106 DEVELOPMENT OF THE FLOWER. 
are minutely and slightly granular. The mass is quite free 
from the cell, and is as usual enveloped in grumous amorphous 
matter. Pl. 29.— Fig. 8. e. 
A little earlier, and the mother cells present a nucleus only, 
no aggregation into groups ; the mother cell is of a fine 
blueish-colour, the disc tinged fleshy; it looks just like the 
membranes, > a granules do not reach the margin of the 
inner disc. 9.—Fig. 8. f. 
The parent ien are now evens: hyaline blueish, scarcely 
separable although distinctly marked out, containing an irre- 
gular grumous nucleus. Pl. 29.— Fig. 8. i. 7. 
Nucelles separated into young grains, with a simple edge 
and granular disc ; mass still slightly coherent at the circum- 
ference. Pl. 29.— 
The grumous mass doirn to the shape of the cell, and 
shews traces of cellular sub-division. Pl. 29.—Fig. 8. l. k. 
Cactus. Quadrilocularity 1st developed, tissue of nearest 
anther quite continuous ; pressure causes the escape of a little 
grumous fluid. (Pl. 38.—Fig. 3.) Less developed anthers 
cells contain a coriform grumous mass, with faint traces of 
cellularity. Those immersed in the mass as it were, appear 
to be the parent cells: membrane hyaline, not very distinct, 
large grumous nucleus. Grumous mucilaginous matter, with 
many subcelluliform bodies. The stigmatic canal of the per- 
fect flower is highly developed; it presents 6 angles and as 
many deep sinuses ; vascular fasciculi 6; opposite the angles. 
Points worth examining—Non-correspondence of ovulifer- 
ous lines with stigmata, irregularity of ovula, identity of the 
whole inner surface of the ovary. 
| In Crucifere there are traces in the very young state of 4 
involutions of leaves, these are pretty equal, but the two 
which subsequently become the ra ap are rather 
the latest. See a, b, c, d. Pl. 29.— j 
When the flower bud is half Aevehiniets there are but 
obscure indications of the limits of these 4 on a transverse 
section. The venation is 1 dorsal, 2 approximate placenti- 
gerous. . During expansion however, the limits on a trans- 
verse section (A) are evident, the placenta appears a grumous 
