54 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
the peripheral or first multiplication zone, the second growth zone, and 
even the maturation zone have disappeared, the gonocysts and cyto- 
cysts having all been converted into spermatocysts variously devel¬ 
oped (pi. 12, fig. 5, /.). 
Each follicle of the testis seems to mature independently of the rest. 
Thus I often find some follicles in which nothing but mature sperma¬ 
tocysts remain as in plate 12, figure 5, /., while neighboring follicles 
may still have the original multiplication and growth zones of gono¬ 
cysts and cytocysts. I shall show later that this difference is due to 
the presence in the latter follicles, of grandmother stem cells from 
which new gonocysts arise. It is quite probable, therefore, that one 
part of the testis may become quite functionless, while other parts 
may continue to produce spermatozoa, the spermatogones in the 
former being exhausted. 
Gonocysts. 
In those follicles where no grandmother stem cell exists, I have been 
unable to find the first stages or two-cell stage of the gonocyst. Even 
in the youngest larva examined by me where the right and left testes 
are still separated by a considerable space, and still comparatively small, 
all the gonocysts are in an advanced stage, judging by the number of 
spermatogones they contain, i. e., not considering those follicles con¬ 
taining a grandmother stem cell. 
I feel justified in inferring from this that all the primary sperma¬ 
togones begin their multiplication at about the same time, in a given 
zone of the follicle and that this multiplication begins near the vas 
deferens, the proximal area of the follicle, and extends gradually 
towards the periphery, the distal termination of the follicle. The 
question then naturally arises how are new spermatozoa produced 
when all the primary spermatogones are thus developed into advanced 
gonocysts at this early period of larval life? 
For a long time this question was a mystery, for there was nothing 
suggesting a solution in published accounts of other insects, and I did 
not at first understand the real significance of the grandmother stem 
cell. There is nothing in the plates of Henking (’ 92 ), Wilcox (’ 95 ), or 
Paulmier (’ 99 ), showing that these writers have solved the problem. 
It seems clear that the same problem exists in those insects, but the 
