Graham. — Nuclear Division of Preissia committal a. 665 
found. Rosen (’95) believes that they grow from the nuclear membrane 
towards a point some distance from it. 
Polar caps such as are described above have been observed by many 
investigators. Many have figured them in the Algae (Spirogyra). Hof (’98) 
describes them in a Fern, a Gymnosperm, and an Angiosperm, and Chamber- 
lain (’03) in the germinating spores of Pellia epiphylla. 
The nucleus in Preissia commutata elongates, as has been said, in the 
direction of the long axis of the cell, and the polar caps and spindle fibres 
appear against the nuclear membrane at these elongated ends. These 
fibres, therefore, appear as portions of the nuclear membrane, against which 
there is a greater mass of cytoplasm than there is on the sides of the 
elongated nucleus. The entire cells from which Figs. 13 and 14 are drawn 
are not represented ; some portion at each end is omitted. The long axis of 
both these cells is parallel with the long axis of the nucleus. These fibres, 
therefore, appear on portions of the nucleus against which there is the 
greater amount of cytoplasm. The behaviour of these bipolar caps accords 
with Nemec’s (’99) theory, namely, that they appear in vegetative cells on 
the sides of the nuclei against which the mass of cytoplasm is the greater. 
While the spireme is segmenting and while the nuclear membrane is 
still intact, a considerable number of spindle fibres have developed in the 
polar caps. Part of the spindle, therefore, is derived from the cytoplasm. 
Diligent search was made for a centrosome at the apex of the con- 
verging fibres, but none was found ; neither were there any cytoplasmic 
radiations. Many granules are in the cytoplasm at this stage, some of 
which may happen to lie at the apex of the cone of fibres, as shown in 
Fig. 13. The poles of the spindle at metaphase, as is shown in Figs. 16, 18, 
and 19, are sharply pointed, and, arising as they do from bipolar caps, it is 
at these poles that a centrosphere or centrosome might be expected, if it 
existed. No structures which could be justly interpreted as centrospheres 
or centrosomes were observed prior to or during the formation of the 
spindle. In one instance a multipolar diarch spindle (Fig. 18) was observed. 
The formation of a spindle of this type was not followed. 
At the time the cone of fibres is fully developed at the ends of the 
nucleus, the spireme begins to break up into segments. Fig. 13 shows the 
spireme partly segmented. At this stage the segments are very little broader 
than the continuous spireme. Soon, however, they become broader and the 
ends of each chromosome curve (Figs. 14, 15). This curving makes it 
difficult to determine their exact form at metaphase. In longitudinal 
section of the spindle at metaphase the chromosomes appear as oval bodies. 
After segmentation of the spireme appears to have been completed, the 
segments are large and there are probably not more than eight (Fig. 14). 
In Fig. 15 the segments have shortened somewhat and there are probably 
not more than eight. In longitudinal section of the spindle at metaphase, 
