384 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
in the integument, the single megaspore, and the indehiscent char- 
acter are evident; we are unfortunately without any evidence as to 
the stage at which fertilization took place. The fossil has long been 
known, but was formerly confused with the Gymnospermous seed 
Cardiocarpon, with which, of course, it has nothing to do. Micro- 
sporangiate cones, probably belonging to Lepidocarpon, are indistin- 
guishable from the cones of a small Lepidostrobus, except that there 
are indications of an incomplete integument around the microspo- 
rangium. 
In Miadesmia, the other genus of quasi-spermophytic Lycopods, 
the sporophyll bears a hgule exactly like that of the vegetative leaf, 
and the lamina is fringed in the same way at the margins. The 
megasporangium is attached, at the proximal end, to the upper sur- 
face of the sporophyll, and contains a single megaspore, filling its 
cavity. Kxternally the megasporangium is inclosed in an integu- 
ment, springing from the upper surface of the sporophyll, and leav- 
ing only a narrow micropylar opening at the distal end, not at the 
top as. in Lepidocarpon. The integument bears long tentacles, 
directed forward, which may have played some part in guiding the 
wind-borne microspores to the micropyle. 
There is thus a general analogy with Lepidocarpon, as regards the 
essential seed-like features, but the structure is quite different in de- 
tail. Of the two the J/iadesmia fructification is perhaps the more 
advanced, for the sporangial wall is less developed than in Lepido- 
carpon, an indication that the protective function had been more 
completely taken over by the integument. Microsporophylls, prob- 
ably referable to the same plant, have been found. They agree with 
the megasporophylis as regards the insertion of the sporangium, but 
no integument is developed. 
It is remarkable that seed-like organs should have been found in 
two genera of Paleozoic Lycopods so different as Lepidocarpon and 
Miadesmia, in each of which the character must, no doubt, have arisen 
independently. We can only conjecture that the circumstances of the 
time may have been peculiarly favorable to the adoption of the 
seed habit. The early development of the prothallus, in the case of 
Lepidocarpon, makes it very probable that pollination, if not fertil- 
ization, took place on the parent plant, but we have no direct evidence 
on the subject. One striking difference from a typical seed is the fact 
that in both genera the whole sporophyll was shed with the mega- 
sporangium, and formed part of its investment. Analogies with the 
achenes and nuts of Angiosperms are too remote to be of service, and 
we must admit that in these Paleozoic Lycopods the participation of 
the sporophyll marks a low grade of seed evolution. In fact it is 
evident that in every respect the seed-like organs in question, even if 
they were functionally seeds, still stand very near the Cryptogamic 
