IX REPULSION AND COUPLING 91 



336 normal fertile, 



1 50 normal sterile, 



143 cretin fertile, 



1 1 cretin sterile. 



Had it not been for the small proportion of sterile cretins 

 the case might have been considered as one in which com- 

 plete repulsion occurred between the factors F and N 

 during the formation of the gametes of the Fj plant. The 

 presence of these sterile cretins may, however, be explained 

 on the assumption that the repulsion is not complete, but 

 that the gametes produced by the Fj plant form the series 

 NF: 3 Nf: 3 nF\ nf. The union of such a series of female 

 gametes with a similar series of male gametes as occurs 

 when the Y^ plants are selfed would result in a generation 

 consisting of — 



33 normal fertile, 



I 5 normal sterile, 



I 5 cretin fertile, 

 I cretin sterile, 



proportions which are almost precisely those actually found 

 by experiment. P'rom this and other experiments there is 

 reason for supposing that in cases where repulsion occurs 

 between two factors A and B the repulsion is not absolute, 

 but that the gametes are produced in one or other of the 

 series — 



AB : 2> Ab : 2, aB : ab, 

 AB : 7 Ab : 7 aB : ab, 

 AB • I S Ab : 1 S aB : ab, etc. 



Such series are in a sense the converse of those in which 

 partial coupling occurs, for in the one case it is the two 

 terms Ab and aB that are reduplicated, while in the other 

 it is in the terms AB and ad where this reduplication takes 

 place. A point of importance in connection with this view 

 of the nature of repulsion is that it enables us to bring the 

 phenomenon more closely into Une with that of coupling. 

 Whatever may be the meaning of repulsion and coupling, 



