MORPHOLOGY OF PHYLLOGLOSSUM DRUMMONDII. 
6 71 
becomes elongated below the insertion of the lowest of them (figs. 22-25). As to the 
arrangement of the leaves on the cone I have nothing to add to tire description of 
previous writers. It will be seen from figs. 22-25 that the first two leaves of the 
cone correspond roughly in position to the second pair of leaves in fig. 13. 
Since in sporangium-bearing plants the apex of the main axis continues its growth, 
and gives rise to the sporangium-bearing cone, it is clear that there must in this case 
be a different origin for the tuber than that described for the purely vegetative plant. 
Up to a stage of development, such as is shown in fig. 23, no trace of a new tuber has 
been observed; but subsequently a depression is seen at a point near the base of the 
pedicel, which is surrounded by a circular weal (fig. 24); this is the young tuber, 
which is again shown in a more mature condition in figs. 25, 26. It would thus 
appear that in those plants in which the main axis develops as a sporangium-bearing 
cone, the new tuber is of adventitious origin, and exogenous. It does not bear any 
constant relation to the leaves, though, as has been observed in mature plants, a leaf 
of smaller size than the rest is occasionally inserted immediately above the attach¬ 
ment of the new tuber (fig. 27). A supernumerary leaf of this sort has been 
observed in four out of eleven specimens of mature sporangium-bearing plants. In 
plants of this type the roots originate in the manner above described. 
The chief points in the development, as seen on germination of the tuber, having 
been described, as far as their external characters are concerned, the description of 
the development of the tuber and of the root, as illustrated by sections, may now be 
proceeded with. If a median longitudinal section be cut from a plant, such as that 
represented in fig. 13, it will present a,n outline as shown in fig. 15. At the point 
marked («) there is a slight depression, which indicates the first origin of the new 
tuber. Seen under a higher power the arrangement of the cells appears as in fig. 28. 
Here it is evident that growth and cell-divisions, especially periclinal divisions, are less 
frequent at the point (a) than in the tissues immediately surrounding that point: the 
result is that the point marked (a) becomes depressed, while the surrounding tissues 
gradually overarch it (fig. 29), leaving only a narrow channel open above the apex. 
Finally the channel which is seen still open in fig. 29 is closed up, and the apex of the 
new tuber is thus completely covered in by the more rapidly growing tissue which 
surrounds it (fig. 30). It is to be observed that the apex thus covered in is in the 
first instance of very small area, and in the case of fig. 29, it appears to be only a 
single cell in width : whether this is always the case I am not in a position to state. 
Subsequently divisions appear in the cells at the base of the depression (fig. 30), 
which are repeated, and as the tuber increases in bulk the apical cone also increases 
in width, until finally in the mature tuber it is a massive structure of convex form, 
such as is seen in fig. 6. Even before the apex of the new tuber is thus covered in 
by the increasing bulk of the tissues surrounding it, growth is localised in a special 
way in the more deeply seated tissues : this growth results in the whole body of the 
tuber being made to project beyond the general surface of the axis, while, since the 
4 R 2 
