158 MUTATION. X 



of chicks thus produced. They comprised 162 Silver males, 

 163 silver hens, 165 golden cocks and 160 golden hens. In most 

 of the farms nothing but golden chicks were kept, and these 

 were all pullets. He was told that the cockerels were weeded 

 out in the first week. When living near Paris we happened to 

 find out what happened to these cockerels. Every spring great 

 consignments of very young chicks arrive at the Halles there 

 from Holland. They find a ready market as baby chicks, but 

 never a single pullet is found among them. 



If we mate a female dove to an albino male, we find that all 

 the daughters are albinos and that the males are coloured. The 

 same state of things has been found by several authors in 

 quite a number of instances. From personal observation we 

 know that the following genes may show this repulsion from 

 the "female-determining" gene : That which Silver Wyandottes 

 and Silver Assendelver fowls have more than Golden-coloured 

 ones, that which Duck-wing bantams have more than Black- 

 reds, that which Brown-red English fighting bantams have 

 more than Black-reds. In pigeons one of us found one of the 

 genes necessary for pigmentation in the wild European turtle- 

 dove to be in this condition. By others a long list of genes have 

 been found to behave in this way. Such are the gene which 

 barred chickens have more than blacks, that which green ca- 

 naries have but brown ones lack, and the gene which by its 

 presence or absence makes the difference between normal 

 Abraxas grossulariata and its variety lacticolor. That ordinar- 

 ily a family of Silver Wyandottes, or of Barred Plymouth 

 rocks breeds true, must be due to this curious repulsion. On 

 the other hand, we observe that if at some day a pure-bred fam- 

 ily of these fowls does produce a Gold-coloured or a black hen, 

 we may not call such an occurence by the name of mutation, 

 because we do not at all know what cause lies at the base of 

 this repulsion. From our experiments with mice we know, that 

 in some cases the genes G and A are inherited quite indepen- 

 dently one from the other, and do not influence each other's 

 distribution over the gametes produced, whereas in other 



