494 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL.. VI. > 
or underlie a corpuscle, giving exactly one of the conditions described 
and illustrated by Janssens and Leblanc. 
The corpuscle does not in the fresh cell react with acetic-methy] green 
like a chromatin body. Preparations of this reagent, which would bring 
out clearly in fresh animal and vegetable cells the chromatin-holding 
structures, left the corpuscles of the yeast cell, even after hours, unaf- 
fected, while the cytoplasm stained readily. Another point of contrast 
is to be found in its affinity for eosin, which is more readily absorbed 
and retained by it than is hematoxylin. 
In the throat yeast the corpuscle was, in the great majority of 
instances, more or less irregular in outline, always excentrically placed, 
and in the great majority of cells, in close contact with a vacuole. 
Sometimes it was crescentic in outline, the vacuole fitting into its con- 
cavity. In this form very little chromatin was found in the cytoplasm 
and as a consequence the corpuscle stood out quite clearly. 
In S. Ludwigit as it grew in the sap of the iron-wood tree, fixation 
with Flemming’s fluid and staining with iron-alum haematoxylin and 
eosin demonstrated the corpuscle frequently as a reddish body, having 
at times a tint of blue-violet and surrounded by a garland-like structure 
formed of deep blue-violet-stained chromatin. This structure, on closer 
analysis, is found to be formed of granules and elongated masses of 
chromatin. Sometimes it appears open or discontinued at one side, and, 
especially if the corpuscle is pear-shaped, its prolongation extends 
beyond the limits of the chromatin structure. Usually, however, in these 
preparations the structure is closely applied to the corpuscle. When 
there are two corpuscles in the cell there is astructure in question about 
each of them. Rarely the corpuscle appears separated from the cyto- 
plasm, and, consequently, from the garland-like structure, by a zone of 
clear space. (Fig. 28). 
I regard this garland-like structure formed of chromatin as quite the 
same as the membrane of fine granules closely applied to, and com- 
pletely surrounding, the “ nucleolus,” as described by Wager, who found 
that the granules and nucleolus stained differently. In my preparations 
of S. cerevistew the membrane does not appear as uniform and as regular 
as Wager figures it; and its irregularity, and sometimes the absence of 
close contact between it and the corpuscle, remind one strongly of the 
conditions in S. Ludwigzi. The point to note specially is the contrast 
in staining presented by the corpuscle and the structures surrounding it. 
The latter has a marked affinity for hematoxylin, while the former 
absorbs eosin readily. Sometimes, in corrosive sublimate preparations 
la ag vy 
