324. M. D. HILL. 
fulfilled. Where careful observations have been made this 
supposition has almost always been found to hold good, e.g. in 
the case of Rhynchelmis (Vejdoffsky, 8), Ascaris (Boveri, 2), 
Axolotl (Fick, 8), and Styelopsis (Julin, 5); while only in 
one instance has it been incontestably denied, viz. by Wheeler 
(9) for Myzostomum. When, however, it is borne in mind 
that nuclear and cell-division can take place without the 
presence of an astrosphere or centrosome, the supposed im- 
portance of the latter as an organ of division is greatly lessened. 
It is, moreover, extremely difficult to offer any explanation as 
to why the first cleavage spindle should have two astrospheres 
and two or four centrosomes, while the polar spindles 
(Phallusia, Ascaris, Sagitta, Ciona, &c.) may have none. 
Again, the relation between centrosome and astrosphere is 
very obscure; are the rays produced by the action of the 
centrosome, or vice versd? The former alternative seems to 
the most probable in the case of Phallusia, where, as is shown 
in fig. 18, while the spermatozoon is quite at the surface of the 
ovum the posterior half of the middle piece has become a 
centrosome, and already acted on the cytoplasm of the egg to 
produce a radial appearance. In spermatozoa simply killed 
and stained there is no sign of a centrosome in the middle 
piece, which points to a direct metamorphosis of part of the 
middle piece into a centrosome as soon as the spermatozoon 
penetrates into the ovum. 
Neither in Spherechinus nor Phallusia is there any evidence 
to point to the centrosome as being artefact. In the former it 
appears to be all that remains of the middle-piece, and is 
brought to light by the latter’s disintegration into granules 
and final disappearance. 
Finally, with regard to a more important matter, viz. the 
relations of the chromatin substance during the maturation 
of the ovum, Phallusia agrees more with the vertebrates 
than the invertebrates. I cannot go into the subject, however, 
in any great detail, because I have been unable to trace the 
history of the eight chromosomes in the nucleus of ovocyte I. 
Still, there are one or two points worth noting. In the first 
