XIV] COMPARISON BASED ON THE GAMETOPHYTE 295 



at the margin of the flattened thallus, giving it an irregular or lopsided form 

 (Fig. 267, B, D). The addition of periclinal divisions in the cells lying behind 

 the apical region, giving the massive cushion, is all that is then required to 

 complete the ordinary cordate type where the apex is distal. In these 

 peculiar lopsided forms, such as are seen in Anemia, or in A^tograrume, which 

 show that peculiar method of perennation already described, the form which 

 is adopted results from the establishment of a lateral apex, followed as before 

 by periclinal divisions. The initiation of such divisions is exemplified in 

 the ontogeny of every prothallus which has a fleshy cushion. All these 

 forms, together with the large and fleshy prothalli of the Marattiaceae and 

 Osmundaceae, can without difficulty be referred back to the simple filament 

 as their ultimate phyletic source. It is indeed, as a rule, from that source 

 that they originate ontogenetically, and to it they may frequently return as 

 a consequence of growth under special conditions (Fig. 266). 



The massive mycorhizic prothalli of the primitive Ophioglossaceae 

 appear to have diverged far from the simple filamentous structure. There 

 is, however, no need to assume that their present state is itself primitive. 

 The key to their origin is probably to be found in the fact that certain well- 

 known Ferns under special conditions of culture produce cylindrical pro- 

 cesses, which bear gametangia {Scolopendrium, etc. Lang). If the massive 

 prothalli of primitive Ferns, such as the Marattiaceae and Osmundaceae, 

 living in the absence of light were to do the like, assuming at the same 

 time a mycorhizic habit, the result would be something of the same nature 

 as the saprophytic prothalli of the Ophioglossaceae : in which case they 

 also would be referable back ultimately to a filamentous origin, which has 

 probably been the source for the gametophyte of all of the Filicales. 



It has already been shown that the two types of gametangia have many 

 points in common, and that it is reasonable to suppose that they diverged 

 as a consequence of sexual differentiation from a single type of gametangium. 

 In the filamentous prothalli of Schizaea and Trichomanes their position is 

 commonly terminal on lateral branches. This brings them into line with the 

 gametangia of Algae, which frequently have a like position. Comparison of 

 them within the Filicales shows that the archegonium was standardised 

 early, assuming that degree of constancy of structure which is seen even in 

 primitive Ferns. The venter is always sunk in the tissue of the thallus, 

 except in the simplest examples of filamentous prothalli, such as Schizaea. 

 But there is some variation in the protrusion of the neck. In Eusporangiate 

 types it may be almost wholly immersed, as in Marattia and Ophioglossuvi ; 

 but in Leptosporangiate types the neck projects freely. The same is the 

 condition of the antheridia. As shown in the table above there may be 

 differences in number of the spermatocytes, and where the number is large 

 the antheridium is massive and thick-stalked, or even wholly immersed 



