OF THE FOSSIL PLANTS OF THE COAL-MEASTTEES. 
247 
micropylar cavity. Fig. 91 shows still more clearly the tissues of the sarcotesta ( a ) and 
the nucular membrane (j f) ; but we now see, in addition, a small portion of the surface 
of a yet more internal membrane ( g ), which I presume is the perispermic membrane. 
At the micropylar end we still find what may be termed the 
exostome, or orifice of the sarcotesta, open ; but the two rings h f f ? f e \ 
and h' are now linked together by a continuous band of paren- 
chyma, though their open canals are still distinct. In fig. 92 
we advance still further. The section has now passed through 
the perispermic membrane (</), whilst the two rings h, h! are now 
united into a common oblong cavity, h. These sections make 
it clear that the rings of parenchyma of the other figures and 
the blended portions of fig. 92 form parts of a tube or canal 
bent into the form of a tube, which the section fig. 89 has 
intersected at a in the accompanying woodcut, fig. 90 at b, 
91 at c, 92 at d, and 93 at e. 
Whether the flexed condition of this tube was a normal 
one, or whether it has been the result of partial desicca- 
tion is doubtful ; but I incline to the opinion that it was 
originally straight. In none of the sections was I able to trace any connexion between 
this tube and the membranes investing the nucleus ; but unfortunately I had only half the 
seed to work upon ; had I had the other half, so as to have been able to follow up the 
section fig. 93 with another one, in all probability the transverse thin band of sarcocarp 
seen at fig. 93 a! would have disappeared, and the direct continuity of the micropyle 
from the exostome to the nuclear membrane have been demonstrated. Fig. 93 represents, 
as I have already pointed out, the section nearest to a central one of any of the series. 
At its apex, a, the exostome of the micropyle has disappeared. A thin extension of the 
sarcotesta now completely arches over the micropylar canal, from which it seems that 
this latter organ was somewhat curved, as well as the parenchymatous tube which it 
contained. In the body of the seed we see the nucular area, n, large and oval, nearly 
filling the interior of the seed, and enclosed within its perispermic membrane (g). 
External to this is the delicate prosenchymatous nucular membrane (f) lining the 
sarcotesta (a). But we now obtain proof that the sarcotesta was hardened at its inner 
border ( b , b) into an endotesta, the tissues of which were much more firm than the 
parenchymatous elements which surrounded it. This is very distinctly seen at the 
chalaza (i), where the hardened layer of the sarcotesta is interrupted, allowing a direct 
communication to exist between the softer parenchyma of the sarcotesta, which is here 
very thick, and the prosenchymatous nucular membrane at/ 7 . At this end of the seed 
we further see the sarcotesta prolonged into what looks like a funiculus (Jc). 
Fig. 88 represents the upper part of fig. 92, enlarged 50 diameters. In this figure the 
differences between the three tissues of the sarcotesta ( a , a), the nucular membrane 
and the parenchymatous structure (h) within the exostome are well shown, especially 
2 m 2 
