APEX OF THE ROOT OF OSMUNDiA AND TODEA. 87 
against an approximately uniform structure of the individual 
root throughout its development. It is impossible to decide 
the question with absolute certainty, but some collateral 
evidence may be gathered from the study of the origin and 
early development of lateral roots; if there be uniformity in 
the first divisions of the rhizogenic cells, while there is irre- 
gularity of structure in the older roots, a transition of struc- 
ture must necessarily have occurred. With the object of 
ascertaining this point, the origin of lateral roots has been 
investigated, and certain well-marked examples of the arrange- 
ment of the early divisions of the rhizogenic cell are shown in 
figs. 19 to 22. As is the case in other Ferns, the lateral root 
of Osmunda originates from a single initial cell, which 
belongs to the endodermis, and is situated opposite one of the 
groups of Xylem. Fig. 19 shows how this rhizogenic cell has 
divided by walls inclined to one another, so that from the very 
first there is a pyramidal apical cell in this young root. Fig. 
20 shows a similar pyramidal cell, but in this case the young 
root was further advanced, and its apex had penetrated to the 
outer limit of the cortex of the main root. On the other hand, 
in the young lateral roots represented in figs. 21 and 22 there 
are no pyramidal cells to be seen, the young roots in these 
cases appear to show an arrangement of initial cells (x, x) 
similar to that in figs. 17 and 18. Thus, in the very first 
stages of development differences of meristematic arrangement 
may appear, which are quite as great as those between the 
most extreme types of structure described above for the apex of 
the more mature root. This observation, it is true, affords little 
more than negative evidence ; we may conclude from it that 
since as great varieties of meristematic structure are to be 
found in the roots at their first stages of development as those 
seen in more mature roots, it is therefore unnecessary to 
assume a transition of meristematic structure from one type 
to another in the individual root, in order to explain the 
differences of structure in the mature root. We have, how- 
ever, no evidence which precludes the idea that such a transi- 
tion may actually take place ; all the evidence being taken 
