i 
1906] LEWIS—DEVELOPMENT OF RICCIA 131 
centrosome and the blepharoplast are homologous, and this is the 
conclusion of keno. In all other plants in which blepharoplasts 
are known to occur, centrosome-like bodies are not present in any 
cell divisions except those immediately preceding the formation of 
the sperms. That centrosomes occur in liverworts in cells outside 
the antheridium is open to question. The conflicting reports of 
those who have investigated Pellia epiphylla make it clear that no 
distinct body occurs there which can be regarded as a centrosome, 
although aggregations of kinoplasm, called centrospheres by most 
authors, do occur. 
In Riccia natans, it seems very evident that centrosomes do not 
occur in the divisions of the spore mother-cells. The spindle poles 
are broad, and there is not even a suggestion of a centrosphere such 
as has been described ‘for Pellia. In the cells of the sporophyte 
GARBER reports centrospheres but no centrosomes. I have never 
been able to observe them in my preparations. When the spindle 
is fully formed there are no fibres radiating into the surrounding 
cytoplasm (figs. 48-52). 
Although the thallus of Riccia natans does not present favorable 
material for cytological study, a number of cells showing nuclear 
division in the gametophyte have been observed near the growing 
point. The greatest difficulty here is the presence of numerous 
deeply staining granules in the cell. In some cases granules resem- 
bling centrosomes appear at the poles of the spindle, but they do 
not differ in appearance from the other granules of the cell, and it 
seems probable that their occurrence here is accidental. 
Summing up, we find that in Riccia natans centrosomes are not 
found in the cells of the gametophyte, sporophyte, or spore mother- 
cells, but that bodies occur in the dividing cells of the antheridium 
which seem to function as centrosomes. In Riccia and Marchantia, 
the blepharoplasts certainly have much more the appearance of 
centrosomes than in any other plants in which blepharoplasts have 
been described. The bodies have every appearance of centrosomes 
when at the poles of the elongated nucleus or at the poles of the 
spindle. Perhaps the strongest objection to regarding these bodies 
as centrosomes lies in the fact that in Riccia natans they occur only 
in the cells of the antheridium, while the blepharoplasts reported 
