XXXIX] COMPARISON 95 



removed from it (Fig. 644, A, B). On the other hand, in MoJiria the indusium 

 is a new formation initiated superficially (z, Fig. 645), while in NotJiolacna 

 the indusioid margin continues as the direct 

 outgrowth of the marginal cells. The same 

 argument may be applied here as in the sori 

 of the Pteroid Ferns (pp. 54, 55), and the 

 conclusion drawn that in both cases there has 

 been a phyletic slide of sporangial production 

 from the original position on the margin of 

 the blade to its surface. The ontogenetic 

 comparisons accord with this, and the physio- 

 logical probability of the better protection of 

 the young sporangia supports it. In other 

 Gymnogrammoid Ferns than Notkolaena the Y\g.64i,. Mohrmcaffronim [U)T)t&w., 

 sporangia arise farther from the margin, and drawings after Prantl showing the rek- 



^ ° o ' tion of the sporangia (s) to the marginal 



they are liable to be spread inwards along series, and to the indusium (?), as "seen 

 the veins, giving the usual character of the i" surface view, and in section. ( x 145-) 

 group. All steps in this invasion of the lower surface can be illustrated by 

 comparison of the species among themselves. In all such questions it is to 

 be remembered that the sporangium is the most constant and important 

 part; the indusia or leaf-margins, and even the flattened blade itself, are 

 ancillary when a broad view of ultimate origin is taken. 



Against any close alliance between the living Schizaeaceae and the 

 living Cheilanthoid Ferns there is the fact of the sporangia of the former 

 being themselves relatively archaic, while those of the latter are definitely 

 of the Leptosporangiate type. This fact raises again the general question 

 of phyletic changeof the sporangia themselves. The broad lines of comparison 

 now before us for Ferns at large m.ake such a change appear more than ever 

 probable: indeed the conclusion that a reduction in size and spore-output has 

 happened repeatedlyseems tobe inevitable. Here theextraordinary variability 

 in the individual sporangium seen in certain Gymnogrammoid Ferns has its 

 intimate bearing on the general problem {Liavea, Ceratopteris, Jamesonia, 

 Cryptograimne). In all of those genera the sporangia appear variable as well 

 as relatively archaic: but they all bear their sporangia upon the leaf-surface, 

 far removed from the margin. What may be concluded from that is, that 

 in these Gymnogrammoid Ferns changes in position and in structure of the 

 sporangium have not progressed along strictly parallel lines, though still they 

 have affected them all. 



Anatomically the Schizaeaceae are as varied as any family of the whole 

 Class of Ferns (Vol. Ii, p. 157). The most advanced in this are Anemia and 

 Mohria, those genera with which the Gymnogrammoids may best be com- 

 pared as regards their sporangia. On both sides of the comparison the 



