486 MISS T. L. PRANKERD ON THE STRUCTURE OF 
at the slightly constricted junction of the plinth with the lagenostome wall. 
But whether this took place first or last is a small matter compared with the 
necessity that dehiscence should be secured right to the base so that the 
sperms might have free access to the archegonia. Complete circumscissile 
dehiscence was not absolutely necessary and did not seemingly always take 
place, except at the base, where we have an organ specially modified to secure it 
Finally, a word may be added on the morphology of the lagenostoine. It 
is generally admitted that this unique structure is no doubt the modified apex 
of the nucellus, and indeed it is difficult to see to what else it could be 
attributed. Now the nneellus is an enclosed, more or less specialized, 
megasporangium, and we do not lack fern sporangia showing features which 
suggest the type that may well have formed the basis for such specialization. 
The sporangia of -Ingiopteris, Osmunda, and Schizea all have apical or sub- 
apical annuli, of which genera Osmunda is particularly interesting, since it is 
a synthetic type combinin g in itself several of the characters of the eu- and 
lepto-sporangiate ferns. But i in the fossil Senftenbergia we find what would 
seem a closer parallel in a multiseriate, apical annulus. The time needed to 
evolve other features of our seed, such as the integument, is at least sufficient 
for the modification of such a type as Senftenbermia into the nucellus of 
Lagenostoma, from which it is surely not so far removed in form or 
function * 
The comparison very tentatively put forward by Lang (8) between certain 
thickened cells in the microsporangia of Stangeria paradora and the annulus 
of the Sehizæaceæ, and further between the micro- and macrosporangia of the 
former, is strengthened by the hy pothesis of the retained annular dehiscence 
in Lagenostoma, the mechanism for which is perhaps vestigial in its nearest 
ally among recent plants—the Cyead ovule, cf. (8) p. 302. 
It has already been suggested (10) that the trac ‘heal mantle at the periphery 
of the nucellus—so common a feature in fossil Gymnosperm seeds—is homo- 
logous with the vascular tissue found within the wall of a fossil sporangium. 
There is surely no greater difficulty in deriving the wall of the lagenostome 
from the annulus of another fossil sporangium. Both are specialized parts of 
the sporangium wall: and here too, as in the analogous case just cited, there 
would be some change of function “in harmony with the modified fate and 
enlarged functions of the macrosporangium ” ; for, whereas in the ancestral 
fern sporangium dehiscence was effected to let the contained spores out, in 
the megasporangium it takes place in order to let the microspores in. 
If there is any truth in this surmise, the Lyginodendres have preserved 
* The suggestion is put forward as much for Z. Lomaxii as for L. ovoides ; for though not so 
well marked as in the latter, the former species shows the radial thickenings, and even the 
curved outer walls, while the inner tangential walls are generally retained. The transverse 
section of the lagenostome wall in R. 16 b is, to me, remarkably suggestive of an annulus. 
