THE SrERMATOGEXKSLS OF DOIIKVI'IC MAMJIALS. 227 



cloarl}- lie seen -with Auerbach's rccthod, wliilc ^^ ith iron-baematoxylLii tLi3 

 appears a<3 a cliroraatin micleolus. Tlius tlio cliiuorpbisiu us regards to the 

 existence of the chromatin nucleolus, c;in not bo seen. 



Li tlio cytoplasm the chromatoid Ixxi}- or " chromatoicler Nebenköiper" 

 is always conspicuous, stixincd deeply with acid-fuclisiu by Auehbacii's method, 

 while with iron-ha'matoxj'liu it is diflicult to find the distinction between this 

 body and the chromatin mass (Figs. G7, 143). 



The centrosome can clearly be seen in this stige, usually lying near the 

 nuclear wall (Fig. 67). At the beginning of this stage the idiozome distinctly 

 appears in close contact with nuclear A\all, showing a sharp contour (Figs 



68, 143). As has ah-eady been described by Bexda ('91) and by Metos ('99) 

 small granules can bo seen within the idiozome (Fig. 67, 68). A small 

 quantity of the mitochondrial gi'anules ap^xjars in this stage, being distiibuted 

 all over the cell body (Fig. 69). 



13. The formation of the spermatozoa in the mouse. 

 Period' I. 



The nucleus : — The small spheric nucleus gradually enlarges and the 

 chromatin granules begin to disappear, leaving a centi'al mass of chromatin 

 behind (Fig. 67-70). This enlargement of the nucleus is probably due to 

 the fact that at this sbxge the spherical nucleus begins to flatten. Simultaneous- 

 ly with this the nucleus gi'sidually moves towards one side of the cell which 

 is destined to become the anterior end of the spermatozoon, while a large 

 amount of the cv-toplasm accmnulates at the posterior part of the cell body 

 (Figs. 70, 71). 



The cenfa."osome i — In this period two centrosomes become clearly visible ; 

 these lie at first side by side, but soon move towards the nuclear wall (Figs. 



69, 70). One of these centrosomes which is destined to become the anterior 

 one, comes to Ix; placed in contact with the nuclear wall, while the fine 

 filament which later becorce? the axial filament of the spermixtozoon, begins 



1. The four periods into wbii^h the formation of the spermatozoa is divided, is in occordnnoe 

 with the work of Me^es ('99), although it ia not x^ossible to draw any definite line of demar- 

 <ntion between the two oucccssivc periods. 



