CHROMOSOMES IN THE SPERMATOGENESIS OF THE HEMIPTERA HETEROPTERA. 101 



sion, their equational division there, and their conjugation and separation in the sec- 

 ond mitosis. In my first paper on this species the spermatogonia! number of cln-o- 

 mosomes was correctly given, wliile in the later paper I was misled by one of the 

 unusual cases, here described, of IG chromosomes in the equator of the spindle. 



2. P]uscHisTUs TRisTiGMUs Say. 



Spermaf agonic Divisions. — Always 14 chromosomes (Plate IX, Fig. 15), 3 (/>/, F, f) 

 being noticeably smaller than the others. When these elements lie suitably 1 2 of them 

 are seen to compose G pairs (.1, u-I<\ f) each pair with components of approximately 

 equal volume and form ; these are the maternal and paternal autosomes. There re- 

 main two elements, Di and di, one of which is the smallest of all, the other larger 

 than this and also larger than either component of the smallest autosome pair; these 

 two elements of such different volumes are the diplosomes. 



Growth Period. — The autosomes unite to form 6 bivalent ones as previously de- 

 scribed by me. The diplosomes also unite regularly and remain so during the earlier 

 part of the growth period {DI, Di, Fig. 16), but they later separate. 



First Maturation Division. — There are always 8 elements (Figs. 17, 18), 6 of these 

 are bivalent autosomes {A, a-F,f), and these divide reductionally. And 2 are the 

 separated and univalent diplosomes {Di, di) which also divide and hence equationally. 

 A pole view of a daughter chromosomal plate of the ensuing anaphase (Fig. 19) be- 

 fore the chromosomes have taken their place in the equator of the second spindle 

 shows the two diplosomes unconstricted, and each of the six autosomes with a con- 

 striction that is the longitudinal split. 



Second Maturation Division. — In the ecjuator of the spindle (Fig. 20) are seen the 

 6 autosomes dividing along the line of the longitudinal split ; but the two diplosomes 

 have conjugated end to end and form a bivalent element with unequal components 

 that divides reductionally. Each spermatid receives 7 chromosomes, half of them 

 receiving the larger (Fig. 22) and half the smaller diplosome (Fig. 21). 



In this species each chromosome pair can Ije followed with great certainty during 

 all its changes, thanks to the marked differences in V(dume of the different pairs ; and 

 this I have illustrated upon the figures by correspondence in the lettering. 



Literature. — My first account was entirely correct (19016), and I described how 

 the diplosomes divide separately inthe first maturation mitosis. But I failed to notice 

 their conjugation in the second spermatocytes. Wilson's account of this and the pre- 

 ceding species is correct. 



