The Spermatogenesis in Pentatoma up to the Formation of the Spermatid. 3 



already mentioned, haematoxyline (Ehrlich's) and eosin, and the 

 Ehrlich-Biondi-Heidenhain mixture were employed. Hermann's 

 double stain is one of the most excellent for the study of mitosis, 

 differentiating the chromatin of the resting cell from that of the cells 

 in divison; I have used it on many other objects and in all get the 

 same good results (e. g. Plethodon and Harpalus spermatocytes, pole 

 spindles of Cavia, and other objects). In many respects (as in the 

 differentiation of different kinds of nucleoli, and of the spindle fibres) 

 it appears to be absolutely the best method yet employed. 



It is my intention to follow these observations with others on 

 other families of the Heteroptera, with the hope of thereby arriving at 

 some facts of interest from the phjdogenetic standpoint, as well as 

 from the standpoint of the reduction of the chromatin. 



I. 

 Review of Papers on the Spermatogenesis in Insects. 



(With especial reference to the reduction phenomena. 

 Authors' names alphabetically arranged.) 



Carnoy ('85) gives a large number of valuable observations on 

 mitosis in testicle cells of various Arthropods, though without re- 

 ference to chromatin reduction. Stenobothrus and Acridium: the 

 continuous spirem thread segments transversely into 12 to 18 bent 

 chromosomes, which wander in two groups to the opposite poles of 

 the spindle without splitting; in cases where the chromosomes are 

 straight or lobular, they first undergo a longitudinal splitting. Forfi- 

 cula: 10 to 14 chromosomes, the metakinesis as in the preceding. 

 Bacillus: here there are 8 to 10 short chromosomes, which lie parallel 

 to the axis of the spindle. Each splits longitudinally, but the ends 

 remain unsplit, so that an elongated ring form is attained (figs. 292 

 — 296 would show that each ring then divides transversely). Coleo- 

 ptera {Harpalus^ Procrustes, Feronea, Steropus, Cetonia, Hydrophilus) : 

 in most cases the long chromatin loops, formed by segmentation of 

 the spirem, become arranged parallel to the spindle, and there sub- 

 sequently become short and curved ; at the equator they may either 

 divide, or may pass undivided to the poles; the elongate ring-shaped 

 chromosomes are first parallel to the spindle, and then after passing 

 through an angle of 90°, divide into two (his figures do not show 

 this process of torsion). Lepidoptera (larvae of Chelonia and Arctia) : 



1* 



