466 Bower. — S hi dies in the Phytogeny of the Filicales. 
Fig. 27, I, and adding a sporangium facing the first, so as to show what is 
actually seen in Gl. flabellata or any of the Mertensia section. It will then 
be seen that the annulus in Fig. 27, 1, and in the right-hand sporangium 
of Fig. 27 bis , is in the same position relative to the receptacle (pi), and that 
is the proper basis for the comparison. 
But the point which is the most striking in the sporangium of Loxsoma , 
as compared with that of any other Gradate Fern, is the median dehiscence. 
The median dehiscence is, however, the rule in the Simplices, such as the 
Schizaeaceae and Gleicheniaceae. The question will then be whether or 
not it is probable that this character of Loxsoma is a survival from some 
simpler type of Ferns. Professor von Goebel’s opinion is that it is not 
a survival, but a secondary acquired condition, derived from some Cyathe- 
aceous source. Against this it may be urged that no such modification of 
the Cyatheaceous type is known elsewhere ; nor does it seem probable, for 
it would, in fact, be a reversal of such a progression as we have reason to 
believe has actually taken place, having been determined by the mechanical 
requirements of the gradate sorus. The case made out in No. II of these 
Studies (‘Ann. of Bot., 5 vol. xxvi, p. 31 1) for a probable progression from 
the Gleicheniaceous type, through Lophosoria to the Cyatheaceous type, 
involves a transition from a simple to a gradate sorus, together with a change 
from a median to a lateral dehiscence. Similarly, I suggest that in the 
Schizaeaceae, Loxsoma , and the Dicksonieae there has been a progression 
from simple (marginal) to a gradate sorus, and from a median to a lateral 
dehiscence. But, in the case of Loxso 7 na the two changes have not syn- 
chronized. While it has acquired the gradate sorus, it has retained the 
median dehiscence characteristic of the living Schizaeaceae. In these latter 
Ferns, the stomium is not of an elaborate type, as shown by Prantl’s figures 
(‘ Schizaeaceen,’ PI. V, Fig. 81 ; PI. VI, Fig. 98 ; PI. VII, Fig. 104 ; PI. VIII, 
Fig. 141). In all of the four genera the split appears as little more than an 
interruption of the continuity of the indurated ring, without a definitely 
constructed stomium such as is seen in the Cyatheaceae ; and this is what 
is the state in Loxsoma. Further, the form of the sporangium of Loxsoma 
resembles markedly that of Lygodium , though the lopsidedness is not so 
extreme (cf. Fig. 26, ^r, with Prantl’s figure of Lygodium , 1 . c., PI. VI, Fig. 97). 
But the annulus in Loxsoma is a wider ring, as, in point of fact, it is in 
Aneimia also, and the ‘peripheral’ face larger as a consequence. On the 
other hand, the spore output is only sixty-four, as against the larger numbers 
of the Schizaeaceae, though the form of the spores is similar. The general 
conclusion is, then, that there is a substantial similarity of the sporangia of 
Loxsoma to those of the Schizaeaceae, though not to any one genus of that 
family. If this be a true comparison, then we may look on Loxsoma as 
a Fern having, like the Schizaeaceae, a marginal origin of the spore-bearing 
members, and as having progressed to the state of a gradate sorus, but 
