354 Bower. — Studies in the Phytogeny of the Filicales. VIII . 
On this last point the fact is significant that in Loxsomopsis notabilis 
the stomium may lie either right or left of the stalk (Figs. 4, 6). If such 
a difference as this be found between the individual sporangia of a single 
plant, it suggests that there is in them just that sort of plasticity of develop¬ 
ment of the annulus which would be required to effect the transition from 
median to lateral dehiscence. It is probable that such a transition has 
actually occurred more than once in the evolution of the Ferns, and the 
plasticity above suggested would support this view. A very clear instance 
is to be found among the Superficiales in Lophosoria quadripinnata , Gmel. 1 
Here the annulus is of essentially the same type as in Loxsomopsis , though 
with a less highly specialized stomium. It appears probable that while 
Loxsoma and Loxsomopsis illustrate the result of transition from the median 
to a lateral dehiscence in the Marginales, Lophosoria, and in the more exact 
degree Metaxya , show it among the Superficiales. The former instance 
might be traced to such a source as the living Schizaeaceae still illustrate: 
the latter clearly owe their origin to some such source as the Gleicheniaceae. 
These considerations appear to me to strengthen the view previously 
expressed 2 that the Leptosporangiate Ferns progressed from very early 
times along two parallel lines, the one characterized by marginal, the other 
by a superficial position of their sori. As already shown elsewhere, 3 the 
distal position was probably general in the first instance for all Ferns. But 
as the leaf-blade expanded the marginal sorus slid to the lower surface, and 
it is probable that this has happened along many distinct phyletic lines, 
sometimes early, sometimes late in their evolution. In the Superficiales we 
may believe that it happened early : in the Marginales the marginal position 
was long retained ; but ultimately the transition occurred, as indeed it is 
foreshadowed in the Schizaeaceae, and was carried out more fully in 
the Pteroideae. 
The effect of the observations on Loxsomopsis , imperfect though they 
are, is to cast a doubt upon the validity of the median dehiscence of the 
sporangium of Loxsoma as a point of comparison with the Schizaeaceae. 
But nevertheless these two related genera will retain their interest as 
synthetic types of clearly primitive nature. They occupy in the marginal 
series of Ferns a position between the Schizaeaceae and the Dicksonioid- 
Davallioid-Pterid sequence: and this is comparable to that occupied by 
Lophosoria and Metaxya in the superficial series, between the Gleicheniaceae 
and the Cyatheoid-Dipterid-Dryopterid sequence. 
1 Ann. of Bot., vol. xxvi, PL XXXV, Figs. 17-20. 
3 The Ferns, Cambridge Press, vol. i, 1923, pp. 216-25. 
2 Ibid., vol. xxvii, p. 470. 
