4o8 BREAKDOWN OF GENETIC SYSTEMS 



This is not by any means inevitable, since the first timing abnormaUty 

 leads to others (Table 62, Fig. 124), and sometimes, as we shall see, 

 upsets the development of the spindle, which itself determines 

 metaphase. We find, however, that long " semi-mitotic " chromo- 

 somes are characteristic of several cases of reduced chiasma-forma- 

 tion (Table 62 (i) (a) ). 



In the absence of secondary abnormalities, the effect of geno- 

 typically controlled failure of pairing on the later course of division 

 is in no way different from that of structurally controlled failure of 

 pairing. They will be considered together later. 



(iii) Failure of the Spindle. In particular races of a number of 

 species pairing of the chromosomes at meiosis is complete and 

 regular, but the chromosomes or daughter nuclei fail to separate 

 at anaphase of the first or second division. Or if they separate, 

 in plants, a cell-wall fails to develop between them. Now, as Belar 

 and others have found, the spindle changes in shape in relation to 

 both the anaphase movement of the chromosomes and the formation 

 of the cell-wall. This relationship suggests that both depend on the 

 activity of the spindle. 



In Triticum the breakdown of the spindle leads to the formation 

 of numerous small nuclei. In inter-race hybrids of Drosophila 

 psendo-ohscura there is a genotypically controlled failure of pairing. 

 Correlated with this abnormality is a premature stretching of the 

 spindle which bends within the confined space of the cell. This is 

 sometimes a consequence of lack of pairing, as we shall see later. 

 But in this case it is not a consequence, for nuclei with complete 

 pairing have bent spindles. It is correlated with it inasmuch as it 

 occurs in the same organisms and therefore proceeds from the same 

 prior conditions. Probably it is due to the spindle developing too 

 early in relation to the chromosomes. It seems that to prevent the 

 spindle stretching and bending it is necessary not only to have 

 paired chromosomes, but also to have them there at the right time. 

 On the other hand, stretching does not occur in another hybrid in 

 this species when unpaired chromosomes are ready to divide and 

 form a plate at the first division ; that is, when the spindle is less 

 precocious in relation to the chromosomes. 



The opposite case of the spindle failing before the chromosomes 



