342 BOTANICAL GAZETTE [NOVEMBER 
unfortunately for this view a very large proportion of the amitotic 
spirems are intermediate, or so closely resemble the mitotic spirems 
that they can be distinguished from them only by the character of the 
cyst in which they occur (fig. 41, a, b,c). It seems impossible therefore 
to determine certainly whether these two classes are distinct or 
whether they are of the same nature. 
The origin of the spindle 
In those cells where the spindle is intranuclear, there seem to be 
two types of spindle formation. In the first type, which occurs in 
the ascomycetes (e. g., Phyllactinia, HARPER 8) and brown algae 
(e.g., Fucus, YAMANOUCHI 21), the centrosomes are permanent 
organs of the cell and form the spindle by moving around the nucleus 
so as to include the chromosomes between them, organizing the 
spindle by the union of those astral rays which penetrate the nucleus 
in a manner very similar to that prevalent among animals. This 
process is very conspicuous and has been observed and figured by 
numerous investigators. In the other type, which occurs in the 
oomycetes (e. g., Saprolegnia, Davis 3; and Albugo, STEVENS 16), 
centrosomes are either small or absent in the early stages; to this 
type the spindles of Synchytrium belong. Here the determination 
of the origin of the spindle is very difficult. This probably accounts 
for the fact that very few of those who have figured such mitoses have 
given a series of figures of the prophases complete enough to throw 
much light on the formation of the spindle. 
When first seen, the spindle of Synchytrium, which is thrown 
directly across the cavity, is not distinguishable by its staining reaction 
or otherwise from a strand of the spirem (figs. 6-8). It is therefore 
difficult to be certain just when the spindle appears or what it comes 
from, but it gives every indication of being differentiated from 4 
spirem strand. Very soon, however, it becomes sharply pointed and 
quite different from the spirem, which now begins to be drawn 1D 
around its equator (figs. 9, 10). Contraction continues and more 
definite connections between the chromatin and the spindle fibers are 
established (jigs. 11, 12). In this condition the spindle often appe@t> 
bipolar (fig. 9), which emphasizes the similarity of the linin strands 
and the spindle fibers. By further contraction this much shortened 
