Se Pie eee ee ee eee "25. ai 
Z Pere Nl i dl : pa 
ia 
Senn ee eee ee DEL RL gy an eee ee i en ee a Ces. ae Se el ee ee rs | 
e 2 ; Baie iy Colo ie efi w : 
DR. J. D. HOOKER ON WELWITSCHIA. 25 
cells, with minute points on their surfaces, but no formed crystals (fig. 27). The outer 
surface of the scale is provided with stomata; the inner (fig. 23) has none. 
The rachis of the cone (Plate VIII. fig. 1; Plate XIII. figs. 1 & 2) is fusiform, broadest 
below the middle, and almost terete, of a soft spongy texture, and covered with a very 
delicate epidermis. The arrangement of the vascular cords is curious: of these there 
are, I believe, normally twelve principal ones, in four sets of three each, corresponding to 
the positions of the four rows of scales, and various smaller supplementary ones; they 
all run just within the periphery, in wavy lines from the base to the summit of the 
rachis, anastomosing here and there by lateral branches. On removing the outer layers 
of cellular tissue (Plate XIII. figs. 1 & 2, which are diagrammatic sketches), three nearly 
parallel wavy bundles are seen beneath the scale, of which the middle one is supple- 
mentary, not communicating directly either with scales or flowers, but anastomosing 
here and there with the lateral ones. Below the insertion of the scale, each of the 
lateral nerves gives off externally one ascending branch, which enters the scale, and 
ramifies in it (fig. 1), as described under that organ. The two nerves of the perianth, 
on the other hand (fig. 2), are derived from the lateral nerves themselves, and not from 
branches of these. It follows, from this arrangement, that the bases of the nerves 
that ramify in the scale are further apart than the bases of those that ramify in the 
perianth. I have seen a scale with an additional nerve on one side, placed nearer 
the margin than the normal one of the same side, and this was supplied by a branch 
from the adjacent bundle of the adjoining triplet, which anastomosed with its neigh- 
bours. Other deviations no doubt occur. 
The smaller vascular cords which traverse the rachis anastomose together; they gene- 
rally run between the main ones, but are for the most part free ; more rarely they overlie 
the main cords, and sometimes they unite with them. 
The parenchyma of the rachis (fig. 3) is by far the most delicate in the whole plant, 
and consists of loosely packed, very thin-walled utricles (figs. 4 & 5), with pitted walls ; 
these pits are remarkably oblique (fig. 5), so as to present the appearance of an overlap- 
ping double boundary. 
The female cone of Welwitschia resembles that of Hphedra in very many particulars, 
and especially in the remarkable bilateral venation of its scales, in the abundance of 
stomata, in the liber-cells ramifying through its wings, and in the presence of spicular 
cells; furthermore, in Z#. alata and other species the scales are as beautifully mem- 
branous, and the two vascular bundles enclose a similar hyaline transparent area. 
Ephedra, as elsewhere stated, differs in all the scales being connate at the base, and in 
their becoming larger upwards to the terminal pair, which alone are floriferous. 
Female Flower, Perianth, and Pericarp (Plates VIL., VIII., IX., & X.). 
Perianth.—This, which ultimately attains a great size, as the pericarp of the seed, does 
not appear till after the formation of the nucleus of the ovule. Ihave traced its develop- 
ment in the immature flowers taken from the apical scales of the half-ripe cones. In all 
these the eight or ten uppermost scales (Plate IX. fig. 1) were extremely minute, and 
; E 
