2ü8 Mr Jiatexoii's Revis'iDiis qf MendeVs Thcori/ <>f Ileredity 



gametes of niore than oiie kind, this niay have happened in either of" two ways: 

 there may have been two kinds of gametes, ono kind produced by each sex, as 

 supposed by Mr Bateson in the case of the Horse or the Ass, in which case the 

 race wouid breed true, every individual being produced by a heterozygote which 

 lost half its heritage during developnient. Sucl» a form of gamctic heterogeneity 

 would be consistent with tho observed fact that the mice bred true, but it would 

 involve difference between the crosses Albino % x waltzer ^ and waltzcr ? x 

 Albino (/■ which has not been observed, and it would fail to aecount for the 

 observed difforences between individual hybrlds produced by the saine parents. 

 The hypothesis that each mouse of either sex produces gametes of more than 

 oue kind involves either the hypothesis that these gametes do not obey Mendcl's 

 law at all, or that a constant percentage of " rogues," dominant or recessive in one 

 or more elementary characters, is produced in each generation. No such rogues 

 have been observed either in a very considerable number (many hundreds) of 

 young cxamined before Mr Darbishire's experiments were begun or in pure 

 waltzing families raiscd during the course of the experiments : so that this 

 hypothesis also falls to fit the facts. 



The second new assumption raade by Mr Bateson, that coat-colour is a 

 "Compound" character, is probably true; but it corapletely falls to aecount for 

 the observed variability among the offspring of the first cross, since no " resolution 

 of Compound characters" can occur during the formation of the first hybrid 

 generation, and it involves a reduction of the number of ))ure recessive albinos, 

 among the offspring of hybrids, to a maximum far bclow the observed proportion. 

 Mr Bateson's new hypothesis therefore fails to fit the facts any better than his first. 



I am as anxious as Mr Bateson can be to test the possible ways of bringing 

 the behaviour of these mice into accord with Mendels " i)iinciples " or indeed 

 with any other theory of heredity. The two hypotheses already put forward fail, 

 as we have seen. I earnestly appeal to Mr Bateson, now that the facts are fully 

 before him, to produce somo final formula, expressed in terms of catogories so 

 definite that they really describe the mice included, which seems to him cjipable 

 of bringing the behaviour of these hybrid mice and their offspring into harmony 

 with the "principles" he so strongly supports. 



[A letter to Nature, May 14, containing a new hypothesis, has appeared 

 while this article is going through the press. This letter and a paper before the 

 Zoological Society of London, of which Mr Bateson has given notice, could not 

 be discussed in the present number of BiometHka without unduly delaying its 

 publication.] 



