14(5 KANSAS UNIVERSITY QUARTERLY. 



more regular in outline, soon becoming of uniform diameter and often 

 twisted like a rope. ( Fig. 16; fig. A, ./•.) About this time the ordinary 

 chromosomes show a tendency to lose their spiral form ( figs. 17, 18, 

 11), and 39), and in some of them may be seen a distinct longitudinal 



split. 



Near the stage shown in fig. 16, there may be seen, close to that 

 portion of the nuclear wall which faces the greatest mass of cytoplasm, 

 two extremely minute but deeply stained centrosomes, each with its 

 radiate arrangement of the cytoplasm and joined together by a very 

 small centrodesmus. These move wider apart (fig. 39) as the chro- 

 matic threads assume more and more the form of chromosomes, and 

 soon come to lie diametrically opposite one another. This moment is 

 marked by the disappearance of the nuclear membrane but not always 

 of that of the accessory chromosome (fig. 21), which now has the 

 form of a rough double rod (fig. 15), or, as often, of a two-strand rope 

 (fig. 21^0- I ^^^^'^ noticed a splitting of the accessory chromosome at 

 a stage as early as that shown in fig. 16, in which case, however, it is 

 impossible to determine its condition. 



Soon the chromosomes, including the accessory, assuming the form 

 of short, thick, straight or slightly curved double rods, take their places 

 in a radial position in the equatorial plate and the central end of each 

 becomes attached by several fibers to each of the centrosomes. ( Figs. 



22 and 23. i 



We have now completed the cycle of the ordinary spermatogonial 

 division. The telophases of the last division, however, differ from 

 those of the others, as we would naturally expect, in preparation for 

 the growth stage of the spermatocyte. This dilt'erence consists in the 

 reduction, in the late telophases, of the sacculated nucleus to a mere 

 irregular vesicle, which finally becomes smooth in contour ; and in 

 the regular diffusion of the chromatin throughout the nuclear space. 

 The accessory chromosome becomes more and more condensed, and 

 its membrane, closely applied to that of the nucleus, still encloses it 

 at the time of the change to the spermatocyte. 



Fig. 38 represents an early spermatocyte prophase in which the 

 hollow center of the vesicular stage of the assessory chromosome 

 seems to be not entirely obliterated. 



DISCUSSION AND COMPARISON OF LITERATURE. 



I. The Origin of tin- Spermafocyi^ts.— 'TliQ occurrence of the 

 sperm-cells of animals in groups or spermatocysts has been recog- 

 nized and mentioned by investigators upon the most various forms 

 since it was first noted by St. George in the Amphih!a, in 1876. Few 

 authors, however, have attempted any explanation of their origin, or 

 even a description of their structure, many seeming to take their 



