R. C. PUNNETT AND P. G. BaILEY 33 



940 up to 1710 grams, and lo $ $ between 765 and 1190 grams. The 

 wide range of variation accorded with the nature of these two hens sent 

 by Dr Hagedoorn, and this was confirmed by the following year's work. 

 The largest and the smallest pair were mated up in 1912. From the 

 largest pair (Pen 14, 1912) were produced 12 ^/c/ and 7 $ $. The 

 average weight of the (/</ was not very different from that of the 1911 

 family, but the 7 $ $ were a distinctly heavier lot (cf. Table I). The 

 smallest pair of birds (Pen 4, 1912) produced 7 </</ and 5 ^ $. As 

 the table (p. 26) shows all of these birds were small, the majority being 

 smaller than the parents. In the case of the pullets the range of varia- 

 tion of those from the small parents is entirely outside the range of 

 those from the large parents ; in the case of the cockerels the two 

 ranges just overlap. These experiments were discontinued after 1912 

 in favour of the Hamburgh-Sebright cross. The general results how- 

 ever, the great range of variation in F.;,, and the fact that the extremes 

 tend to breed true to size, offer an interesting corroboration of the 

 Hamburgh-Sebright story. 



We may now proceed to suggest a scheme by which the phenomena 

 we have described may be interpreted in terms of genetic factors. 



H ypothetical Explanation. 



In discussing such a tentative scheme several distinctive features 

 about the results must be borne closely in mind. Among them the 

 most important are : 



(1) The fact that from the Sebright-Hamburgh cross the range of 

 variation in F^ is beyond that of either of the original parent forms. 

 The smallest birds are definitely smaller than the Silver Sebright and 

 the largest are well above the weight of the Gold Hamburgh. This 

 phenomenon was more marked in the F^ generation when the smallest 

 and largest F^ birds were bred fi'om. It must therefore be supposed, 

 in devising an explanation in terms of genetic factors, that the Gold 

 Hamburgh does not contain all the factors concerned, and that the Silver 

 Sebright does not lack them all. 



(2) The Fi birds are not quite so large as the Gold Hamburghs. 

 They must however contain a single dose of all the factors concerned. 

 Hence we must suppose that the factors with which we are dealing 

 produce a less marked result when the bird is heterozygous for them 

 than when it is homozygous. 



(3) Extreme variants in F^ are relatively scarce. Indeed the 

 smallest and largest extremes as shown by the F^ generation did not 



Journ. of Gen. iv 3 



