2G4 PROCEEDINGS OF THE ACADEMY OF [jNIarcll, 



sixteen larger and one smaller (N. 2). The sixteen larger ele- 

 ments in the first mitosis are all bivalent, and probably are all true 

 chromosomes; they all divide transversely. The two small ele- 

 ments (those marked JV. 3 in the figures) of this mitosis do not 

 divide, but one of them goes nndivided into the (me daughter 

 cell (second spermatocyte), the other one into the other — this 

 explaining why in the first spermatocytes there are eighteen ele- 

 ments, in the second only seventeen. On account of these small 

 elements not dividing, each of them must be considered univa- 

 lent ; for 60 far as my observations on the Hemiptera have gone, 

 all bivalent elements divide transversely in the first maturation 

 mitosis. 



The species of Capsidce thus far examined (compare the preced- 

 ing paper, l. c. ) show a remarkable agreement in the number of 

 their chromosomes. Thus, if we count each bivalent chromatin 

 ' element of the first maturation mitosis as two, there would be the 

 following number of univalent elements (counting in also chroma- 

 tin nucleoli) in this mitosis of the following species: Lyg us 2)1X11611- 

 sis, 34; Leptopterna dolobrata, 34; Calocoris rapidus, 33; Poecilo- 

 c((psus lineatas, 35?; P. goJiipJiorus, 34 or 36. There is not found 

 in the Capsidce such a disparity in the number of chromosomes as 

 is found between the species of some other families {e.g., the 

 Lygceldce and Coreidin), so that the Capsidce wouJd appear to be a 

 mure homogeneous group. Then if the number of the chromo- 

 somes may be loosely taken as a criterion of the degree of speciali- 

 zation, a smaller number of chromosomes marking a more special- 

 ized stage (and this I hold to be true within certain bouuds), the 

 Capsidce, like the Reduviidce and Phymatidce, may be considered 

 relatively prindtive Hemiptera heteroptera, in comparison with the 

 Pentatomidcc, Lygceidce and Coreidce. This, it seems to me, is a 

 vital interest in the study of the chrotnosomes — to find critei'ia for 

 testing relationships. 

 6. Nabis annulatus Kent. 



I had only a single testis for examination, and it showed no 

 spermatogonic mitoses. 



There is no complete rest stage in the growth period of this 

 species (in which regard it is like certain of the Coreidce). In 

 the late telophase there is found in the nucleus (PI. X, fig. 11) a 

 large, usually centrally placed chromatin nucleolus (iNT. 2), with 



