Bower . — Studies in the Phytogeny of the Filicales. 463 
the cursory observations now made indicate that in all these Ferns the 
sorus is phyletically marginal in origin, but that the more advanced types 
may show a shifting of the sorus towards the lower surface. The biological 
significance of this is obvious, for protection of the young sorus is thereby 
secured, and this consideration therefore supports the view here advanced. 
Three lines of argument should be used in arriving at such a conclusion 
as that above stated. First, the comparative use of detailed observations 
which form the immediate foundation upon which it may be based, by pro- 
viding intermediate steps from the originally marginal state to the superficial. 
It is admitted that the series here given is an incomplete one, and that the 
whole question will have to be gone over again, and tested by numerous 
examples. The second line of argument is that based on the comparisons 
and conclusions of the systematists, chiefly relating to external characters, 
upon which the current grouping has been based. The probability is that 
such conclusions are correct. The third is the argument from physiological 
probability. In the present case, these three lines of evidence coincide in 
support of the conclusion as stated. Accordingly, the proper place of 
Nephrolepis (and probably also of Oleandra ) in a phyletic system should be 
in near relation to Saccoloma and Davallia , notwithstanding the position of 
the sorus on the lower surface of the leaf. But their place in the system 
will be more distal than those genera where the sorus Is more nearly 
marginal. 
The similarity in position of the sorus, and in form of the indusium 
between Nephrolepis and the Aspidieae, is very striking. Commonly, there 
is this difference between Ferns of the Davalloid and those of the Aspidioid 
affinity : that where the sorus is ‘ phyletically ’ marginal the vein terminates 
in its receptacle, while, where it has been c phyletically* superficial, the sorus 
is seated on a vein which still continues its course. But this does not 
appear to be a perfectly dependable rule. Any exceptions to it would 
still further accentuate the striking parallelism which these Ferns show. 
But, nevertheless, it must be concluded that the two groups have been 
distinct in their descent. The one is essentially marginal, the other essentially 
superficial, in the position of the sori. They illustrate in a remarkable 
degree the convergent development of two distinct phyletic lines. 
Loxsoma. 
This Fern, as interesting as it is enigmatical in its relationships, has 
again come into prominence through a recent paper by Professor von Goebel 
(‘Flora,* 1912, Band cv, p. 33, &c.). He has there not only figured the 
sorus afresh, but also described the prothallus, hitherto unknown. After 
a general survey of the known characters of Loxsoma, he concludes (p. 45) 
that there is no ground for placing Gleichenia in near relation with Loxsoma , 
and that the same holds for the Hymenophyllaceae. He considers that it 
