256 
PROFESSOR W. C. WILLIAMSON ON THE ORGANIZATION 
them having a mean diameter of '004 ; a cluster of them, enlarged 50 diameters, is 
represented in fig. 120. 
This perisperm is invested by a very thin membrane, g , in which I can find no struc- 
ture, and which, I conclude, is identical with the perispermic membrane of the previously 
described seeds. It is rounded at its apex, not being prolonged into the micropyle, 
resembling in this respect the Cardiocarpons figured by M. Brongniart *. External to 
this perispermic membrane we have two layers, which, though they accompany each 
other in their course, differ in their structure. The inner one ( b ) is very thin and deli- 
cate, consisting of a single layer of thin-walled prosenchymatous cells. Superiorly it is 
prolonged through the micropyle, and, in this instance, extends beyond the outer layer 
at b' ; but this is doubtless an accidental result. At its base ( b ") it has evidently been 
closely united with the chalazal tissues. The outermost membrane, or exotesta (a), is 
dense and well defined ; though in this instance its parenchymatous structure is not very 
clearly shown. Interiorly it is prolonged into a funiculus or peduncle ( k ), whilst at 
its apical extremity it projects in a long tubular micropyle. 
I cannot determine with certainty the relation which the inner membrane (b) in this 
species bears to those seen in the other seeds which I have described. In its structure it 
corresponds closely with what, following M. Brongniart, I have designated the nucular 
membrane. But in its prolongation to the extreme tip of the micropyle it differs com- 
pletely from the nucular membrane of the other seeds. Hence I am inclined to regard 
it as a thin and delicate endotesta. The other specimens to be described confirm me in 
this conclusion, because, though the perispermic structures have entirely disappeared 
from them, this membrane remains in perfect integrity. Thus fig. 118 represents a 
transverse section of a specimen enlarged 22^ diameters, its mean size being about T. 
The exotesta ( a ) is very sharply defined both externally and internally, and has a mean 
thickness of about '007. In the body of the seed, w r here it has invested the nucleus, it 
expands broadly into two rounded wings ; but at the apex of the nucleus it has rapidly 
contracted into the tubular prolongation a forming the micropyle, and which has been 
curved laterally near its exostomal extremity, d. The inner membrane or endotesta [b) 
exhibits the same delicate prosenchymatous structure as is seen in the corresponding 
membrane in fig. 117. In like manner it is firmly united with the exotesta at the cha- 
laza (7), and is prolonged as an inner layer of the micropyle to the point d. The nucleus 
and nucular membrane are wanting in this example. 
Fig. 119 represents one of the finest specimens I have met with, so far as the preser- 
vation of the tissues of the testa are concerned. It is enlarged 22 diameters. Whether 
it is a distinct species from figs. 116 & 118, or whether it is merely a large example of 
the same, I am unable to say ; but it is more than twice the diameter of fig. 116, and 
more than three times that of fig. 118. In every other respect it corresponds closely 
with those figures. We now see that the exotesta (a) consists of innumerable very 
minute cells with dense cell-walls, and which in some parts, as at a assume a prosen- 
* ‘ Etudes sur les graines fossiles/ pi. xxi. fig. 1. 
