EMBRYOLOGY 355 
is still obscure. The type of leaf seen in the cotyledon is repeated in 
the “protophylls,” but without definiteness of position or number upon the 
enlarging tuber: their sequence is closed at last by the activity of the 
stem-apex, close to which in time and in position the first root appears. 
It is as though a rootless phase of morphological anomaly, initiated by the 
parenchymatous swelling in the upper tier, were intercalated in the regular 
embryogeny of the Se/ago type, immediately after the origin of the cotyledon: 
and after a period of digression the normal embryogeny were then resumed. 
The swelling is associated in Z. cernuum and inundatum with the entry of 
a mycorrhizal fungus, which occupies the tuber: it must at present remain 
uncertain whether or not this symbiotic state is the cause or a mere 
concomitant of the tuberous condition: and what the relation of it to 
the late appearance of the root; but given the tuberous state, the other 
anomalous foliar conditions readily follow. The proneness of the Lycopod- 
embryo to such secondary swelling as contemplated is seen also in the 
embryos of the clavatwm-type: it is also shown by the repetition of such 
swelling upon the roots in Z. cernuum itself, as have been fully described 
by Treub. 
The cernuum-type of embryo is shared by Z. zxundatum, but not in 
its extreme form. It is this species rather than Z. cernuum itself which 
gives the link to Pxyloglossum. The strobilus of the latter is like a very 
simple strobilus of Z. zmuwndatwm: this species, as is well known, perishes 
in winter, excepting the tip of the trailing stem, which perennates. If 
such a condition were still further prepared for, and condensed by the 
formation of an adventitious protocorm in cases where the plant has been 
fertile, or of a similar body as the product of direct apical growth where 
the plant of the previous year was sterile, the condition of Phy/loglossum 
would be attained. It is interesting to note in this connection that Goebel 
has found that adventitious protocorms are formed in L. ixundatum, a fact 
which strengthens the suggestion here made.!' It would thus appear that 
Phylloglossum, so far from being a prototype of Lycopodinous development, 
is more probably a specialised offset from it. I still adhere to my thesis as 
stated in 1885, that “it is a permanently embryonic form of Lycopod.” But 
it may now be added that the characters which it repeats each year appear 
to be those of a secondary rather than of a primitive embryonic type. 
And thus the embryogeny of the Eligulate Lycopods, so far as at 
present known, conforms to a single central scheme with variations upon 
it. The type of Z. Sedago, the only species of the ‘Se/ago” section of 
the genus in which the embryo has hitherto been observed, is held to 
be the most primitive, as it is certainly the simplest. The rest may be 
held to be secondary variants on that type, due to changes for the 
most part biologically intelligible. s 
1 Bot. Zett., 1887, Plate II., Fig. 32. 
