310 THOMAS M. MONTdoMKRV, JR. 



two classes of cells, just as they do in the Sertoli cells and sperma- 

 tocytes of mammals. 



The material consisted of testes, seminal vesicles, vasa defer- 

 entia, and oviducts, some fixed in strong Flemming's fluid diluted 

 with an equal part of distilled water, and others preserved in 

 corrosive sublimate-acetic; all were originally received from my 

 friend, Dr. Purcell, of Cape Town. The mitochondria appear 

 pale red after Hermann's safranine-gentian violet, and after 

 the Ehrlich-Biondi-Heidenhain method; shades of gray or black 

 after iron hrematoxylin, according to degree of destaining; and 

 after Benda's stain they are deep violet, while the chromatin is 

 brownish and the centriole red the typical reaction to this stain. 

 There is only one other object known to me on which they are 

 equally readily demonstrated, namely, the spermatocytes of 

 A scar is. 



Fig. I, PI. I., exhibits the position of the mitochrondia at the 

 end of the second maturation mitosis where, as after the first also, 

 they lie at the distal poles (equatorial ends) of the daughter cells. 

 Until their later fusion takes place they are chiefly peripheral, 

 next the cell wall, spherical or slightly elongate, and in the form 

 of hollow vesicles. In the earliest spermatids they always form 

 a layer at the distal pole, but sooner or later move forward, along 

 the cell membrane, so as to take a position on the side of the 

 nucleus (Figs. 2-16); at the same time the sphere (s) always ad- 

 vances from its original position and the cytoplasm comes to 

 make a lobe around the nucleus and entirely in front of the cen- 

 triole (c). These movements do not occur synchronously on 

 cells of the same stage, there is much variation in the process, yet 

 the end result is the same in all. By reason of tin- mitochondria 

 remaining generally in a single layer they may be readily counted, 

 and their number is found to differ in different spermatids, which 

 shows their mass cannot be accurately quartered by the matura- 

 tion divisions. In Figs. 4, 6, 11-13, a " () ' each cell are drawn 

 with care, and their numbers here are respectively: 33. 45. ()S - 

 64, 49. Their volumes also differ considerably, as the figures 

 show. They do not blacken with osmic acid. 



In the nucleus the chromosomes are at first peripheral and 

 quite distinguishable (Fig. 2), then coalesce to produce a hollow 



