38 Arber.—On the ‘ Squamulae Lntravaginales ' of the Helobieae . 
tion needs restatement. If we interpret Fig. 5 A, p. 37, on Saunders’s 
theory—which I adopt here because it appears to me to clarify our ideas of 
the spermophyte shoot—we should say that the surface of the ‘ axis 5 on the 
left-hand side (inside/.j), which is seen giving rise to squamules, is, in reality, 
the ‘ leaf-skin ’ belonging to leaf 2 ; this leaf-skin is regarded as extend¬ 
ing downwards from the exsertion of leaf 2, so that it clothes the axis 
throughout the abbreviated internode separating leaf 1 and leaf 2. I should; 
thus, describe the squamules of Triglochin , not as belonging to the leaf 
beneath them, but as appendages of the dorsal (lower) surface of the leaf 
next above them ; they are developed near the extreme base of its down¬ 
ward extension—that is to say, in the region in which it has no free exis¬ 
tence, but forms the leaf-skin for the internode. The only case, in which 
an explanation, more or less of this type, has been already offered, is that 
of Cymodocea aequorea , Kon. (Potamogetonaceae), described by Bornet ( 2 ) 
under its old name PJmcagrostis major , Cavol. Bornet’s work predates the 
appearance of the Leaf-skin Theory by many years, so he naturally did not 
express his conclusions in these terms, but in describing the squamules he 
definitely related them to the leaf above their place of origin, and spoke of 
them as left behind by the elongation of the internode. Irmisch ( 14 ) dis¬ 
cussed Bornet’s view, and attempted to refute it. His chief argument was 
that squamules occur in certain positions (e. g. above the highest leaf of an 
inflorescence axis) in which there is no foliage leaf above them to account 
for their presence. But it seems to me that the squamules in such cases may 
possibly be related to bracts or sepals occurring at some distance above them, 
for Saunders has brought forward evidence that modified leaves of this type 
may provide a downward prolongation of leaf-skin as effectively as if they 
were foliage leaves. Bornet’s conclusions were based on a close and 
delicate examination of the squamules of Cymodocea aequorea , Kon., as 
solid objects. In the case of another species of the same genus (C. isoeti- 
folia , Asch.) I have shown (p. 36), by the radically different method of 
serial sections, that it is possible to convince oneself that the squamules do 
not belong to the leaf in whose axil they are found, but arise from’the 
apparent axis above this leaf (Fig. 4 D, p. 36); this observation is thus con¬ 
firmatory of Bornet’s view. 
In the other Potamogetonaceae which I have examined ( Potamogeton 
natans and P. sp.) the squamules arise from a zone of tissue forming the 
boundary between the sheathing leaf-base and the * axis ’ which it encloses. 
Whether this boundary tissue should be treated as belonging to this leaf or 
to the axis is, in this case, rather a subtle question. In the broad-leaved, 
submerged Potamogeton represented in Figs. 1 and 2, p. 33, there are 
clear indications that the squamules, on the side remote from the opening of 
the leaf-sheath, are still fused with the growing apex, after the leaf becomes 
detached from it. Examples of this attachment are seen in the squamules 
