certain Monocotyledonous Leaves. 339 
the median bundle comes to occupy the base of a channel running the 
length of the blade, while the main laterals, l. x and l'. v occupy the ridges 
which on either side bound the median channel, and the laterals, /. 2 and /'. 2 , 
/. 3 and /'. 3 , are located in the ridges separating two pairs of channels which 
run on either side of the median channel and parallel to it. In Figs. 13 A, 
C, and D I have labelled the vascular bundles in f. x as well as in /. 3 , in order 
to demonstrate that development proceeds on the same lines in both cases. 
In Figs. 13 c and D the median bundle of f. x (m.b'.) is seen, as in f. 3 , at the 
base of the groove between the two main laterals, b. x and b'. v while b. v b . 2 , 
b. 3 , b'. x , b'. 2 , b'. 3 , occupy ridges separated by invaginations. That b. x is not 
placed at the actual summit of a ridge, is an accident due to the pressure of 
the more external leaves which have been omitted in the drawing. It will 
be noted in all three leaves that the deepening of the invaginations is 
associated with a thinning of the leaf-substance ; this can be seen with great 
distinctness on comparing, for instance, the segment of the leaf f. x between 
the bundles b'.% and b '. 2 , in Figs. 13 A and D. The two bundles remain 
practically as fixed points, but whereas in Fig. 13 A they are connected by 
a simple segment of relatively thick leaf-tissue, in Fig. 13 D they have 
between them a deep fold, whose substance is attenuated to about one-third 
the thickness of the corresponding uninvaginated region in Fig. 13 A. 
In Figs. 13 E and F we witness the changes which the blade undergoes 
in passing towards the apex. In the case of /. 3 , all the ventral invagina- 
tions, except the median one, have died out, while the dorsal grooves are 
now reduced to inconspicuous indentations. The apex is slightly hooded, 
so that a cavity (cav.) occurs in the transverse section at this level. The 
apex thus corresponds to that of the plicate first leaf of the Palm, 
Pritchardia filifera , Lind., which I have described elsewhere. 1 In Fig. 10 F 
the apex is solid and all the invaginations have disappeared, while the 
bundles, which were oblique in Fig. 13 E, are now running together, meeting 
one another almost horizontally. 
Smilax. 
In a previous paper 2 I have considered the general morphology of the 
Smilax leaf, without, however, discussing the origin of the ‘ blade ’, the 
question with which we are here concerned. Serial sections through the shoot 
apex of S', herbacea , L., show that at a very early stage — in fact even before 
advanced stage, and tendrils, t. and t’ . ( x 14) ; Figs. 14C-E, series of sections through developing 
‘ lamina ’ of another leaf ( x 23) ; the points marked with a cross in Fig. 14 C will eventually become 
the margins of the ‘ lamina ’ ; Fig. 14 F, section of another leaf ( x 14) to show developing lamina, 
and tendrils, t. and t'. Figs. 1 5 A and B, Smilax laurifolia , L. , two sections of one young leaf ( x 47) ; 
Fig. 15 a passes through leaf at level just above sheath and shows wings of sheath, j. and s'., tendrils t. 
and t'. , and petiole, pet. , penetrated by ventral invagination, i. ; Fig. 1 5 B, same leaf at higher level, just 
below tips of tendrils ; lamina. Figs. 16 a and B, Rhipogonum album , R.Br., two sections from 
transverse series through apical bud ( x 23) ; ax., axis; the petiole, pet., of youngest leaf is just 
detached in Fig. 16 a, and the same leaf is represented by the lamina, in Fig. 16 b. 
1 Arber, A. (1922 1 ), Fig. 5 e, p. 256. 
2 Arber, A. (1920 2 ). 
