R. C. PUNNETT 313 



puzzling in character." It is however what we should expect on the 

 hypothesis that I have outlined above, and since this case offers a good 

 test between Castle's interpretation and my own we may examine it in 

 more detail. 



Let it be supposed that the self-coloured rabbit is homozygous for 

 the three following factors: 



P, the factor for increased pigmentation (as explained above). 



T, a factor which in the presence of S leads to the production of a 

 Typical Dutch. We may further suppose that when S is not present T 

 leads to some increase in the pigmentation of the White Dutch. 



S, a factor which, when added to White Dutch, leads to an increase 

 in the pigmentation, and the formation of the Spotted Dutch character. 

 The heterozygote Ss is intermediate in form. 



The F^ rabbits ex Self-coloured x White Dutch will be PpTtSs in 

 constitution, and will give rise to the series of gametes PTS, PTs, PtS, 

 Pts, pTS, pTs, ptS, pts. The zygotic series produced by two such 

 series of gametes is represented on Fig. 2, on which are also indicated 

 the grades to which the various zygotes may be referred. It will be 

 understood of course that any rigid system of assignment is impossible ; 

 for, after all, the grades made use of are but convenient symbols. In 

 making the various assignments I have been guided as far as possible by 

 the experience gained through many varieties of mating extending over 

 a period of more than ten years. In Table I the various grades have 



TABLE I. 



Grades ' 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 



Distribution ] 

 calculated j- 36 43 20 16 10 6 7 6 8 5 3 2 2 5 8 8 6 1 

 from Fig. 2 ) 



Actual ) 

 figures from j- 41 45 22 14 9 5 .5 4 5 5 8 4 8 .5 6 4 1 — 

 Castle j 



been collected together from Fig. 2, and brought into tabular form 

 in order that the hypothetical result may be compared with Castle's 

 actual figures^ {Table 28, p. 4^). The general agreement is evidently 

 close, and it is clear that the experimental data are more in accord- 

 ance with the interpretation I have suggested than with that which 

 Castle has put forward. 



1 The number of rabbits in Castle's F^ generation is 191 ; accordingly I have multiplied 

 each zygote in Fig. 2 by 3, bringing the total to 192. Where a zygote is given as of 

 several adjoining grades each is represented in Table I. Thus where a zygote is given in 

 Fig. 2 as of grades 4-6, I have reckoned one of each of these three grades in Table I. 



