36 DR. J. D. HOOKER ON WELWITSCHIA. 
Though well convinced that the contents of the bulbous base of the fertilized secondary 
embryo-sac thus gives rise to the suspensor by the Giyaton of its mass and the formation 
therefrom of long tubular cells that sheath the retiring portion above the bulbous base, 
I am by no means clear as to the exact process. It would, probably, be necessary to 
examine fresh specimens to prove whether the germinal vesicle may not burst through 
the base of the sac, and whether it may not divide longitudinally into four, as in the 
Conifere, and thus give rise to four rudimentary suspensors, of which, in Welwitschia, 
only one is developed. All I can affirm is, that I saw no evidence of either of these 
phenomena, and that the figures (22 to 24) are faithful (camera-lucida) representations 
of the structures, whose deficiencies in matters of detail are owing to the specimens 
having been preserved in alcohol. 
The number of cells thus developed upwards, as it were, from the terminal cell of the 
suspensor, varies at different parts of the suspensor and in different individuals ; the first 
series often consists of six, eight, or ten, and the subsequent series always of more, till, 
towards the end of the fully formed suspensor, they are thrown off (so to speak) in great 
numbers, and coalesce into a fleshy mass continuous with the apex of the embryo. The 
first formed (Plate X. figs. 18 & 24) are of nearly equal length, and elongate most of all, 
attaining sometimes nearly jth of an inch; the next series (fig. 19) are more irregularly 
developed. Finally, towards the lower end of the suspensor, as the power of production 
declines, they lengthen upwards less and less, and form an imbricating bundle, of which 
the outermost are free and patent, and sometimes recurved wholly or in their upper parts 
(fig. 25). After this last stage of funicular development, the terminal cell commences 
rapidly to enlarge and form an embryo. 
The suspensor appears to be very rapidly formed: it is developed wholly within 
the cavity at the apex of the endosperm, and in a loose tissue of endosperm-cells ; 
it is extremely tortuous, and sometimes attains the great length of three inches. The 
terminal cell appears to burrow amongst the endosperm-cells at the sides of the cavity ; 
sometimes it gets entangled amongst the coalescing cells of the rapidly-forming albumen, 
but I think, in all such cases, it never produces a mature embryo. 
In most cases it is easy, after carefully dissecting away the tissues of the neck of the 
endosperm, to free the cone of the nucleus, and in removing it to draw out the whole 
suspensor from the cavity of the endosperm (Plate IX. figs. 26, 28; Plate X. fig. 20). 
As the endosperm hardens and the embryo becomes developed, this can rarely be done ; 
for then the suspensor usually is retained within the endosperm, whilst the tubular pro- 
longation of the secondary sac is stretched (Plate IX. fig. 27) and may often (as observed 
above) be withdrawn, quite entire, from out of the apex of the suspensor. 
During these changes the endosperm has been rapidly acquiring bulk and consistency, 
becoming an obovoid, compressed, granular mass, with an obconic central cavity (Plate 
IX. fig. 32), whilst its fleshy annular neck detaches itself from the cone of the nucleus 
above. The membranous remains of the embryo-sac may often be found on the surface 
of the nearly mature endosperm: arrested secondary sacs may up to a very late period 
be found adhering to the margins of its fleshy annular neck; and numerous unimpreg- 
nated secondary sacs are always present within the canals of the cone of the nucleus. 
