CHROMOSOMES IN- THE SPERMATOGENESIS OF THE HEMIPTERA HETEROPTERA. 119 



gate during tlie synapsis (Fig. 152) and in the later postsynapsis (Fig. 155) shows a 

 split along its length which, as is the case also with the autosomes, is widest at its 

 middle ; this split becomes temporarily closed a little later. The diplosomes {Dl, di, 

 Figs. 153, 154) remain very small during the growth period but retain their red stain 

 and dense structure ; usually but not always they are close together, and like the 

 monosome lie against the nuclear memljrane. There is always one large plasmosome 

 (Figs. 154, 155, PI) and frequently one or two smaller ones. 



First Maturation Mitosis. — In the spindle there are 1 1 elements so placed that 

 within a circle of 9 autosomes is the bivalent diplosome (Di, <U, Fig. 15G), and outside 

 of this circle the univalent monosome (Mo) which lies with its long a.xis in the equa- 

 torial plane ; the annular constrictions of the autosomes found upon pole views mark 

 their longitudinal splits. All of these are shown on lateral view in Fig. 157, and 6 of 

 them in Fig. 158. The 9 autosomes divide reductionally, and so does the bivalent 

 diplo.some because its parts that separate from each other are unequal in volume and 

 in the preceding stages we found this dissimilai'ity characteristic of the two. The 

 monosome, however, lies with its long axis in the plane of the equator (Figs. 157, 158, 

 Mu), and divides through its length. 



Secovd Maturation Division. — Here again there are 11 elements (Fig. 159), but 

 grouped differently from those of the preceding division in that there are usually 2 

 within a circle of 9. They are 1 univalent diplosome {Di), 9 univalent autosomes, 

 and the half of the monosome. The autosomes and the diplosome divide again and 

 equationally (Fig. 160), but the monosome {Mo, Figs. 160, 161) passes undivided into 

 one of the spermatids and usually lags behind the others in reaching the spindle pole. 



Literature. — Paulmier's monographic account of the spermatogenesis of this 

 species (1899) was in the main a very correct one, save that he stated the normal 

 number of chromosomes to be 22, and consequently identified the allosome of the 

 growth period and the chromosome that does not divide in the second maturation mito- 

 sis with the minute diplosomes. I (1901/>) followed Paulmier in these mistakes, and 

 because the monosome of the spermatogonium is constricted counted it as two. Wil- 

 son (1905c, 1906), in whose laboratory Paulmier's work was done, was the first to cor- 

 rect these errors, and to trace the history of the monosome distinct from that of the 

 diplosomes. But Wilson failed to note that the diplosomes are not quite of the same 

 size, and that they may be distinctly recognized during the greater part of the growth 

 period. 



22. Anasa sp. (from C'alifornia). 



Spermatogonic Divisions. — In every case there are 21 elements in the spindle 

 (Plate XI, Fig. 164). These are : 2 diplosomes of unequal volume {Di, di) ; 1 mono- 



