10 DUBLIN 



fact is most strikingly brought out in connection with the older 

 testes. Here mature spermatozoa are often found surrounding 

 spermatogonia of this early generation. The latter evidently 

 remain latent and develop only when room is made for them by 

 the discharge of ripe spermatozoa. 



Leaving out the slight differences in size, these different gen- 

 erations are all of the same type in both the resting and division 

 stages with the important exception of the one, which from its 

 smallest size, I take to be the last. In the prophases and meta- 

 phases of all but the last of these, the chromosomes appear 

 everywhere as V's, with their apices directed toward the spindle, 

 (Figs. 2—6) and cannot be distinguished by any structural dif- 

 ferences from the figures in the somatic mitoses. The number 

 is clearly unreduced, and by the study of many division figures 

 is determined as twenty-two. They split longitudinally, and in 

 the anaphase (Fig. 7), are seen with their angles turned toward 

 the poles. In the telophase (Fig. 8), the twenty-two small 

 bodies are concentrated very closely, and with the chance 

 exception of some belated ones, cannot be made out individually. 



In the last generation, the character of the mitosis shows a 

 marked contrast to the preceding ones. The chromosomes 

 are, in the metaphase, in nearly every case, no longer V-shaped 

 but thickened dumb-bell shaped bars (Figs. 9—12). It is im- 

 portant to observe that these often show a constriction in the 

 middle which, may correspond to the apex of the earlier V's. 

 These split longitudinally, and in the early anaphase (Fig. 13) 

 the unreduced number of rods appears, presenting a striking 

 difference from the other previous spermatogonial figures. It 

 must be stated at this point, however, that in one or two 

 instances I have found one or two chromosomes in the meta- 

 phase of the last division not as yet converted into the bars. 

 This, together with the fact that some straight rods are occasion- 

 ally found in spindles, which, from consideration of size, I 

 think are of the penultimate generation, would tend to show 

 that this conversion of the type of the chromosome is not a 

 sudden process, but is perhaps carried over a considerable 

 period, being most pronounced and finding its completion in 



