184 



Wm. S. Marshall, 



it not for two things which it would be hard to reconcile to this 

 view. It was easy to assume that all the cells of a group came 

 from the repeated division of a Single cell and that, after each 

 division, the resultant cells remained attached to each other. When, 

 however, in a larva 13 mm in length, groiips of cells were found 

 similar, except in size, to the groups found in the ovaries of older 

 insects, I began to doubt as to their origin. Even in this young 

 larva, some of the groups contained as many as seven cells (Fig. 8). 

 The cells all come from a common central portion, each is pyriform, 

 and they are all attached by the narrov^ stalk like part. In Megachüe, 

 LuDw^iG (23) observed that the eggs have for some time a stalk, and 

 earlier, somew^hat similar appearances, had been observed by insects 

 other than Hymenoptera, Huxley (17), Claus (7), Lubbock (22). We 

 have already called attention to the very few dividing cells present 

 in the gonads of embryos, or of young larvae; in the larva we now 

 describe, but very few were noticed, and none of these occurred in 

 groups, as we shall notice in older larvae and in pupae. In young 

 larvae of Polistes we never found more than five or six dividicg 

 cells in any ovarian tubule; if all the cells in a group have come 

 from repeated divisions of a siugle one, it is certain, that some of 

 the numerous embryos and young larvae, would have shown many 

 dividing nuclei. Then again if these groups originate as above 

 mentioned, some dividing cells must be found near together, whereas, 

 as we have already said, up to this age the very few dividing cells 

 were scattered in the tubules. It might be said that the cells of 

 these groups arise by amitotic division; Giardina (8) has shown in 

 Dytiscus that the cells divide mitotically, and the few mitotic figures 

 Seen in Polistes^ would prove that this is also here the method of 

 division; that both mitosis and amitosis occur in the ovaries of these 

 young larvae, I do not believe. 



In Dytiscus^ each group consists of one oöcyte and several nurse 

 cells. In the group found in Polistes^ not only here but in older 

 larvae and in pupa, all the nuclei are similar, and there is no 

 difference in the size of the cells. We shall learn that, during the 

 growth of the oöcytes and the nurse-cells, they soon become easily 

 recognizable as such, and if, in older larvae, each group contained 

 one oöcyte and several nurse-cells, this would be apparent. We do 

 not believe that these groups of cells originate the same [way in 

 Polistes as in Dytiscus^ but cannot offer a very satisfactory explanation 

 to account for their origin and persistance in the former insect. It 



