12 BOTANICAL GAZETTE [JULY 
structure of paleozoic seeds of any sort, it would be rash indeed 
to suppose that the known types of seeds were the only ones found 
among Cordaitales, and even more rash to generalize more than 
provisionally on the assumption that they did not possess a particu- 
lar type. We do not know whether the seed habit was developed 
in the phylum once or more than once, much less whether it was 
developed before or after the organization of cones. In short, this is 
today a strong argument for the lycopod theory, that the dis- 
coveries of tomorrow may become an equally valid argument for 
the cordaitean theory. 
A third group of resemblances between lycopods and conifers 
is presented by the leaves. The small, narrow, uninerved type 
of leaf so characteristic of lycopods is very common among the 
conifers. The arrangement in many cases is also similar. The 
gradual transition from leaves to sporophylls in the lycopods pre- 
sents a very close resemblance to certain araucarians. Most, 
or perhaps all, of these resemblances, indeed, can be explained away, 
but that is just where their strength as evidence for this theory 
lies, they do have to be explained away. 
On most other points the theory appears to be on the defensive. 
It can, to be sure, offer more or less plausible explanations and 
possibilities for some of the evidence that appears to be against it, 
but still it must explain them in some other than the obvious way 
to bring them into harmony with itself. The first and most 
serious objection in the opinion of most of its opponents appears 
to be the structure of the stem, more particularly the stele. Not- 
withstanding STILEs’s ingenious and convincing exposition of the 
evolutionary tendencies of the lycopod stele, it yet remains true 
that no known lycopod did attain to the possession of a mesarch 
or endarch siphonostele with leaf gaps. That they might have 
done so appears very probable, but there is yet no evidence that 
they actually did so, and much less that any one that could be 
supposed to be a form ancestral to the conifers had even nearly 
approached it. The same sort of objections apply with even greater 
force to the attempts to explain the origin of the staminate cone 
structures. It is admitted by most botanists that septation has 
probably occurred in certain lycopod sporangia. It may even be 
