272 Prof. H. Karsten on the Formation, 
species of (idogonium which will hereafter be described. This 
may explain why Mohl and Sanio did not observe it, and why I 
only succeeded a few times in meeting with stages of develop- 
ment corresponding with that shown in PI. V. fig. 8. 
The presence of a lamina of cork of some thickness retards, 
on the one hand, the flow of nutritive juices and their concen- 
tration by evaporation, and, on the other, restricts the action of 
the atmosphere on the cells assimilating it ; the newly generated 
cells then extend themselves more slowly, and even appear to 
remain in a half-evolved condition; at least, in the line of de- 
marcation between the cork and the unchanged cells of the 
tissue, some of the latter are occasionally noticed to contain 
delicately walled cells and vesicles, which possess the chemical 
constitution of cork, and present all the intermediate stages to 
that of the completely developed periderm, as is shown in Pl. V. 
fig. 2 x, in the case of a thickened porous cell of the medullary 
sheath, and, in fig. 5, in that of a vessel. 
The most recently developed peridermic cells behave like cellu- 
lose with a solution of iodine and chloride of zinc, but soon lose 
the property of becoming blue on the addition of iodine and 
corrosive reagents, which property they indeed do not possess in 
the first period of development. Probably cell-nuclei occur in 
all cork-cells during a certain stage of development ; and their 
duration seems to depend on the quality of the nutritive fluid 
absorbed by the cell-tissue, and on the chemical composition of 
the plasma within the cells, as well as on the more or less im- 
mediate access of atmospheric air. 
The first-formed cork-cells contiguous to the withered cell- 
layer mostly remain simple, whilst those subsequently produced 
contain a further generation of two or more new cells. In this 
manner such a cork-cell becomes occupied by a complete cell- 
tissue, the very delicate walls of which are in such close ap- 
position that they leave no intercellular spaces between them, 
and cannot be recognized as consisting of a double membrane. 
Some layers of these cork-cells in the immediate vicinity of 
the cut surface, soon after their evolution, acquire stratified 
thickenings of their secondary cells, which are penetrated by a 
few pore-canals. 
On resolving the tissue containing the cork-cells into its sepa- 
rate cells, by boiling it with nitric acid and chlorate of potash, 
and on subjecting it to the action of ammonia, the cellulose 
walls of the parenchyma become more or less greatly swollen, or 
are entirely dissolved. In the latter case, the solution of am- 
monia also acts upon the coherent groups of cork-cells derived 
from the dissolved cells, and causes them to swell up, to assume 
a spherical form, and to part asunder (figs. 14 & 15). In this 
