444 FRANK R. LILLIE 
sperm aster arises around an extranuclear centrosome of spermatic 
origin. There is, therefore, no reason for assuming that the 
experiments on Nereis have revealed an exceptional condition. 
Kither, then, the centrosome theory of fertilization must be 
rejected in toto, or the sperm nucleus must be regarded as a cen- 
tronucleus in Boveri’s sense (Boveri ’00), i. e., as including the 
centrosome. If we examine the latter point of view we see that 
it would be necessary to assume that centrosomes exist in the 
sperm nucleus at every level, for the sperm aster can form at any 
level as demonstrated by the experiments. The diagram (text fig. 
Text fig. 4 
4) will make this clear. It represents the sperm nucleus in the 
form that it possesses as penetration is nearly completed 
(cf. fig. 8); normally the sperm aster arises at point a, but if ab be 
removed then at point b, if ac be removed at c, if ad be removed 
at d,etc. But the assumption that centrosomes exist through- 
out the nucleus and condition this phenomena would seem to be 
merely an exaggerated morphological point of view with refer- 
ence to a problem that is, after all, fundamentally physiological, 
viz.: by virtue of what properties does the sperm nucleus exercise 
this effect? If the great mass of experimental cytological data 
favored the view of the permanence of the centrosome, the con- 
