343 
Phylloglossum Drummondii. 
the U-shaped stele of which is remarkably like that now under discussion. 
This analogy to Tmesipteris is enhanced by the presence of what appear 
to be imperfectly developed tracheides bridging over the gap. 
If there be any significance in this analogy, and the shape of the stele 
demands some such explanation, we must suppose that the gap indicates 
the presence, in an ancestral form, of a leaf, typically ‘ megaphyllous ’ in 
its behaviour to the stele. In this case we have a complete analogy to 
Tmesipteris , the lower leaves of which are clado-, and the upper phyllo- 
siphonic ; and Phylloglossum presents yet another link between the 
Pteropsida and Lycopsida of Jeffrey. 
The leaf-member which causes the gap must be regarded as entirely 
suppressed in the modern form, and there is no indication of the presence 
of a branch in its axil; the structure is, in fact, typically Filicinean. The 
gap never closes above; the U-shaped stele breaks up directly into the 
isolated strands of the strobilus pedicel. 
It must be mentioned at this stage, however, that Jeffrey (8) denies 
that the gaps in the stele of Tmesipteris are foliar, and he points his denial 
by a strict definition of the term ‘ leaf-gap ’; he emphasizes the closeness of 
the relation between the gap and the outgoing leaf. More than one botanist 
of distinction is less certain than Jeffrey upon this point; but, whatever 
may be the true case in Tmesipteris , the gap in the stele of Phylloglossum 
certainly complies with Jeffrey’s criteria of a leaf-gap. In his paper (8) he 
evades this point by assuming, without offering any discussion upon the 
matter, that the opening in the stele of Phylloglossum is caused by the 
exit of the bundle which supplies the young tuber; the fact that this 
opening is found to face the site occupied by the latter at a lower level 
seems to lend some support to this contention. In this problem, how¬ 
ever, we may derive considerable assistance from the stalked tuberous 
structures met with in certain Monocotyledonous plants, and described 
by Miss Robertson (10) in the case of the genera Tulipa and Erythro - 
nium. These ‘ stalked bulblets ’ or f droppers ’ are the precise analogues of 
the Phylloglossum tuber, and serve the same fundamental purpose, namely, 
the exploration, so to speak, of the soil; the analogy is further enhanced 
by the fact that more than one tuber may be formed during a season 1 —so 
that the tuber of Phylloglossum is essentially a means of vegetative re¬ 
production ; this consideration seems to lessen very considerably the 
fundamental phylogenetic importance which has been assigned to it. 
Like the tubers we have described above, the stalks of the droppers 
are tubular, and bear small bulblets at the tips. Miss Robertson points out 
that the dropper is actually a continuation of the base of the leaf; and we 
may therefore suppose that the Phylloglossum tuber is a continuation of the 
suppressed leaf which we have postulated above ; this contention is sup- 
1 See supra , p. 338, and Thomas ( 13 ). 
