OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 
259 
Fig. 127, which is a section made along the line fig. 121 x-x , shows the endotesta, b , 
extending upwards into the thin lenticular expansion of the testa which borders each 
side of* the micropyle; and the same condition is seen in fig. 128, which is a similar 
section made through the testa bordering the nucular cavity, i. e. along the line y in 
fig. 121. Fig. 128 is enlarged 50 diameters, to show the cells of the endotesta (b) 
arranged in long vertical lines, whilst the exotesta ( a ) exhibits its usual disorganized 
condition. The line a ' indicates the former boundary of the latter tissue, much of 
which has disappeared. The length of the specimen, fig. 126, has been about -21, and 
the diameter of fig. 124 about '25. All the specimens agree in giving to this species 
much of the general dimensions and contour of the Cardiocarpum acutum of Lindley 
and Hutton, and of the C. Lindleyi of Mr. Carruthers, excepting that the latter 
observer describes his seed as having a central longitudinal ridge, which my specimens 
certainly have not. Since these differences exist, it may be well to distinguish my type 
under the name of Cardiocarpon compressum. 
Since the above descriptions were written, Mr. A. Butterworth has found, near Old- 
ham, the specimen from which he has prepared the section represented in fig. 128*. It 
is a transverse one, made immediately above the cordate base of the lenticular seed. In 
it we have a thick testa, a, within which we find the very distinct membrane/, which, 
I presume, is identical with f in fig. 124. This, in turn, encloses the shrivelled remains 
(g) of what, I presume, has been a perispermic membrane identical with fig. 124, g ; 
but between the two membranes/ and g we have a small number of scattered, free cells, 
l, l , which closely resemble the supposed pollen-grains seen in other seeds. Unfortu- 
nately the micropylar extremity of Mr. Butterworth’s specimen was lost ; hence I cannot 
determine whether it represents a distinct species or is identical with the C. compressum, 
but if distinct the two types are very closely allied. Supposing the objects l, l to be pollen- 
grains, we obtain some light respecting the cavity into which these grains are received 
in Cardiocarpon, and which plays the part of the lagenostome seen in the other seeds. 
A priori, the organization of figs. 116, 117, 118, 119, & 126 would suggest the infer- 
ence that the pollen-grains w T ould pass down the micropyle through the tubular part of 
the membrane (b) of the first four figures, and into the cavity bounded by the membrane 
(S') in fig. 126, in each of which cases the grains would be brought into immediate 
contact with the nucular membrane. If this is a correct explanation, these seeds of 
Cardiocarpon differ little from the simpler types of Coniferous ovules seen in Juniperus 
and Callitris. Thus, in each of the figs. 117 & 126, the pollen would reach the triangular 
cavity ( x ) without the intervention of the more complex arrangements of the lagenostome 
seen in many of the other seeds which I have described. In their internal organization, 
and especially in the great length of the micropyle, some of these seeds bear a strong 
resemblance to the ovules of Welwitschia f. 
Mr. Butterworth was also fortunate enough to find in one of the Oldham nodules 
t “ On Welivitschia, a new Genus of Gnetaceae,” by Joseph Daxton Hooker, F.E.S., Transactions of the 
Linnean Society, vol. xxiv. pi. ix. figs.'ll & 12. 
