( 414 ) 
and thicker ones have much contracted and form at both poles, 
adjacent to the daughter-nuclei, two small caps which soon disappear 
in the protoplasm. The long threads on the other hand remain between 
ihe daughter-nuclei, extending from one to the other and hence are 
often called connecting threads. They occur in numbers from 200 to 
300 and cross-sections show that they form a massive bundle lying 
free in the surrounding protoplasm, which can freely penetrate 
between them. 
Hence Mr. SYPKENS arrives at the conclusion that the nuclear 
spindle is entirely formed from the cytoplasm within the nuclear 
space and so agrees with what has been found by most other inves- 
tigators and on main points also with the results obtained by STRASBURGER 
and Heeser for the nuclei of Pritdllaria. 
Now with regard to the part played by the nuclear spindle in 
cell-division zoologists and botanists have divergent views. Con- 
cerning animal cells the general opinion is that the nuclear spindle 
is dissolved in the cytoplasm after the nuclear division has been 
completed and takes no active part in cell-division, the cell subse- 
quently dividing by constriction. Botanists on the other hand, attach 
great importance to the spindle in cell-division and especially in 
the formation of the wall. Their generally accepted representation 
is that the above-mentioned connecting threads of the spindle grow 
thicker in the equatorial plane and form so-called dermatosomes. 
By fusion of the dermatosomes the so-called cell-plate is then formed, 
which subsequently participates in some way or other in the for- 
mation of the new cell-wall between the nuclei. STRASBURGER is one 
of the chief representatives of this much spread conception. 
It is a consequence of the fact that the study of this phenomenon 
has for the greater part taken place with nuclei that were seen 
from the outside. By means of his sections Mr. SYPKENs was enabled 
to prove that, for the objects studied by him, the opinion now 
prevailing in botany is incorrect and that, at any rate as far as the 
behaviour of the nuclear spindle is concerned, the phenomena have 
great resemblance with those of animal cells. 
In describing the later phases of the nuclear spindle it is desirable 
to distinguish three different cases of nuclear division. In the first place 
we have the free nuclear divisions in the parietal layer of protoplasm of 
the embryo-sac of Fritdlaria, whieh will be followed by still other 
nuclear divisions before there is question of cell-divisions. Here in the 
beginning a system of connecting threads between the daughter-nuclei 
appears, as in all other cases, but this soon becomes narrower at the 
equator and so assumes the shape of an hour-glass and is then 
