Kashycip. — Notes on Equisetum debile , Roxb. 441 
ducing small islands of parenchymatous tissue of one to six or seven cells. 
These isolated rings may be close to each other so that one cell is actually 
common to the two rings (Fig. 1, x ), or the endodermis round the parenchy- 
matous cells may be quite independent (Fig. 2, a). In Fig. 1 the two rings 
a and b are connected by one cell having suberized bands on four of its 
walls, two belonging to one ring and two to the other. In other cases two 
of the bands may actually occur on one wall, the four bands belonging to 
two rings occurring on three walls as seen in Fig. 3 at a . 
It is easy to understand the formation of these islands if we suppose 
that the second type of the endodermis is gradually passing into the first by 
the fusion of the two layers at various points. But that is hardly likely in 
view of what has been said and what is to follow. It is more difficult to 
understand how these islands could be formed if the endodermis of the first 
type at the node is giving rise to the endodermis of the second type in the 
internode. 
In the aerial fertile region below the cone, and in the base of the cone 
itself, the endodermis is of the first type everywhere. It may be due to the 
fact that the internodes here are very short. Since the fertile part of the 
shoot, however, is known to be more conservative and retains its ancestral 
structure for a longer time than the vegetative shoot, it is likely that this 
arrangement indicates the ancestral features of the stem, especially when 
taken along with the structure of the rhizome. In the cone itself, even near 
its base, the characteristic bands on the radial walls of the endodermis are 
lost, but the layer can be distinguished on account of the large size and 
regular arrangement of the cells. As we ascend higher, however, even these 
