392 BOWEN. 



The carrying out of this study on a comparative basis was made 

 possible by the comprehensive collection of preparations of hemipteran 

 germ cells which Professor E. B. Wilson had accumulated during his 

 years of chromosome study. This entire collection he very kindly 

 placed at my disposal, and for its use I am greatly indebted to him. 

 I am further indebted to Professor Wilson for many valuable criti- 

 cisms, appreciation for which I wish to express in this place. I am 

 again indebted to Mr. H. G. Barber for the identification of much 

 material of my own collecting which was employed for checking and 

 other purposes. 



Observations. 



I. The Phenomena of Polymegaly. 



Hemiptera of the Family Pcntatomidac possess two compact testes, 

 presenting typically a rather oblong contour when seen in the proper 

 longitudinal section (Text-figs. lA and B). Each testis is enclosed in 

 a connective tissue sheath which is continued into the body of the 

 gland as septa or partitions dividing it into a number of compartments, 

 or, as I shall call them, lobes, arranged parallel to the long axis of 

 the testis. These lobes var^' in number from three to seven (in the 

 forms examined by me, — see Table I), but for any particular species 

 the number is constant. The lobes are typically arranged side by 

 side in a single series (Text-figs. lA and B), each lobe passing from 

 one side of the testis to the other (Text-fig. IC); but in some cases 

 {Perillus (= Mincus) bioculatus for example) a tendency for the lobes 

 of the testis to be bunched together is evident, and in Stiretrus an- 

 chorago, the serial arrangement is completely lost, all the lobes being 

 visible at one time only in cross-sections (Text-fig. ID). In a few 

 cases of this kind, some of the lobes appear to have arisen by the sub- 

 division of an originally single, typical lobe. In any case, the lobes 

 all open at one end into a common collecting chamber, from which 

 the efPerent duct {d in the figures) of the testis is given off. In the 

 typical case, this duct arises laterally (Text-figs. lA and B), but in 

 Apateticus crocahis, for example, it is more nearly median, and in 

 Elasmostethus cruciatus it is quite central, the testis spreading out fan- 

 wise like that of a coreid. Euschistus and Murgantia are good ex- 

 amples of the typical arrangement, and Text-figures lA and B give an 

 idea of the general topography as seen in longitudinal sections of the 

 testis of these bugs. It will be noted that the lobes differ in width, 

 and these differences are in general definite and specific ones, which 



