Aster ella hemisphaerica , Beanv . 537 
larger masses of chromatin. In most cases, however, the original nucleolus 
seems to enlarge through a concentration of the chromatin. As the nucleolus 
continues to enlarge and becomes more loosely arranged (Figs. 4 to 7), the 
chromosomes assume definite form. The number of chromosomes seems to 
be not less than eight. A continuous and distinct spireme was not observed. 
The Formation of the Spindle. 
The first indication of the development of the spindle which was 
observed was the formation of cap-like areas of cytoplasm at opposite sides 
of the nucleus (Figs. 8 and 9). These caps may be closely applied to the 
nuclear membrane (Fig. 8), or only slightly removed from it (Fig. 9). These 
caps may remain for some time slightly beyond the extremities of the poles 
of the spindle. Soon after their appearance the nucleus becomes drawn out 
along the axis in which these cytoplasmic caps lie (Fig. 5). From this 
material, very delicate intranuclear spindle fibres soon become organized 
(Figs. 5, 10, and 11). As the spindle reaches maturity, remnants of the 
polar cap may occasionally be seen (Fig. 10). Occasionally a small granule 
may occupy the sharp-pointed pole of the spindle (Fig. 10). However, the 
poles may be sharp-pointed without the presence of either polar caps or 
granules (Fig. 1 1). Or, again, the poles may be broad and truncated (Fig. 1 2). 
Sometimes the spindle may be multipolar (Figs. 12 and 13) with more or 
less conspicuous granules occurring where the fibres converge. A careful 
study of Figs. 8, 9, 11, 12, and 13 would lead to the conclusion that, while 
polar granules occur with irregularity, no bodies are evident which partake 
clearly of the nature of centrosomes. Similar granules may occur with 
irregularity in number and position throughout other portions of the 
cytoplasm. 
A careful examination of the stages represented in Figs. 1,5, 8, and 13 
has convinced the writer that, in Asterella, polar caps and granular bodies 
arise de novo as the nucleus enters upon division. There seems to be no 
basis for assigning to them morphological rank more than to the fibres of 
the spindle. They seem to be temporary structures of protoplasmic 
development, as are the spindle fibres. The caps no doubt actually become 
a part of the spindle. There is ho evidence that the polar caps or granules 
(Figs. 5, 8, 9, and 10) are permanently organized structures. Their appear- 
ance and behaviour suggest rather a visible structure which is built up 
temporarily and which performs definite functions, and then is resolved into 
less differentiated cytoplasm. This conception, supported by cytological 
evidence, seems to the writer to throw some light on the questions sug- 
gested above. 
The reaction of the delicate protoplasmic structures to killing fluids is 
also significant, as a comparison of different cells in different parts of the 
antheridium will show. 
