ON THE SPERMATOGENESIS OF THE SILK-WORM. I35 



of three parts, the two polar parts which stain faintly with 

 hematoxylin and the median part which stains deeply with the 

 same dye. 



Beside this change there is to be seen a deep colouring spot 

 surrounded by a free area either within the cytoplasm of sperm- 

 mother-cells or in the boundary between two such cells. It is 

 either round or ellipsoidal in shape, and sometimes it has a con- 

 striction in its centre (fig. 33, v) giving it the appearance of a 

 dumb-bell (figs. 32-37 v, 39-43 v, 47 v). This is surely the "Ver- 

 bindungsbriicken " of Plainer (27). Nothing can be said of its 

 first appearance or its disappearance, but as it is found in cells 

 in which the centrosomes are already present, it can not be the 

 centrosome, although it appears very much like it, especially 

 when it is found in the cytoplasm. 



A little before the appearance of the centrosomes in sperm- 

 mother-cells the chromatin granules as shown in fig. 44 gra- 

 dually collect here and there and assume ring-shaped struc- 

 tures. These rings have already been described by Hcnking (15) 

 and vom Rath. (28). Each of these rings again dissolves into 

 four small chromatin granules (fig. 45) as vom Rath observed in 

 a sperm-mother-cell of Gryllotalpa and Salamandra (28, 29). 

 After the breaking up of the chromatin ring, the separated 

 chromosomes are now ready to divide. 



In the preparation fixed by Flemming's strong chrom- 

 osmium acetic acid, the " Verbindungsbriicken " is clearly to 

 be seen without any staining, while the spindle fibres can not 

 be seen unless they are stained by hsematoxylin. I am not able 

 to observe these structures in the preparation fixed by picro- 

 acetic acid and stained by picro-carmine. 



In the sperm-mother-cells of the silk-worm, no accumula- 

 tion of yolk granules is to be observed in the cell-body, although 

 in other animals, such as in Ascaris and in Pyrrhocoris, this is 

 the case. 



III. The Ripening Stage. In this stage, we find fully grown 

 sperm-mother-cells each of which divides two times succes- 

 sively and produces four cells which I will call the sperm- 

 daughter-cells. These change themselves, without any further 

 division, into spermatozoa. 



