44 BOTANICAL GAZETTE [JULY 
diatoms, and his beautiful preparations, which it, was our pleasure 
to examine, show these bodies as sharply defined as in most 
animal mitcses. Both Lauterborn and Karsten agree that a 
centrosome, or at least a body derived from it, becomes cylindri- 
cal or ring-shaped, and functions as a spindle during mitosis. 
The centrosomes of diatoms stain intensely and are not sur- 
rounded by acentrosphere. Lauterborn found centrosomes even 
during the resting condition of the nucleus and cell, but Karsten 
was not able to identify the body positively until the radiations 
began to appear. Davis (5) describes a centrosphere, but no 
centrosome, in the tetraspore mother-cell of Corallina. The 
centrospheres give rise to the spindle, and consequently play an 
essential part during nuclear and cell division. No centro- 
spheres could be recognized during the resting-stage of the 
nucleus. 
Thus it appears that in many of the algae well-defined centro-. 
somes are present, at least during certain phases of the life- 
history, and that the centrosomes may divide and persist from 
one cell-generation to another, while in other algae the centro- 
some does not show such a degree of permanence. In the algae 
which we have mentioned the centrosomes are not surrounded 
by aclear area. In Corallina it is to be noted that there is no 
centrosome, but only a centrosphere. In none of the algae 
have centrosomes been traced throughout the life-history of the 
plant. In some fungi centrosomes are present during the mitoses 
concerned in the development of spores. Among the liverworts 
we doubt whether there is, at any period in the life-history, a 
centrosome like those described for the thallophytes. The cen- 
trosphere, appearing and functioning during = a few mitoses, 
has replaced the functional centrosome. 
The polar radiations which are often conspicuous during 
mitosis in pteridophytes, gymnosperms, and angiosperms are 
of the same nature as those of thallophytes and bryophytes, but 
in the higher groups (and, possibly, in most mitoses in the lower 
groups) a definite centrosome, or even a centrosphere, is lacking. 
Centrosomes and centrospheres in vascular plants have been 
described and figured so frequently by such competent observers 
| 
, 
| 
