26 THOS. H. MONTGOMERY jr., 



Just after the preceding metaphase (Fig. 50) evidences of astral 

 radiation persist in the cytoplasm , and the latter has a more 

 or less loose texture, except at the periphery of the cell, 

 where it is denser and more granular. At this stage I have been 

 unable to detect any centrosomes, or more than slight traces of the 

 idiozome substance. A little later, when the chromosomes lengthen 

 out, an idiozome appears at one point on the surface of the nucleus, 

 apparently usually at that pole where the greatest mass of cytoplasm 

 is situated {Id. s Figs. 53, 55, 58—60). Its substance is nearly homo- 

 geneous and shows no demarcation into different zones, and though 

 sharply marked off from the cytoplasm, it does not appear to possess 

 a special limiting membrane. It does not color with the use of Her- 

 mann's stain, but becomes browned more deeply than the cytoplasm 

 by Hermann's fixative ; with the Ehrlich-Biondi stain, or with the 

 use of haematoxyline-eosin, it stains more deeply than the cytoplasm. 

 Thus in the spermatocytes it would seem to have the same chemical 

 consistency as in the spermatogonia. The idiozome in the synapsis 

 (Figs. 63 — QQ, 68—70) is a body of more or less concavo-convex 

 shape placed close to the outer surface of the nucleus, and produces 

 either a flattening or a depression of the nuclear membrane. Within 

 the idiozome, though not at any particular point in it, may be detected 

 in the majority of cases one or two minute granules, which are prob- 

 ably centrosomes. Since two such granules are most usually found, 

 it may be concluded that division of the centrosome occurred in the 

 preceding metaphase, or in the early anaphase. My observations do 

 not positively prove these granules to be centrosomes; the proof of 

 such identity can only be brought by one working with higher powers 

 of magnification, and with a greater variety of staining methods, — 

 lack of material prevented me from employing as many of the latter 

 as I should have desired. The great regularity, however, with which 

 one or two such granules are found in the idiozome, and the analogy 

 with other cases in which they are proved to occur (e. g. the sala- 

 mander spermatocytes), renders it at least possible that they represent 

 centrosomes ; and that the idiozome must certainly include the centro- 

 somes, would seem to be proved by the following observations. 



The cytoplasm immediately surrounding the idiozome appears in 

 the early anaphase, as a rule, to be less dense than at other points 

 in the cell, and very frequently this more or less clear portion is 

 seen to be transversed by delicate fibres which radiate from the surface 

 of the idiozome (Fig. 55). These are, in all probability, persisting 

 pole fibres; they disappear before the synapsis. More pronounced 



