STUDIES IN THE EXPERIMENTAL ANALYSIS OP SEX, 167 
It is possible^ therefore, that the immature reproductive cells of 
hybrids may be affected by the fact of their hybridity more 
readily than the ordinary somatic cells, and that this sensi- 
tiveness may exist independently of the occurrence of 
synapsis and the reduction division. But this does not 
detract from the importance of the cases we have been con- 
sidering, where everything was normal in the hybrid until it 
came to the first reduction division. 
Theoretically, the facts brought out by Guyer and by this 
paper emphasise the fundamental importance of synapsis and 
the reduction division, and support the view that this process 
is a vital one in the orderly sorting out of the properties of 
the two germ-plasms derived from the two parents. 
A further theoretical consideration is suggested by the 
facts. The reproductive organs and even a large number of 
the spermatozoa of these sterile hybrids had the outward 
appearance of being normal, and yet, according to our inter- 
pretation, the spermatozoa were impotent owing to the 
abnormal distribution of the chromosomes to them. Is it not 
possible that the sterility of certain individuals within any 
particular species may be due to similar causes ; that there may 
exist intra-.specific hybrids, so to speak, which, owing to their 
combining in their germ-plasms incompatible qualities, are 
not capable of forming functional gametes ? Again, it is 
possible that the disturbances in the expected Mendelian 
proportions which are known to occur so frequently in intra- 
specific crosses may be due to the impotence of particular 
gametes carrying particular characters. If this were so, if 
the combination of certain characters in a germ-plasm led to 
abnormal distribution of the chromosomes during maturation 
and to the consequent impotence of certain types of gamete, 
we would obtain what in effect would be tantamount to 
selective fertilisation. At any rate it is certain that sorne- 
occur in many female animats at a very early stage, even in the embryo, 
long before the oocytes enlarge. If this is the case in birds we obtain a 
I’oady explanation of the complete sterility and abortion of the ovary in 
hybrid females, and possibly also of the high death-i'ate of hybrid females. 
