SHAPE OF THE NUCLEUS 595 
ended nuclei—so characteristic of human heart muscle—can 
only be explained by assuming the presence of these membranes 
lying invisible in the sarcoplasm at each end of the nucleus. 
TexT-rie. 4. 
Oocyte of Esox lucius—a pike. Note the pouches in nuclear mem- 
brane usually in relation with the nucleoli. Nv, nucleolus; yp, yolk- 
discs. Technique: Bouin and iron haematoxylin. 
4, NucLEAR SHAPE AND THE CENTROSOME. 
We deliberately confine ourselves to the consideration of 
the centrosome and nucleus in the resting cell, as the question 
of the spindle fibres, amphiaster and chromosome formation 
is beyond the scope of these observations. In the resting 
cell the centrosome often lies very close to the nuclear mem- 
brane and opposite an indentation in it. And since this body 
often does not touch the nucleus, one must surmise that the 
depression is due not to mechanical causes but to repulsion 
between nuclear membrane and centrosome. When an 
amphiaster is present, its influence upon the nucleus is still 
more marked, as is shown in Text-fig. 6, which depicts a blasto- 
mere from an egg of Triton. It will be seen that here 
nuclear shape is due partly to pressure from the yolk-dises 
(as already pointed out), partly to invaginations in the nuclear 
membrane in the vicinity of the centrosome. The astral rays 
in fact deeply indent the nucleus wherever they come into 
contact with it—a point possibly in favour of the view that the 
cytoplasmic radiations around the centrosome are of a relatively 
solid nature, 
