50 Papers from the Marine Biological Laboratory at Tortugas. 
into two. Even shortly before this division astral rays are seen to take form 
round the single centrosphere. These are undoubtedly formed from the 
cytoplasmic reticulum, as Wilson has shown in the sea-urchin’s egg. The 
rays abut the nuclear wall and indent it slightly. The archoplasmic mass 
now begins to elongate. Presently it assumes a dumb-bell shape, a bridge 
of delicate fibers showing between the two globes. Each of the globes now 
contains a centrosome (centriole of Boveri) surrounded by a centrosphere 
and astral rays. The two centrosomes separate more and more, moving 
over the surface of the nucleus, their astral radiations elongate and press 
upon the nuclear wall, indenting it deeper and deeper until it finally ruptures, 
allowing thus an intermingling of cytoplasm and nuclear content. Rupture 
seems to be due to pressure plus some solvent influence of the rays. It is 
apparent that no part of the achromatic structure of the first polar spindle 
came from within the nucleus, for the wall in some cases is still intact when 
the two centrosomes and asters (from which the spindle is spun) are 
already formed (fig. 50). While the centrospheres enlarge and the astral 
rays lengthen, the latter also increase perceptibly in thickness. This 1s par- 
ticularly true of the traction fibers (Zugfasern), to which the chromosomes 
become attached and by which they are drawn into the central spindle 
(figs. 51, 52, 61, 62). The spindle is spun between the two centrosomes 
by the union of the astral rays very much in the manner described by 
Child (1896) in Arenicola marina. The outermost rays and those that do 
not unite or blend to form the fibers of the central spindle interdigitate with 
each other in the median plane (figs. 52, 64). The astral rays merge at 
their distal ends into the general cytoreticulum. When the first polar 
spindle is fully formed the centrospheres have a reticular or alveolar struc- 
ture, and occasionally two centrosomes are seen (fig. 66). 
NUCLEOLAR CHANGES. 
When maturation is imminent the nucleolus is usually in a position on that 
side of the nucleus nearest the cell wall (fig. 37, 50). In this position many 
of the astral rays of the first polar spindle are extended directly upon it 
(figs. 53, 58, 59). The fact that the nucleolus as well as the mass of chro- 
mosomes and the occasional problematical body are all concentrated into this 
narrow space, combined with the further fact that the nucleolus begins to 
fragment and the chromosomes to scatter among these fragments, while the 
astral rays are extended among them indiscriminately, makes it difficult 
ordinarily to trace the fate of these several structures. However, excep- 
tionally favorable conditions permit of observations which leave no doubt 
as to the correct interpretation of the more complicated processes. Par- 
ticularly favorable for such study is the condition where the chromosome 
mass and nucleolus at the beginning of maturation are widely separated 
(figs. 32, 38, 43; 46, 47; 48, 49). 
