258 
PROFESSOR W. 0. WILLIAMSON ON THE ORGANIZATION 
is imperfect, enough remains to show that it, too, had a lenticular outline. Within this 
testa we now find two membranes, of which the inner (g) probably represents the peri- 
spermic membrane. Fig. 125 is one surface of another transverse section, made parallel 
to the last, but near .the base of the seed. Being the opposite surface of the thick 
section, fig. 124, its position is reversed — the thick layer of the exotesta, a', being now 
on the upper instead of the lower side of the figure. The parenchyma of the chalaza 
is seen at i, on each side of which are two basal prolongations of the cavity enclosed 
within the testa, and which project below the chalaza, indicating the cordate shape of 
the endotestal cavity at the chalazal extremity of the seed. 
Figs. 126, 127, & 128 are three serial longitudinal sections of another specimen made 
vertically to figs. 121 & 123. The first of these is a central one. Unfortunately the 
base of the specimen was lost ; but we now find that the testa is divisible into two 
layers. The outer one (a) is more or less disorganized, but various indications show 
that it consisted of a parenchymatous tissue. Thus at a' we have a line of cells which 
might easily be mistaken for an epidermal layer ; but we shall find similar cells in 
fig. 127 at a!, and still more at occupying the innermost portions of the same tissue. 
Combining these sections we appear to have evidence that the exotesta consisted of a 
coarse parenchyma, many of the cells of which have a diameter of ’003. Within this 
is the more dense endotesta, fig. 126 b , which is here about ‘01 in thickness, and consists 
of parenchyma, the cells of which are of much smaller size than those of the exotesta. 
At the apex of the nucleus these united portions of the testa suddenly contracted to 
form the micropyle, d, which was the narrowest at its nucular extremity, and gradually 
expanded into a trumpet-shaped aperture at its open mouth, corresponding in this 
respect with fig. 123 ; but it must be remembered that the latter section is made along 
a line between the points y,y of fig. 124, whilst fig. 126 is made across the shorter 
diameter of the seed from x to x. At fig. 126, b', b\ we see a tendency to the separation 
of a thin membrane from the interior of the endotesta, which may be the same as 
fig. 124, b. The tissues enclosed within the testa have become much contracted and 
shrivelled, especially towards the apical half of the seed — a condition which was equally 
seen in fig. 123. In fig. 126, f, we have an outer membrane, and at g there appears 
to be an inner one. The former is shrivelled up, as in fig. 123, to little more than a 
mere line at the upper half of the nucular cavity, whilst inferiorly the two membranes 
f Sc g unite to invest the perisperm n. Thus in both the specimens examined all traces 
of the nucleus (n) seem to have wholly disappeared from the upper half of the interior 
of the seed, whilst more of it remained at the lower half, where, both in figs. 124-126 
& 127, we see clear indications of what may be regarded as a double perispermic 
membrane, /A/. Unfortunately none of my sections show whether or not the nucleus, 
with its investing membrane, originally occupied the interior of the upper part of the 
membrane f, or whether the latter was largely detached from the perispermic membrane 
to form a lagenostomal cavity. This point requires further investigation when favour- 
able specimens can be obtained. 
