THE SPORE-PRODUCING MEMBERS 699 
seen in Ferns is similar in kind to that of strobiloid types, but modified 
in accordance with the great amplification of the sporophyll, with its 
continued apical growth and often profuse branching: this was accom- 
panied by increase in number of the sporangiophores (sori), fission being 
one prominent source of that increase, and also by a tendency for the 
sorus to diffuse itself as scattered sporangia over the enlarged surface, 
producing thus the non-soral state as a secondary condition: moreover, 
the position of the sori shows frequent tendency towards the lower leaf- 
surface. From this point of view the Fern-type does not stand apart 
from the rest in the essentials of its morphology, but only in the fact- 
that it has proceeded to a larger-leaved state, and that this has brought 
with it secondary changes of the spore-producing members. 
There is a considerable bulk of evidence to show that, apart from 
fissions, the sorus or sporangiophore has also been capable of extension 
in the course of descent: this is to be found in certain points of structure 
which have not been satisfactorily accounted for on any other footing, 
It has been noted that in the fossil Equisetales the number of sporangia 
on each sporangiophore is commonly four (p. 425), but that modern 
Equiseta have usually more. In the Psilotaceae and Sphenophylleae the 
number may be from one to six, the lowest number being probably in 
some cases due to reduction: thus fluctuating, numbers are a common 
feature in the simpler types. In the Ophioglossaceae the fluctuations are 
within wider limits, and the larger numbers of sporangia are associated 
with an apical growth of the sporangiophore, which is either of very short 
duration or entirely absent in other cases. The result is in Op/hioglossum 
the elongated spike, with its lateral rows of sporangia partitioned some- 
times imperfectly from one another (Fig. 361 L). The structure bespeaks 
a progressive condition in which septation has played a leading part 
(p. 404). In Botrychium, profuse branching parallel to that of the 
sterile leaf, occurred, and it is very closely related with septation of the 
individually projecting sporangia (p. 454); lastly, in Aelminthostachys 
the rows of sporangia of Ophzoglossum are replaced by dense ranks of 
sporangiophores (pp. 455, 485), and their origin is believed to have been 
virtually a repetition of that process of septation and upgrowth above 
recognised in the origin of the sporangiophore from a simple sporangium 
(Fig. 361 N). All these amplifications of the sporangiophore are consistent 
with physiological probability, as shown in Chapter XXXI. 
In the Ferns also similar extension of the sporangiophore (or sorus) 
is seen, but it has taken a different form in accordance with the expansion 
of the Jeaf-surface to which it remains attached. It is exemplified in the 
simplest form in the Marattiaceae, in which the structural condition of 
Danaea seems plainly to be the result of elongation and progressive septation 
of a sorus of the same type as that of Maratéa (Fig. 278 c, £); the 
partial septations are themselves specially convincing evidence of how the 
highly septate state has been acquired (p. 518). The progression has 
