1904 } MERRIMAN: VEGETATIVE CELL DIVISION 185 
the reason for the failure of the hyaline periplasts to appear in 
the cells of the wounded tubercles is because of the greater size 
of the nuclei. Nuclei occurring in the tubercles experimented 
upon were 30 per cent. larger than those found in the growing 
points of root and stem. 
' His statement as to the cytoplasmic fibrillae in the cells of 
wounded tubercles growing out directly from the nuclear sur- 
face agrees with what I found in normal cells from the root tips 
of Allium fixed in Flemming and picro-sublimate solution. As 
to the appearance of the periplast, I cannot agree with Némec in 
considering it a normal feature in the process of vegetative cell 
division, and its absence caused by experimentation or otherwise 
disturbed conditions. In some cases its presence seems to be 
due to the action of fixatives or to the sectioning of an irregu- 
larly shaped nucleus. 
Also as to the cytoplasmic aggregations which Strasburger 
and Hof consider as being made up of a specific substance, kino- 
plasm, the evidence derived from comparing the structural 
appearance of these aggregations with the appearance of the 
remainder of the cytoplasm in the various preparations does not 
seem to uphold such a conception. Their appearance rather 
justifies the view expressed by Wilson (5) and others that alveo- 
lar spheres, microsomes, granules, and fibrillae may be mor- 
phologically considered as but gradations of one structure. 
CENTROSOMES. 
In one case only, out of the many preparations of Allium 
examined, was there a substance visible which could be construed 
as a possible centrosome. This case is shown in fig. 34 of a 
nucleus in the spireme stage with a dumb-bell-shaped body lying 
a little at the right of the lower pole. No astral rays are 
present, there being only a clear space about the body inclosed 
by a plasma membrane. The fact that the body was discrete 
and that its shape might indicate its undergoing a process of direct 
division led me to think that here might be a centrosome which 
by reason of its temporary character had before eluded observa- 
tion. The section was from material killed in chrom-acetic and 
