10(_) CHROMOSOMES IN THE SPERMATOGENESIS OF THE IIEMIPTERA HETEROPTERA. 



appearance of a totality of 10. It was on the basis of cases of this kind that I had 

 previously decided that the normal number is 1 6, whereas 1 now find that the usual 

 number is 14. Whenever all the chromosomes lie with their long axes in the plane 

 of the equator their arrangement in pairs of like components may be readily made out. 



(Jrowth Period. — In the synapsis the 12 autosomes conjugate to form 6 bivalent 

 ones as I previously described in some detail (1898, 19016). The diplosomes also 

 always unite then end to end. At first each diplosome may become more or less 

 irregularly bent (Fig. 5), later becoming more spherical. After the synapsis period 

 they are at first in intimate contact, each is a little longer than wide with a slight 

 constriction around the middle (Fig. 6, Di, di) ; this probably represents a longitudinal 

 split of each. The two may lie parallel or slightly divergent, or frequently with their 

 long axes making a right angle. When they are so placed a small space is seen 

 between them, and this I erroneously described in 1898 as a vacuole within a single 

 element ; now I can decide that no such vacuole is formed, and that the diplosomes 

 swell but little in size during the growth period. Though the two may often be so 

 near together as to appear to form an apparent single sphere, they never seem to 

 actually fuse, for a line of separation can always be found. 



First Maturation Division. — The behavior of the autosomes was described in full 

 in the papers already referred to. In the late prophase, just before the dissolution of 

 the nuclear membrane, or at that time, the diplosomes separate. After they separate 

 each may continue to show the longitudinal split (Fig. 8) or may not (Fig- 9) ; in the 

 latter case there is, that is to say, a temporary closure of the split, just as happens reg- 

 ularly with the autosomes. In the monaster stage are found 8 elements, and all of 

 these are shown on latei'al view in Fig. 10. Six of them are bivalent autosomes and 

 these divide reductionally. But each of the two smallest chromosomes is a univalent 

 diplosome, and their division is probably through the plane of their earlier longitu- 

 dinal split. Each second spermatocyte receives 6 univalent autosomes, and half of 

 each of the diplosomes. 



Second Malwration Mitosis. — In the equator of the spindle (Figs. 11, 12) all the 

 6 autosomes become placed with their constrictions (longitudinal splits) in the plane 

 of the equator, and they all divide equationally. But the two diplosomes conjugate 

 in the middle of the chromosomal plate where they compose a bivalent element with 

 components of unequal volume {Di, di), and this double element divides reductionally. 

 Consequently each spermatid receives 7 chromosomes, whereby half the spermatids 

 get the larger diplosome (Fig. 13) and half the smaller (Fig. 14). 



Literature. — In my previous papers, 1898, 190U>, I made the serious mistake of 

 failing to note the separation of the diplosomes just before the first maturation divi- 



