68 



Mr. L. Hogben. 



Throughout the foregoing account no mention has been made of the 

 behaviour of the unpaired chomosome of the spermatogonia. As will become 

 evident later, this chromosome represents the X-element or accessory of other 

 forms. Throughout the entire synaptic period it can be identified as a club- 

 shaped body (figs. 11, 15), becoming abbreviated in the metaphase I. In every 

 case twelve chromosomes could be counted in the metaphase of the first 

 heterotype division : usually they are arranged so that a central group of three 

 is surrounded by a group of nine, including the accessory (fig. 22). In lateral 

 views the X-chromosome can never be seen to pass to one pole undivided, but 

 appears to undergo division rather later than the bivalents (fig. 21). 

 Metaphase counts, however, of the homotype division display in some cases 

 twelve, in other cases eleven chromosomes (figs. 23, 24). Their arrangement is 

 generally the same as in heterotype metaphase, i.e., an outer ring of nine (or 

 eight) and an inner group of three. Lateral views indicate that the X-element 

 passes to one pole undivided in the second reduction division of Libellula 

 depressa, as in Sympetrum and Anax (fig. 25). Smith thinks, however, that 

 this is not the case in Z. basalts : it would nevertheless be remarkable if such 

 a difference were confirmed in two species of the same genus. 



It follows that in Libellula depressa the second division is reductional for 

 the accessory chromosome. It has been stated by several writers of the 

 parasynaptic school that where both divisions are longitudinal with respect 

 to the diplotene filam.ents, it is impossible to identify one or the other as that 

 in which the segregation of the autosomes is effected. But unless- the 

 pachytene filaments rotate about their long axes in the contraction phase, the 

 line of cleavage in the diplotene stage must be in the same plane as that in 

 which the conjugating elements approximate. It is, therefore, most reasonable 

 in the absence of evidence to justify such a supposition to assume that the 

 mitosis which separates the longitudinal halves of the diplotene segments is 

 that in which the segregation of homologous chromosomes takes place. In 

 general the pre-reductional view is the more satisfactory : but in the case of 

 the Ascarid tetrad of which each joint corresponds to a longitudinal fissure of 

 the bivalent filament, it is possible that either pre-reduction or post-reduction 

 occurs. Possibly in the latter case the segregation of some of the bivalents 

 occurs in one division and some in the alternate mitosis. The genetic ratios 

 inferred would be the same. 



Synapsis in the Oocyte. 

 Attention was originally attracted to a study of the gametogenesis of 

 Libellula, on account of the remarkable phenomena described in connection 

 with synapsis in the oocytes of two other genera of Odonata by McGill. 

 Examination of the living ovaries showed that the nucleolus of the oocyte is 



