34 2 Arber . — On the Nature of the ‘ Blade ' in 
sinuses between the main lateral bundles, l. x and /. 2 , /. 2 and /. 3 , l'. x and /'. 2 , 
t'. 2 and /'. 3 . In addition there are, as in Veratrum , dorsal sinuses opposite 
and /'. 2 . The result is that the final arrangement of the main 
bundles in relation to the ridges and grooves exactly follows the scheme 
outlined in the case of Veratrum . 
It is interesting to notice that in Curculigo recurvata the first sign 
that invagination is about to take place is given by a local enlargement of 
the epidermal cells. This hypertrophy begins before the leaf shows any 
change of form (Fig. 17 C), and in later stages the enlarged epidermal cells 
can still be recognized at the base of the grooves (Fig. 17 e). 
Liliiflorae — Dioscoreaceae. 
Tamus . 
The leaf of Tamus communis , L., is peculiar in the complete absence 
of any basal region of a sheathing character. Fig. 18 A shows the petiole 
in the act of becoming detached from the axis (ax.) ; in Fig. 18 B it is free, 
while in Fig. 18 c the basal lobes (c. and c .) of the cordate 4 lamina ’ are 
seen on either side. Fig. 18 D marks the junction of stalk and blade, while 
Figs. 18 E and F show how this ‘ blade’ is produced by marginal winging 
of the petiole and separation of the main bundles. The ribbing of the lower 
surface of the leaf is brought about by extremely slight dorsal invaginations 
between the bundles. 
Dioscorea. 
The ontogeny of the leaf in the case of Dioscorea sativa , L., and 
D. divaricata , Blanco, is similar to that of Tamus communis , but it will be 
recognized from Fig. 19 that invagination plays a much more conspicuous 
part than in the latter genus — giving rise to the deep grooving of the 
under surface of the young leaf ; in this sketch, the section with almost 
entire outline (marked with a cross) is that of the leaf which is nearest to 
its point of attachment, while the other leaves, which are cut at higher 
levels, show deeper grooving. 
A noticeable feature of the leaves of the Yams is the differentiated 
apical region (ap. in Fig. 20), which in the young leaf looks almost like 
a separate lamina. Sir David Prain, F.R.S., was so kind as to draw my 
attention some years ago to this curious structure ; in order to try and 
understand it, I have studied serial sections through leaf-buds, and I have 
come to the conclusion that the distal region of the Dioscorea leaf has no 
special morphological significance. It cannot, I think, be treate<d as a case 
apart ; it belongs to that class of leaf-apex which Raciborski 1 named 
‘ forerunner tips ’, and which he showed to be associated with the climbing 
habit, not only in the Dioscoreaceae and in Smilax , but also in various 
Dicotyledons. I find that the leaf-tips of the Yams do not differ from the 
1 Raciborski, M. (1900). 
