696 THE GENERATIVE ORGANS. 
and some are seen to be surrounded by a distinct sheath, which 
occasionally exhibits a large nucleus. These sheaths are pro- 
bably the remains of the daughter-cells, which have been only 
partially converted into spermatozoa by endogenous cell for- 
mation. 
These observations agree with those of Gilson [857], who 
holds that when the spermatozoa are seen enclosed in a cap- 
sule, this consists of the remains of a cell of a previous gene- 
ration; he further regards each spermatozoon as a distinct 
differentiated cell, and says that in all other points variations 
exist in the manner in which spermatogenesis occurs in 
different animals. I would go even a step further, and think 
that not only do such variations occur in different animals and 
groups of animals, but that the process of cell multiplication 
which precedes spermatogenesis varies under different condi- 
tions in the same animal. Thus, the size of the polyblasts in 
the Blow-fly varies within wide limits. In some cases they 
are comparatively small and contain few nuclei when the for- 
mation of spermatozoa commences; whilst others are very 
large, and contain numerous nuclei. Again, some of the sper- 
matospores remain enclosed in a distinct nucleated envelope, 
and others are separated by the exogenous division of the 
mother-cells. It appears to me that exogenous and endo- 
genous cell multiplication occur in all stages of the process. 
Sometimes a spermatospore continues to multiply by exogenous 
division for a long time; whilst in a sister-cell endogenous 
multiplication takes place without previous exogenous division. 
In the former case small, and in the latter large polyblasts 
arise; small bundles of spermatozoa originate in the small 
polyblasts, whilst large ones are formed in the large poly- 
blasts. Some of these remain enclosed within the mother- 
cells, whilst others are apparently free and unenclosed. 
There is no doubt active karyokinetic changes occur 
throughout the whole process, but I have not attempted the 
study of the nuclear figures, as I feel that the complexity of 
the phenomena hitherto described, and the great discrepancies 
which exist in the statements of various observers, are such 
