1905.] of Permanent Forms among Chromosomes, etc. 569 



of each type have a constant numerical relationship to each other. Thirdly, 

 so far as the investigation has at present gone, certain types of gemini appear 

 to be common to all the widely sundered forms examined. Still further, it 

 will be seen that the number of different types of gemini is less in the oldest 

 evolutionary form Periplaneta. 



Whether this last indication will be found to hold, good is a matter upon 

 which it would at present be useless to speculate ; but the fact itself opens 

 up a line of future inquiry which is certainly full of possibilities. 



It seems to us, moreover, that it should be emphasised that both in regard 

 to the permanent types of gemini and their numerical relationships, as well 

 as with respect to the numerical constancy in the chromosomes themselves 

 and their periodical reductions, we are face to face with constant arrange- 

 ments in the parts of the unit of living substance (the cell) which seem to 

 underlie and to be quite independent of those external interactions that are 

 supposed to have helped to build the grosser features of living things. 



With regard to the different types of gemini, it should further be pointed 

 out, that the existence of these types implies substantive differences between 

 the chromosomes that can unite to form the different kinds. It must be 

 remembered that each of the gemini arises through an association of optically 

 similar premaiotic chromosomes, but that at the time the nuclear membrane 

 is about to disappear these associations have assumed different forms. They 

 cannot do this unless they are of a different nature. The fact that there exist 

 in those nuclei which we have examined groups of similar gemini shows that 

 there must be sets of premaiotic chromosomes which in the synapsis can 

 conjugate with each other, but not with the remaining individuals. 



The present position may be in part summed up as follows : — In the 

 fertilised egg the paternal and maternal chromosomes divide independently on 

 the spindle of the first segmentation figure. And they go on dividing in a 

 similarly independent manner throughout the soma, and during the pre- 

 maiotic history of the reproductive elements themselves. In the synapsis 

 which ushers in the maiotic phase the chromosomes unite in pairs, and in 

 those cases we have as yet examined only certain individual chromosomes 

 are capable of uniting with one another to form differing group of gemini ; 

 in each of these groups the number of gemini is more than one, and it varies 

 in the different species hitherto observed. 



Thus whether the conjugation of the chromosomes in the synapsis is 

 really the final consummation, after many generations long delayed, of the 

 copulatory intentions of the paternal and maternal elements, is a matter 

 upon which there is as yet no actually conclusive evidence. 



