving grey Fj, with 

 Mendelian ratios 3 grey to 1 black, or 1 grey to 1 black in F^, 

 according as F^ was mated with F,, or with black. 



The fact that the original black did not appear in F^ or in Fg 

 calls for elucidation. It suggests a possible difference between the 

 albinos used either in producing F^ or later, some individuals 

 having the power of resolving the grey, while others ha-d not that 

 power. 



Cuenot next recoi'ds the new and important fact that the 

 coloui- of the Fj produced from his black strain x albino differed 

 according to the class of albino used. (1) Albinos extracted from 



