86 
F. 0. BOWER. 
very early stages (figs. 11, 16, 17, tr .) ; but the outer limit of 
the procambial cylinder does not appear to correspond to any 
definitely recurring wall in the young segments, as is the case 
in the roots investigated by Naegeli. The same may be said 
with respect to the limits of the cortex and epidermis ; the 
latter tissue, which can readily be recognised in the older part 
of the root, and can be traced as entering below the layers of 
the root-cap, loses its identity at a considerable distance from 
the actual organic axis of the meristem. This fact is illus- 
trated in the figs. 11 and 18; in neither of these examples can 
a clear distinction be drawn between those cells which will 
develop as tissue of the root-cap and those which will form 
epidermis, or cortex. 
It remains to suggest which of the several types of arrange- 
ment seen in longitudinal section correspond to the several 
types above described as being seen in transverse sections, but 
it must be understood that what follows is only put forward in 
a tentative manner. The observation of thick transverse sec- 
tions appears to show that where there is a three-sided apical 
cell, it is of pyramidal form; whether the same is the case with 
four-sided apical cells is uncertain. On the other hand, where 
the number of initial cells is three (or possibly in some cases 
four), transverse sections lead to the conclusion that they are of 
the truncated pyramidal form. Thus it is probable that where 
a pyramidal cell is seen in longitudinal section, it is a cell of a 
three-sided or four-sided pyramidal form. Probably, also, the 
pyramidal cells marked (o) in figs. 13 and 14 correspond to 
such cells as those marked also (o) in the transverse sections 
fig. 5, b, and fig. 4. Further, on the grounds above mentioned, 
it seems probable that arrangements such as those in figs. 17 
and 18 correspond to such meristems as those represented in 
transverse section in figs. 4 and 7. 
A question of considerable importance is whether in the 
individual root a transition may occur at any time from one 
of these types of meristematic structure to another. That a 
number of roots of a given species or individual differ in 
meristematic structure from one another is no argument 
