90 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
are certainly not identical. The limiting layer is thin at first, 
becomes gradually thicker, but is never as clearly definable at its 
edges as are the membranes that follow. It consists of a substance 
that possesses a different staining capacity from the surrounding 
cytoplasm and while the cytoplasm is more or less coarsely reticu¬ 
lated or foamy in structure, it on the other hand is apparently struc¬ 
tureless or very finely granular. But it is never separable from the 
protoplasm on either side. The cases of separation figured by 
Harper are evidently separations between the plasma membranes. 
Now it is not improbable that with the increased growth and differ¬ 
entiation of the limiting layer, and a pull on the part of the nucleus, 
that the cleavage is first affected in the part adjacent to the centro- 
some. It is here that the membranes first begin to form, and it is 
this part that would be directly affected by a nuclear tension. A 
continuance of cleavage would be favored by the formation of mem¬ 
branes on the surfaces already separated and much more by the 
differentiation of the spore plasma which by this time at least is in 
course of progress. 
The membrane covering the epiplasm undergoes no further 
change except in extent and shape according as the spore changes 
in this respect. It continues to line the epiplasm until the spores 
are developed, and is the last of the epiplasmio substance to dis¬ 
appear. The membrane covering the spore plasm likewise con¬ 
tinues for a time without other changes than in extent and shape to 
meet the demands of the growing spore, and the various pressures 
brought to bear upon it. When the spores are nearly matured, 
however, the exospore begins to be laid down, and with this process 
perhaps both plasma membranes are intimately connected. 
Subsequent to the formation of the plasma membranes one of the 
chief matters of interest until the spores are mature, consists in the 
resumption of a spherical form by the nucleus. This change takes 
place very soon after the completion of the membrane, though the 
nucleus may first elongate still farther as shown in figure 32 (pi. 8). 
The neck of the fiask-like nucleus contracts, and the result is to 
thicken it and draw the centrosome away from the periphery 
towards the center (pi. 8, fig. 34). In one respect Hydnobolites 
differs from the forms already described, for as yet the astral rays 
have not vanished. Indeed, they are demonstrable until the beak 
is fully withdrawn (pi. 8, fig. 35), although at that time they begin 
