IO 
Birbal Sahni. 
III. Theoretical Considerations. 
a. The Primitive Branch. 
Although the branching of Ferns has long been a subject of 
detailed investigation, and nearly all the different forms of the 
vascular relations between stem and branch described above have 
been known since the publication of De Bary’s “ Comparative 
Anatomy,” it has not hitherto been suggested that these relations 
can be arranged in such a regular series as that shown in Fig. 1, 
and the interpretation of the series in terms of the phylogeny of 
the filicinean branch has consequently not been attempted. Gwynne- 
Vaughan was the first to suggest that the ontogeny of the vascular 
system of the main axis is often more or less imperfectly repeated 
in the development of the branch; and subsequent work, chiefly that 
of Professor Bower, has added considerably to the data on which 
that suggestion was originally founded. It still remains to be seen 
which of the two conditions is the more primitive for the branch— 
is it that in which this repetition is more perfect, in which the branch, 
starting life with a thin protostelic strand, gradually “ works its way 
up” to the morphological level of the main axis; or that in which 
it is from the very beginning at or near that level ? These two 
conditions are represented by the two ends of the series C—F; the 
ends are connected by means of stages morphologically intermediate 
between them; and the problem before us is to decide whether the 
series is a descending or an ascending series. 
It is evident that this problem is inseparably connected with 
the wider and much debated question of the primitiveness or 
otherwise of dichotomous as opposed to monopodial branching, and 
that the answer to the first will give the clue to the second. 
Before proceeding to our discussion, the conclusion arrived at 
may be stated at once, namely, that the forms of branching depicted 
in Fig. 1, C-F, constitute an ascending series. In our discussion 
we shall give particular attention to the branching of the higher 
Ferns (those with reticulate steles), as affording a more complete 
illustration of the parallelism in structure between stem and 
branch, and a greater variety in the behaviour of the branch vascular 
system. 
Naturally, the most direct way to attack the problem would be 
to enquire into the circumstances which lead to the production of 
the respective forms of branching. Let us conceive of the growing 
stem of a dictyostelic Fern which is about to branch, and ask 
