sash Ss a 
ea re 
12 DR. J. D. HOOKER ON WELWITSCHIA. 
are so abundant in the former, that in a full-grown specimen the tissue cannot be. cut 
smoothly with the sharpest knife, and a saw works through it with considerable diffi- 
culty, leaving very ragged surfaces. The root especially, owing to the masses of liber 
in the wood-wedges, and their waving courses, can scarcely be either cut or sawn, but 
may be riven if sufficient force be applied. ; 
The spicular cells above alluded to are the most singular feature in the tissues of the 
plant, and, whether in their size, form, or surface, are quite unlike anything I have seen 
elsewhere. They are often }th of an inch long, extremely rigid, fusiform or acicular, 
acute (even pungent), or more rarely obtuse, at both ends, straight or branched, often more 
or less angled, curved, or hooked, and more or less thickly covered with minute crystals 
(Plate XII. figs. 5-8). They are found entangled in the parenchyma of all parts of the 
plant; but their walls become at a very early period free from the surrounding cell- 
walls. Their average diameter is that of four or five of the parenchyma-cells. The oldest 
ones are yellow. In atransverse section their walls appear concentrically laminated, the 
central cavity becoming all but obliterated with age: in the younger tissues they are 
smaller, paler, have thinner walls, and the crystals also are smaller and more remote 
(Plate XTT. fig. 8); and, as I shall hereafter show, in the more membranous tissues of the 
plant (hermaphrodite perianth, &c.) they remain permanently thin-walled, are wholly 
void of crystals, and resemble the corresponding acicular cells in Gnetum, Balanophore, 
and other plants (Plate VI. fig.9; Plate VII. fig.8; Plate VIII. figs. 26, 27). 
The crystals are flat, and applied to the surface of the cell by their broad faces ; they 
are either rhomboidal or short six-sided prisms, always thin, of various sizes, zo5pth of an 
informs me that they are neither phosphates nor oxalates, but that after incineration they 
become soluble in acids, which indicates silica; but there is no test for silica sufficiently 
lar and often branched forms, surfaces roughened with crystals, and irregular arrange- 
ment suggest that this is the purpose they are intended to serve. In the leaf they brace 
together the several layers of parenchyma; and in the perianth, under a’ modified form, 
they give strength to the cellular tissue, acting in this respect like vascular bundles. 
Under this view they are analogous to the siliceous spicula of sponges, though formed 
on a totally different plan. 
The gum which exudes from the trunk, peduncles, and various parts of the inflores- 
