DR. J. D. HOOKER ON WELWITSCHIA. 35 
more of the secondary embryo-sacs is found to have elongated downwards into the cavity 
in the neck of the endosperm, and to have commenced to develope a suspensor at its base 
(Plates IX. fig. 28; X. figs. 20,21). In almost every such case, when I have removed the 
secondary sac entire from the cone (which I have done many times), I find a pollen-tube 
adherent to its apex (Plate X. figs. 15-19). In every such case, also, I have found pollen- 
grains on the apex of the nucleus, and their tubes penetrating the tissue of the cone just 
within its margin, and passing downwards near the periphery. I have never succeeded in 
removing a pollen-tube entire from the apex of the secondary sac to the pollen-grain, 
but I have on several occasions so nearly done so, as to have seen its continuity before 
rupturing it ; and in two instances (sketched before the dissection was completed, Plate X. 
figs. 15, 16) I failed at one point only, leaving no possible room for doubt. The ferti- 
lized secondary sacs, in every case examined, had advanced from one-half to two-thirds 
up the cone; and their apices occupied the termination of the canal, which was always 
close to the periphery of the cone, in the direction of the descending pollen-tubes. 
The number of secondary embryo-sacs which are thus impregnated varies. Usually, 
I think, only one is so; but there are often two or three (Plate IX. figs. 26-29 & 32), 
and I have several times found four, six, and even eight (Plate X. fig. 20). I have, how- 
ever, never found more than one embryo in the seed, nor seen a secondary sac branch- 
ing downwards, as might be expected to occur in the one figured at Plate X. fig. 14. 
After the contact of the pollen-tube, great changes commence in the bulbous base of 
the secondary embryo-sac, which elongates, and its contents collect into an obovoid mass 
the germinal vesicle. 
After elongating, the bulbous base of the secondary embryo-sac first presents a con- 
striction at the neck, and its basal contents, now bounded by a cellulose wall, are ob- 
scurely lobed, first on the summit and sides, and later at the apex (Plate X. fig. 22). Soon 
the constriction becomes more marked, separating the bulb, which is now a rounded 
cone with the broad end uppermost, from the tubular portion above it, whilst the lobing 
has so far advanced that the upper part of the contents of the terminal cell consists of 
eight or ten separate masses (fig. 23); these shortly acquire proper cell-walls, and be- 
come the first series of cells of the suspensor (fig. 24). The terminal cell now again 
divides as before, longitudinally and transversely, and adds a second series of cells to the 
suspensor (fig. 19). 
The cells of the suspensor elongate with rapidity, and are developed in great numbers ; 
and as they go on forming and elongating faster than the secondary sac elongates, the 
base of the latter is removed further and further from the terminal cell. At the same 
time that the base of the sac elongates, it becomes excessively attenuated, and finally 
forms a slender tube sheathed by the upper tubular cells of the suspensor (fig. 27). 
I have never had any difficulty in dissecting out the end of the secondary sac from the 
top of the suspensor (fig. 27); indeed, I have frequently pulled it out perfectly entire to 
the tip: but I have found much difficulty in tracing the early development of the sus- 
pensor from the terminal cell at the base of the secondary sac, and especially in proving 
the early retirement of the tubular prolongation of the sac after its bulbous base has 
been cut off by that terminal cell (as shown in Plate X. fig. 24). 
F2 
