BY F. P. SANDES. 385 



the theca, at any rate for a time, may preserve its usual relations; 

 (see fig. 19). Occasionally, it appears that these cysts may remain 

 for a considerable time, but the majority lose their layer of epithe- 

 lium, and are obliterated by proliferation of the rudimentary theca 

 interna and invasion of the cavity by connective tissue cells, as 

 above described. It is probable that the two varieties of these 

 larger atrophic follicles are due to variations of the same process, 

 but the origin of the variation is obscure. The outer layer of 

 the theca undergoes the changes which have been described 

 previously in the consideration of the stroma of the ovary. 



In the case of the smaller follicles, the process is less compli- 

 cated, though similar. The vitelline membrane shrinks from the 

 membrana granulosa cells, the ovum degenerates and is remov^ed, 

 whilst the membrana granulosa cells may persist as a single 

 layer of cuboidal epithelium, or ma-y atrophy, when the theca 

 proliferates and fills up the cavity of the follicle. Occasionall}^ 

 as in the case of other animals, a metaplasia of membrana 

 granulosa cells into spindle- and star-shaped cells takes place. 

 These fill up the space and cause its obliteration (see fig. 21). 



In the case of follicles which are ripe or nearly so, whose ova 

 are not extruded, there takes place a quite different process, which 

 is not seen at all in the atrophy of the younger follicles. Practi- 

 cally, with the exception of the extrusion of the ovum, everything 

 proceeds in the same way as if rupture had taken place. A 

 corpus luteum atreticum is formed in the centre of which the 

 atrophic ovum is seen, sometimes even making an attempt to 

 segment (see figs. 22, 23, 24). The atrophied ovum is invaded by 

 connective tissue and is removed by leucocytes. The membrana 

 granulosa cells hypertrophy, the connective tissue of the theca 

 grows in, in the same way as in the ordinary corpus luteum, and 

 there is thus formed a corpus luteum atreticum. 



It will be seen, therefore, that in Dasyurus there is no difier- 



ence in the formation of the corpora lutea atretica, as compared 



with the mode of formation of the true corpus luteum. Some 



difference in size can sometimes be made out, the atresic being 



24 



