322 Arber . — The Leaf Structure of the Iridaceae, 
Babiana cuneifolia , Baker. The leaf-limb in this genus gives, at first 
glance, the impression of being folded in a fan-like fashion, but this 
impression is shown to be incorrect when a series of sections is cut through 
the junction of sheath and limb. Figs. 50 B-F represent such a series for 
the case of leaf /. 3 in Fig. 50 A. Fig. 50 B shows the sheath region, which 
is quite normal. In Fig. 50 C the limb is beginning to put in an appearance, 
but it is modified from the ensiform type by deep grooves or invaginations 
(g.), which occupy alternating positions to right and left. As we leave the 
sheath region, the grooves become more conspicuously developed (Figs. 50 
D and e) until in Fig. 50 F the mature form is reached. These sections 
show plainly that — as Lindman 1 has already pointed out — the terms 
‘ feuilles plissees 5 2 and 1 foglie piegate ’, 3 which have been suggested for these 
leaves, are inadmissible. We are not dealing with a case of ‘ folding ’, but of 
invagination, which is an entirely different thing. The difference is, in fact, 
as fundamental as that recognized by the geologist between the case of the 
production of certain mountain ranges by the actual wrinkling of the earth’s 
crust, and the case of the carving of a system of hills and valleys, by the 
action of rain and rivers, out of an originally flat surface. For comparison 
I have included Fig. 54 — a section of the basal region of a leaf of Phoenix 
dactylifera , L. (Palmae) — since this leaf has been described as offering a 
genuine instance of folding. It will be seen from this figure that, if the leaf 
were flattened out, all the bundles would be normally orientated, with xylem 
upwards and phloem downwards. And I have also found the same thing 
to be true of the c plicated ’ leaves of Curculigo sp. (Amaryllidaceae, Fig. 55) 
and Carludovica Plumerii , Kunth. (Cyclanthaceae). But, on the other 
hand, if the leaf of Babiana be flattened, the bundles will still be orientated 
in two opposite ways, as in so many other phyllodes. I propose in later 
papers to consider the nature of the ‘ folded ’ leaves of the Palms, 
Cyclanthaceae, Curculigo , & c., and to discuss the relation of these leaves 
to those of the Irids. 
Tigridia (Iridoideae, Figs. 52 A-D) may be taken to represent a 
further stage in the evolution of the foliated leaf. The transition region 
between sheath and limb shows that pseudo-plication arises exactly as in 
Babiana , through lateral invaginations. The form is, however, rendered 
a little more complicated by the development of a slight keel, k ., from the 
leaf surface in the region outside the phloem of each of the main bundles ; 
this keel may itself contain a bundle. Similar keels are also occasionally 
developed in non- foliated leaves. They may, for instance, occur in connexion 
with the midrib of an ordinary ensiform leaf, such as that of Tritonia , 4 or 
with the angles of a tetragonal leaf, such as that of Hermodactylus (Fig. 29 D, 
1 Lindman, C. A. M. ( 1899 ). 2 Chodat, R., and Balicka-Iwanowska, G. ( 1892 ). 
3 Ross, H. ( 1892 - 8 ). 
4 Arber, A. ( 1918 ), p. 486, and Figs. 15 B and c, p. 483. 
