THE ORIGIN OF THE SEED-PLANTS 97 
is 
gin of the seed than what recent plants (such as Cycads) might already 
have told us. 
It is very different where we really have evidence that a sort of seed 
was developed in an otherwise Cryptogamic phylum. Among the Car- 
boniferous Lycopods there are two genera, Lepidocarpon and Miades- 
ma, Which are known to have formed seed-like organs. An integument 
grew up around the megasporangium, in which only a single megaspore 
(embryo-sac) came to maturity. Here, however, the Lycopod characters 
are perfectly evident, both in the sporangial and vegetative organs. 
These plants are truly Cryptogams which had become (after a fashion) 
Seed-plants, but there is not the least probability that their imitation 
seeds had anything to do with the real seeds of the main line of Sper- 
mophyta. The , seed-bearing’”’ Clubmosses, in fact, came too late upon 
the scene; the old-established Seed-plants were already too strong for 
them, and it seems that after a comparatively short career they came 
to nothing. 
The characters which have appeared to support a direct affinity bet- 
ween the Pteridosperms and the Ferns are to be found in the male or- 
gans, the anatomy, and the habit. As regards the first, it is no doubt 
often difficult to distinguish the sporangia of a Carboniferous Fern 
from the pollen-sacs of a contemporary Pteridosperm. Our knowledge 
of the latter is, however, still very imperfect. Our first information came 
from Dr. KIDSTON, who, in 1906, identified the male fructification of 
Lyginopteris Oldhamia. He found that the pollen-sacs, in this case, are 
bilocular. If this should turn out to be general, there would be little real 
resemblance to the sporangia of a Fern. The evidence in other cases, 
since discovered, is usually in the form of impressions, showing little 
of the actual structure. 
Great stress has been laid on the Fern-like anatomy of various Pteri- 
dosperms. For example, the primary structure of the Lyginopteris 
stem has been compared with that of Osmunda, and the structure of 
Heterangium with that of Gleichenia. Here, however, the comparison 
was with recent Ferns, which cannot have any real affinity with Pale- 
ozoic Seed-plants. Osmundaceae, it is true, go back to the Permian, but 
their early reprsentatives have quite a different stem-structure from 
that which was compared with Lyginopteris. 
Medullosa was polystelic, like most recent and many Carboniferous 
Ferns, but the moment we begin to examine the details (development, 
