519 



of two parts, and if we turn the body around 90 " we find its outline 

 like Fig. 19, the fainter constrictions on the top an bottom repre- 

 senting the earlier longitudinal split of the ring, and the one at right 

 angles to this, the transverse division. Thus we have here a tetrad 

 precisely equivalent to the first described, and to that in Anasa. 



These rings, though having the same tetrad value as those de- 

 scribed by RücKEET in the Copepods and by Vom Rath in Gryllotalpa, 

 are of quite a different type: the four constituents being arranged 

 as halves of two small rings, instead of quarters of one large one as 

 in those cases. 



In the equatorial plate of the first spermatocyte division, we find 

 both of the above types of tetrads side by side; the more prominent, 

 that is the transverse constriction, being in the plane of division. 

 Thus the first division is a reducing division. In the Anaphases, the 

 dyad nature of the chromosomes is plainly apparent, but the con- 

 striction, in my preparations, is always transverse to the plane of 

 the first division and never in the same plane as Montgomery de- 

 scribes. This constriction therefore corresponds to the earlier longi- 

 tudinal split, and the second division, which separates these halves, 

 is an equation division. The process in Brochyraena is similar to 

 that in Euchistus. In specimens of four other genera of the family 

 Pentatomid« : Podisus, Mormidea, Coenus, and Nezara, I have found 

 in certain stages a clearly marked longitudinal split, so that this 

 appears to be the rule in that family. I have not yet attempted to 

 trace it through the division stages. 



Somewhat more extended observations upon one species of the 

 family Lygicida), Lyginus turcicus Fab., prove that here the process 

 is similar to that in Euchistus. A longitudinal split is present in 

 the early stages and after the breaking up of the chromatin there is 

 a number of longitudinally split segments, which either contract directly 

 to form a tetrad, or, by fusion of its ends, form a double ring, which 

 then splits in the way described above. 



This species is more closely related to Pyrrochoris, the subject 

 of Henktng's investigations, than any of the forms before described. 

 It is difficult to make an exact comparison between the forms of 

 chromosomes figured by Henking and those described above, but still 

 they seem more like those of Euchistus than like any with which they 

 have hitherto been compared. 



The evidence thus appears to point to the conclusion, that the 

 first division in the Hemiptera-Heteroptera is the reducing division, 



