OF THE FOSSIL PLANTS OF THE COAL-MEASUEES. 
243 
outline, I cannot ascertain its structure ; it seems to merge in the prosenchymatous 
tissue (b) with which it is invested. It is possible that this latter tissue may have 
formed the true endotesta. I have already shown that the nucular membrane of 
Lagenostoma ovoides appears as if little more than the innermost layer of the true 
testa, with which it is often more or less united by torn fibrous threads. Hence it is 
not improbable that in the seed under consideration all these structures, including the 
perispermic membrane, may have remained permanently adherent to each other, 
excepting near the apex of the seed, where the formation of the lagenostome involved 
their partial separation. This seed is also from the Oldham nodules. 
The next examples belong to a somewhat different modification of this Carboniferous 
type ; and since in both the two forms of it which I have found, what appears to repre- 
sent the supposed lagenostome has a curious funnel-shape, I have assigned to them the 
provisional name of Conostoma. Figs. 80, 80*, & 81 are three sections of a distinct species 
from the Oldham nodules for which I propose the name of Conostoma oblonga. Fig. 80 
is a longitudinal section made nearly through the centre of the seed, enlarged 15 dia- 
meters. The testa (a) is here thin throughout the greater part of the seed, but thickened 
at its apex. Fig. 80 a represents a longitudinal section of this latter part, broken away 
from the end (p) of fig. 80, and enlarged 35 diameters. Before I made a section of 
this fragment its inner or lower surface displayed a small central projection which fitted 
into the upper end of fig. 79. It is composed of numerous long narrow cells or fibres, 
arranged in parallel lines more or less perpendicularly to the surface of the seed at its 
apex, but which, at its sides, are curved, so that, at their inner extremities, the fibres 
verge towards parallelism with the surface of the nucleus. Some of these long, narrow 
cells have a diameter of ‘001, others of -0007. Each apparent fibre usually consists of 
more than one cell arranged in each continuous line. Their transverse subdivisions 
are most frequently square, at others they are oblique. Their internal cavities are 
filled with dark carbonaceous matter, hence the dark lines giving to the testa its fibrous 
aspect. The regularity of their arrangement, combined with the equally regular 
external outline of the seed, indicates that this testa has been hard and dry. I have not 
detected the micropyle in fig. 80 a, though I searched for it. It has probably been a 
minute one, and hence, having escaped my observation, the section has passed on one 
side of it. Within the testa, and running parallel with it, is a very thin membrane (f), 
which is apparently the nucular membrane. On reaching the level of the apex of the 
nucleus ( n ) this membrane splits into two layers. The outer one (f") curves upwards 
and then inwards and downwards on each side, terminating in a little mass (p) of what 
appears to be cellular parenchyma, and which I presume is a dwarfed modification of a 
lagenostome. The inner layer (f) of the nuclear membrane proceeds horizontally 
inwards from each side, and then curves upwards so as to form a dome-shaped arch, the 
keystone of which is again the cellular mass ( p ). The breaks in its continuity in this 
specimen are obviously accidental ones. Within this nucular membrane and completely 
detached from it is the perispermic membrane ( g ), investing the nucleus (n). I have 
