58 THE ORIGIN OF THE SEED-PLANTS 
course of the bundles and so on) the resémblance vanishes. The analo- 
gies with recent Ferns can clearly only show a certain parallelism of 
evolution, not a genetic relation. 
When we compare the Pteridosperms anatomically with their con- 
temporary Ferns, we find no approximation whatever between them. 
Fern-stems of the Psaronius type agree with those of Medullosa in 
being polystelic, but differ in every other respect. The older and sim- 
pler Primofilices, now very well known as regards two families, have a 
peculiar anatomy of their own, perfectly distinct from that of any Pte- 
ridosperm. | 
It was an anatomical detail, discovered by Professor Weiss, which 
first suggested to the present writer that the Pteridosperms had no di- 
rect connection with the Ferns. Van Tieghem had shown that one of 
the differences between Vascular Cryptogams and Phanerogams is to 
be found in the orientation of the plate of primary wood in diarch la- 
teral rootlets. In the Vascular Cryptogams the plate is at right angles 
to the axis of the main root ; in Phanerogams it is parallel to it. Profes- 
sor WEISS discovered that in Lyginopteris the latter is the case. It is a 
small point, but a definite one, and shows that in this respect the Pte- 
ridosperm Lyginopteris was a Phanerogam, pure and simple 1). 
As regards the external characters, all botanists know that habit is 
illusory. Yet, when the likeness is very close, it may impress the mind, 
as it has certainly done in the case we are considering. It is still often 
impossible to say, from the fronds alone, whether a given Carbonife- 
rous plant was a „Seed-Fern”’ or a true Fern. We are able, however, to 
demonstrate that in such cases the habit is no guide to affinities. One of 
the Ferns which Hooker cited as particularly close to the modern Bra- 
cken was an Alethopteris (then called Pecopteris heterophylla). The ana- 
tomy of Alethopteris is now well known, and proves to be totally diffe- 
rent from that of any Fern-frond and very similar to the leaf-structure 
of a Cycad. Here, then, the habit was clearly deceptive—the resem- 
blance was merely an external one, not necessarily any more signifi- 
cant than the familiar likeness between a succulent Euphorbia and a 
Cactus, or between a Frogbit and a Waterlily. The special conditions 
under which the Carboniferous vegetation had to grow may well ex- 
plain a similarity of habit in quite diverse groups. 
') F. E. Weiss, „The Root-apex and Young Root of Lyginodendron, Man- 
chester Memoirs, Vol. LVII, No. 16, 1913. : 
