62 



Mr. L. Hogben. 



to be reinforced by the study of the phenomena cited. Despite the unusual 

 character of the process described and the interpretation inferred, no other 

 investigator has repeated this work. 



In 1908 Lefevre and McGiir(4) repeated the earlier work of the latter on 

 the spermatogenesis of Anax Junius. In their amended form the chief points 

 of interest recorded are as follows : (a) the diploid complex of the male germ 

 cells in A. Junius consists of twenty-seven chromosomes (including twelve 

 pairs of rod-like, one accessory, and a pair of micro-chromosomes) ; (b) the 

 chromosomes of the spermatogonial telophase form a continuous spireme; 

 (c) the long axis of the tetrad is equivalent to the line of cleavage of the 

 diplotene filament ; (d) the X-element is segregated in the homotype mitosis. 



More recently, Smith, 1916 (5) has made a careful study of the male sex 

 cells in Sympetrttm semicinetum and Libellida basalis. In both species the 

 spermatogonial metaphase displays twenty-five equal chromosomes ; the 

 tetrads are stated to be formed like the rings of the Periplaneta type and to 

 divide in the same way ; the accessory is said to pass over undivided in the 

 first reduction division of L. basalis, but in Sympetrium to behave as 

 in Anax. 



The Spermatogonia. 



From the examination of gonads taken from spring and autumn nymphs, 

 it appears that L. depressa does not complete its life cycle in less than two 

 entire years. Most favourable material for the study of spermatogenesis 

 is obtained from nymphs in the early spring immediately preceding the 

 final moult. The testes then consist of cysts of spermatogonia undergoing 

 mitotic changes and of spermatocytes ; a remarkable feature of the species is 

 the very short duration of the resting stage of the nucleus. In very young 

 cysts, composed of only a dozen spermatogonia, the dimensions of the 

 nuclei considerably exceed those of the nuclei of cysts of numerous cells ; 

 and, as the cells are in any case small, the former provide most suitable 

 objects for observation. While it would appear that the divisions follow 

 rapidly upon one another, a circumstance which facilitates observation, it 

 is rarely found that the cells of any particular cyst differ to any marked 

 extent in the phase of nuclear history which they display ; this fact is again 

 helpful to an intensive study of individual stages. The material employed 

 was fixed in Flemming's reagent (either with or without acetic acid), diluted 

 with two volumes of aq. dest. For staining, thionin, Gentian violet, and 

 Haidenhein's iron-alum haematoxylin were used. 



The history of the nucleus of the spermatogonial cells may be commenced 

 most conveniently at the metaphase. As more than a hundred individuals 

 were obtained for this purpose, an abundance of preparations, with all the 



