114 CHROMOSOMES IN THE SPERMATOGENESIS OF THE HEMII'TERA HETEROPTERA. 



tively enormous elements, one of which is approximately straight and apparently a 

 little moi-e voluminous, while the other is horseshoe-shaped. Finally there is a single 

 chromosome without a corresponding mate, therefore a monosome (Mo). 



Growth Period. — In the synapsis stage the 10 autosomes become longitudinally 

 split and conjugate to form 5 bivalent ones. But 3 of the chromosomes differ in pre- 

 serving their safraninophilous stain and dense structure; from the later history of 

 these there can be no question that the largest (il/o. Figs. 108-111) is the monosome, 

 the 2 smaller the diplosomes (Di, di). The monosome increases somewhat in volume 

 and in the postsynapsis (Figs. 109, 110) is rod-shaped, sometimes bent, and undergoes 

 a longitudinal splitting ; in the rest stage, that is complete (Fig. 11 1), it becomes more 

 I'ounded and then shows either no trace of this split, or el.se only a mere sign of it in 

 the form of an indentation at either end ; it may or may not lie against the nuclear 

 membrane. The diplosomes are unequal in volume as in the spermatogonium, and 

 undergo but slight increase in mass during the growth period. In the postsynapsis 

 each {Di, di, Fig. 109) becomes bipartite, which is evidently a longitudinal splitting, 

 and they remain so during the remainder of the growth period. The spermatocytes 

 contain each several large plasmosomes {PI, Figs. 110, HI), and the diplosomes, and 

 less frequently the monosome, ma}' be in contact with these. 



First Maturation Divisio)i. — In the prophases there are 5 bivalent autosomes 

 {A, a-E, e, Figs. 114-116), each longitudinally split. One of them, by far the largest 

 {A, a), is in the earlier stages the single one that is regularly ring-shaped (Fig. 112), 

 with a distinct longitudinal split in each arm of the ring; this ring gradually ojiens 

 until it first becomes an angle (Fig. 113), then straight (Figs. 1 14-1 1 6), the longitudinal 

 split still continuing in the axis of each arm (univalent constituent). By the gradual 

 condensation of the auto.somes (Fig. 116) their longitudinal splits become more or less 

 closed, but even in the metaphase it is sometimes clearly indicated (Plate XI, Fig. 

 118), and is then always parallel to the long axis of the chromosome. No animal 

 shows more decisivel}'' than this one that the first maturation mitosis separates whole 

 univalent chromosomes. The monosome can be recognized as a large dyad (7l/o, Figs. 

 114-116). The diplosomes (/);', di, Figs. 114-116) do not conjugate until the later 

 propha.ses, appareiitl_y usually not until the nuclear membrane has disappeared ; in 

 them the longitudinal split becomes temporaril}' closed as in the case of the autosomes, 

 but the monosome continues to show it distinctly. 



There are in the spindle almost invariably 7 elements (Plate XI, Figs. 117, 1 18) ; 

 in a few cases 8 are to be seen on pole aspect, which is then due, as in Harmostes, to a 

 precocious division of two of the bivalent elements, Imt here usually of the bivalent 

 diplosome. There is a central bivalent diplosome {Di, di) and around it a cii'cle com- 



