DR. J. D. HOOKER ON WELWITSCHIA. 23 
cannot trace any direct connexion between these three bundles of vessels in the rachis and 
the six stamens, but it is conceivable that they have some relation to one another. 
Ovule.—The ovule (figs. 11, 14) occupies the axis of the flower, and is connate, at its 
base, with the base of the staminal tube; it is conical, and its integument is narrowed 
upwards into a flexuose tubular styliform body, which terminates in a very broad 
expanded papillose stigma-like disc, depressed in the centre; the depression communi- 
cates with the tube of the styliform body, and thus downwards with the cavity enclosing 
the nucleus. Thereis no other integument to the ovule but this; it consists of elongated 
cellular tissue (fig. 15), and contains a conical, fleshy erect, nucleus (fig. 14), in which I 
find no trace of embryo-sac, nor other contents, except an opaque line in the centre 
(fig. 16). Two vascular bundles (fig. 14) proceed to the outer base of the nucleus (one on 
each side), where they abruptly terminate. 
Not only is there a total absence of an embryo-sac in this ovule, but the whole body 
turns brown and withers after flowering; these circumstances, and a reference to my 
description of the ovule and development of the embryo in the female flower, will show 
how far it may safely be assumed that this organ in the hermaphrodite flower, though 
furnished with an apparently perfect stigmatic summit, is in reality constantly function- 
less. And if (as I shall hereafter attempt to show when treating of the female flower) 
impregnation takes place by the direct application of the pollen to the nucleus of the 
ovule before its integument is produced into a tube, it is a further argument in favour 
of its sterility in the hermaphrodite flower, that I have never yet found a pollen-grain 
within or on the ovule. 
The hermaphrodite cones and their flowers, here described, accord singularly with the 
male cones of Hphedra* in many essential respects. In both the scales are quadrifariously 
arranged, identical in nervation, single-flowered, and formed of the same tissues and on 
the same plan. The outer leaflets of the perianth of Welwitschia are absent in Ephedra ; 
but the two inner are identical in both genera, in structure and appearance. The six 
stamens united by their filaments, with capitate anthers opening by short terminal slits, 
and the pollen, correspond in every respect, except that the stamens of Ephedra vary in 
number from two to eight, that its anthers are usually 2-celled, and the staminal 
column is solid. Lastly, the ovule of the hermaphrodite Welwitschia, with its tortuous 
styliform process and stigma-like apex, is the same in structure and appearance with the 
ovule of Ephedra, differing only in wanting the embryo-sac, and in the stigma-like disc 
of the latter being narrow-oblong, and not papillose. With Getwm the affinity through 
the male flower is less obvious, and as, in so far as I can see, it is only traceable through 
Ephedra, it would be out of place to discuss it here. 
The most analogous case known to me of an apparently highly perfected stigmatic 
organ being absolutely functionless, is that of the curious genus Cardiopteris, which 
abounds in the eastern provinces of India, and which I have examined in a living state. 
In this plant the ovary is unilocular and has two stigmata—one large, globular, and 
* In one species of Ephedra I have found tristichous female cones, the scales being connate at the base in threes, 
