144 KANSAS UNIVERSITY QUARTERLY. 



within the cyst increases by division, the cyst membrane, obviously, 

 becomes more and more extended, the material for this increase being- 

 furnished by those primary spermatogonia which still remain in the 

 resting condition, one or more being always found in relation with 

 each cyst, either in the axis of the follicle or between the cyst and 

 the follicle wall. 



To return now to the developing cell ; the spiremes of the prophase 

 attain, by a process of shortening and thickening, the form of short, 

 heavy, rough-contoured spiral threads (fig. 19), which split longitudi- 

 nally. These assume a form similar to that of the chromosomes in 

 the metaphase already described and divide along the lines marked 

 out in the prophase, giving rise to two cells each, and consequently to 

 a cyst of four cells. 



It is not until the eight- or sixteen-cell stage that I am able, in my 

 preparations, to find all the stages, but such as I do find in the few 

 smaller cysts which I have seen would indicate no difference between 

 the early and late spermatogonial divisions. 



These I will describe in detail, beginning with the anaphases, as I 

 have seen them in the later generations, and as I believe them to 

 commence after the metaphase of the primary spermatogonia. 



At the beginning of the anaphase, the chromosomes are in the form 

 of double rods lying in the equatorial plate, perpendicular to the axis 

 of the cell and attached by their central ends to the spindle fibers. 

 (Figs. 22, 23.) 



The separation marking the commencement of these phases begins 

 at the central ends of the chromosomes, producing a circle of V-shaped 

 figures with their ai^ices outward. The two halves remain connected 

 at their outer ends until the apex of the "V" has been drawn down 

 into line with the limbs, when they separate and the two halves travel 

 through an elliptical course to their respective centrosomes (fig. 25), 

 stretching out between them the connecting fibers, which at first pre- 

 serve the ellipsoidal shape of the spindle figure, but soon begin to draw 

 together in the middle so as to form first a cylinder and then a hyper- 

 boloid. (Figs. 26 and 27.) During this time the cell- wall has con- 

 stricted across and we now have two daughter-cells connected by the 

 persisting spindle. Where this spindle penetrates the cell- wall there 

 is developed on each of its fibers a slight enlargement which stains 

 rather deeply. (Fig. 27.) 



Many authors describe these connecting fibers as being arranged 

 in two cylinders, one within the other, and distinguish them by the 

 names of spindle fibers and mantle fibers, respectively. In this object, 

 however, as shown by fig. 28, almost all the fibers are confined to the 

 mantle, a very few being irregularly disposed within. 



