70 



Mr. L. Hogben. 



later. The meiotic phenomena of Lihellula depressa will be described first, 

 and the question of the " chromidia " discussed in due course. 



The ovaries in Libellula, as in Anax and Plathemis, are a pair of spindle- 

 shaped capsules, enclosing numerous minute ovarioles. They are thus of the 

 type found in Gryllus, consequently, they are not so amenable to very 

 delicate fixation as are the freely-exposed egg tubes of Periplaneta. McG-ill's 

 material was fixed whole — after dissection in fluid — in Flemming's reagent 

 (strong formula). Comparison with preparations of Periplaneta showed that 

 this technique is totally inadequate when attempted with the Libellula 

 material. Improvement was obtained by adopting three precautions : first, 

 the ovaries were removed without immersion in fluid and plunged directly in 

 the reagent ; secondly, they were teased out quickly and thoroughly, so that 

 €ach ovariole was separately and instantaneously fixed ; lastly, the fixing fluid 

 (Flemming's) was diluted to a third the usual strength. When these 

 precautions are observed, the results are in every way more satisfactory. 

 Mutatis mutandis this applies to material preserved in chromosmic without 

 acetic acid, and by the Mann-Kopsch method {vide infra) for the examination 

 of mitochondria. 



The ovarioles of Libellula, Anax, and Plathemis are of the simple type 

 found in Gryllus and Periplaneta : the greater proportion consists of oocytes, 

 progressively arranged with respect to size in linear series. The remainder 

 consists of young oocytes at the inception of growth, lined side by side, a 

 mass of oogonia and oocytes in synaptic stages, and a terminal filament of 

 smaller cells. There are no nurse cells : the follicle cells arise from the 

 primary oogonia and divide mitotically. A few authors have described 

 amitotic division in follicle cells of Hexapoda, but in the majority of cases, 

 probably all cases in which the follicle cells are germinal in origin, they 

 divide in the usual way. 



To obtain mitotic stages of oogonia and oocytes in synapsis, the second-year 

 nymphs provide most suitable objects for study ; for observations on yolk 

 formation, the third-year nymphs, immediately before emerging, must be 

 obtained. Oogonial mitoses are rare, and the chromosomes of the follicle 

 cells are small and crowded. Hence it follows that the female germ cells do 

 not provide suitable material for investigating the phenomena of nuclear 

 division. A limited number of very clear oogonial anaphase, telophase, and 

 prophase figures were obtained in the preparations studied, and in the main 

 the series of phenomena seems to be essentially identical with what occurs in 

 the spermatogonia. That is, in anaphase the chromosomes become more bent 

 as they pass undivided to the poles of the spindle ; in telophase they spin 

 out into loops that are apparently polarised ; in prophase the individual 



