88 OPHIOGLOSSACEAE [ch. 



Ophioglossaceae, it has been seen in many instances that it arises either by 

 marginal abstriction, or by extra-marginal segregation from the supply lead- 

 ing to the sterile blade (Figs. 359, 360, B). These are precisely the alternative 

 methods of supply to the pinnae of Ferns (compare Vol. I, pp. 172—174), 

 and incidentally in B. virginianuin the supply to its own sterile pinnae arises 

 in the same way (Fig. 360, A). From this the conclusion seems probable 

 that the spike is of pinna-nature. But since the vascular supply habitually 

 comes off equally from both sides, it was suggested by Roeper (^Bot. Zeit. 

 1859, p, 241) that the spike really represents two lateral pinnae fused 

 laterally, and Anemia with its two erect fertile pinnae was naturally cited 

 as evidence. Though this suggestion lacks support from external morphology 

 (unless the fissions of the spike in Ophioglossiivi be held as such) it may be 

 held as a possible view, though nothing would appear more natural than 

 that, as a laterally originated pinna moved to a median position, a vascular 

 connection should be established secondarily with the other side of the 

 leaf-supply. The fusion-theory may therefore be held as still "not proven." 

 A consideration that strengthens the Fern-pinna theory as against com- 

 parisons with Lycopods and Psilotales is that in Tmesipt^ris , Cheirostrobus, 

 and Sphenophyllum, in all of which adaxial spore-producing members exist, 

 the origin of the vascular supply is median, and adaxial, not lateral as in 

 the supply to the Ophioglossaceous spike. This seems to be a real and 

 valid anatomical distinction, which ranges the Ophioglossaceae with the 

 Filicales rather than with such plants as those named. Further, the positive 

 and close analogy between the stelar morphology and the venation of the 

 Ophioglossaceae on the one hand, and that of relatively primitive Ferns on 

 the other, gives cogency to the conclusion that the spike is of pinna-nature, 

 and that consequently the Ophioglossaceae are Filical in their alliance. 

 Finally, this view accords with the fact that in BotrycJii2t.in the sterile blade 

 may not infrequently be in part or in whole fertile (Fig. 338, p. 46). The 

 balance of evidence thus appears to be clearly in favour of ranking the 

 Ophioglossaceae as primitive Ferns, and the spike as of pinna-nature. 



In comparing the Ophioglossaceae among themselves it is necessary to 

 be clear which are the more primitive types, and which the more specialised. 

 In forming an opinion use must be made of the criteria established in Vol, I. 

 A very obvious character is the venation. In Botrychiinn and Helviintho- 

 stachys the venation is always open, in Ophioglossum it is as constantly 

 closed: and not only is it closed, but it forms an elaborate small-meshed 

 reticulum, such as appeared only in the Secondary Period, while the venation 

 of the other genera is Palaeozoic in type. This at once marks off Ophio- 

 glossum as probably derivative^ Passing to stelar structure, Botrychium and 

 ^ Among the many disabilities consequent on the Great War has been the difficulty of interchange of 

 scientific literature. To this I ascribe the omission of Prof, von Goebel to notice my memoir on leaf- 

 architecture {Trans. Roy. Soc. Edin. 1916, Vol. XLI, p. 657) in his valuable study Gesetziniissigkeiten 



