122 WHEELER. [Vol. VIII. 



The female larva, like the male, has no external orifice to 

 the sexual organs at the time of hatching. It is even more 

 backward than the male, inasmuch as the terminal ampullae of 

 the oviducts have not yet met in the median ventral line. The 

 first traces of a vagina were found in XipJiidijini fasciatuin 

 larvae about lo mm. long (Fig, 51). Here the terminal am- 

 pullae meet, but the surfaces of mutual contact are limited to 

 the pointed tips. The vagina {vg) is a short and broad in- 

 vagination of the hypodermis between the seventh and eighth 

 segments. Its tip extends to the juncture of the terminal 

 ampullae. In a somewhat later stage the ampullae open into 

 each other and into the vagina. The three pairs of gonapo- 

 physes {pp^-op'i) are already assuming their definitive charac- 

 ters. 



For an excellent resume of the little work that has been done 

 on the embryonic development of the sexual organs in insects 

 I would refer the reader to Heymons' recent paper ('9l). Here 

 I shall consider only three contributions, — two of Heymons' 

 ('90 and '9l) which treat mainly of the sexual glands, and Nus- 

 baum's paper on the development of the ducts ('84). 



Although the results of my study of XipJiidiiun, so far as 

 they go, agree in many respects with Heymons' account of 

 Blatta, several not unimportant differences must be pointed 

 out. The first difference relates to the stage in which the 

 germ-cells make their appearance. Heymons ('9l) claims that 

 he can detect them before the somites are established, at a time, 

 in fact, when the abdominal region of the embryo is still un- 

 segmented. This would correspond to a stage in Xiphidiiim 

 midway between B and C. In Blatta certain mesoderm-cells 

 at this time enlarge and assume a clear and succulent appear- 

 ance. There is apparently no definite relation between the 

 position of these modified cells and the future segments, and 

 even in a later stage when they become integral portions of 

 the somite-wall, they are quite irregular in their distribution. 

 Heymons regards them as largely dissepimental in position. 

 In XipJiidium, which is, on the whole, a far more favorable 

 form for the study of the sexual organs than Blatta, I was 



