9 
be shown, do not belong, however, to the roots, but to subterra- 
nean stems. Plate CXIII, Fig. 25 shows then the complete rhizome 
and the leaves ofa mature specimen ; we see here that the plant has 
been developed from a large tuber, and that three stolons proceed 
from the leaf-bearing axis, all of which terminate in tubers.* The 
tuber has then been formed by the end of a stolon, a manner of 
propagation which is not uncommon, as for instance in Cyperus 
esculentus and others. The stolons are rather thin, with long in- 
ternodes, of which the leaves are scale-like and pointed. The 
direction of the stolons is almost horizontal, and they do not run 
very deeply under the surface of the ground. The tubers them- 
selves consist simply of but two internodes, the leaves of which 
are a little larger than those of the stolons, but show, however, 
the same form. In regard to the shape of the tubers, these are 
almost globular at a younger stage, but later they are of a nearly 
oblong shape with the apex acute, formed by a few pointed 
scale-like leaves. Now, when the tuber is going to germinate, 
the stolon dies away, and one or two relatively long internodes 
will be developed from the apex of the tuber, terminated by the 
flowering scape with a rosette of normalleaves. This is to be seen in 
Fig. 25, where the plant distinctly shows its offspring from a large 
tuber, and we see also one long vertical internode between the 
leafy rosette and the tuber itself. Ali the roots are confined to this 
internode, and it looks, certainly, as if it were a tuberiferous 
root. The stolons do not, however, always terminate in a 
tuber; ina case figured in Fig. 26, a small leafy rosette has been 
developed immediately at the end of a stolon, without being pre- 
ceded by a tuber, and in this same specimen the stolon had been 
developed in the axil of a scale-like leaf, situated on the under- 
ground stem, formed by the large tuber. The plant propagates 
then easily by tubers and, as mentioned above, all the specimens 
collected showed only this kind of offspring. There is, however, 
no doubt that it is able to propagate by seeds also, but the ger- 
mination has not yet been examined, and might probably show 
something interesting in regard to the primary root and the first 
beginning of the formation of stolons. 
[* There was a tuber represented in Mr. Holm’s drawing on the stolon shown to 
the left of the figure, but it was accidentally omitted in the engraving. —Ep. ] 
