6 
which are much larger than the others, but there does not seem 
to be any regularity in regard to their position on the rhizome, 
only that there are often two or three of these larger ones situated 
closely together. We see farther that the vegetative and floral 
shoots at their base are surrounded by some scale-like, membrana- 
ceous leaves. By separating the “tubers” from the rhizome (Plate 
CXIII, Fig. 14) it is plainly seen to consist of a number of very 
short branches, some of which are terminated by an inflorescence 
or carry a few, (two or three,) normal leaves besides the scale-like 
ones. The roots are scarce, slender and proportionally weak. 
Now, in regard to the mentioned tuber-like bodies, these, the 
smaller ones, are almost conical, slightly flattened at their ventral 
face and shortly pointed, in contrast to the larger, which are dis- 
tinctly triangular in outline with obtuse edges and show a scar at 
their apex, which proves that they have carried something. Still 
there is nothing yet to explain their origin, before we have re- 
moved the scale-like leaves. Inside these we _ shall find, 
(Plate CXIII, Fig. 15) in their axils, several clusters of minute 
“tubers,” very much like those observed before, but these show 
at their summit a rudimentary toothed blade, and are nothing but 
leaves (Plate CXIII, Figs. 16-17) situated on a short axis. At the 
same time, during the removing of the scale-like leaves the base 
of the normal ones becomes free and shows nearly the same form 
as the larger tuber-like bodies, only with the exception that they 
are in connection with the long petiole and the large, finely 
divided blade. The large “tubers” represent then merely the 
base of now partly faded normal leaves. 
There is, however, a shorter way to find out the origin of these 
tuber-like organs, if we will contemplate the plant at an earlier 
stage. Numerous minute, one-leafed specimens are always occur- 
ring around the mature plants, and these, illustrated in Figs. 18-20, 
are undoubtedly developed from the above mentioned smaller 
tuber-like leaves, which, as will be shown later, commonly sup- 
port a very minute bud (Plate CXIII, Fig. 23). It was my first 
thought that they were developed from seeds and in a secondary 
stage, probably one year old, but it seems to be the only correct 
explanation to offer, that they were developed from the small 
tuber-like leaves. It is also to be pointed out that these drop easily 
