488 OPHIOGLOSSALES 



axial system by foliar lacunae, after the manner of the phyllosiphonic type. 

 But in its lower parts, Tmesipteris is typically cladosiphonic : it is thus 

 seen that a phyllosiphonic structure may arise in a cladosiphonic stock,, 

 and the illustration is taken from that group of plants which show analogies 

 with the Ophioglossaceae in other respects : on comparison of Fig. 268 

 with Fig. 258 of O. Bergiamun, the essential similarity of the two cases 

 is evident. But in the Ophioglossaceae the structural dominance of the 

 leaf is on our hypothesis a consequence of the advance of the leaf towards 

 megaphylly, combined perhaps with weakening of the axis which bears 

 it. It matters little how the balance between the parts of the shoot is 

 disturbed : the progression would be essentially the same in either case. 

 These considerations show, in the first place, that it cannot be assumed 

 that all phyllosiphonic plants are necessarily derived from a distinct and 

 primitive phyllosiphonic stock, such as Jeffrey assumed for his Pteropsida : 

 and secondly, that analogy of their structure with Tmesipteris indicates a 

 possible origin of the phyllosiphonic type in the Ophioglossaceae, phyletically 

 quite distinct from that shown by the true Ferns. 1 



The leaf-trace in the Ophioglossaceae is typically a single strand, which 

 branches within the cortex into strands which vary according to the pro- 

 portions of the leaf which they serve : 2 these facts accord with a theory of 

 origin of the leaf from the simpler type. It is interesting to note that 

 " the branching of the leaf-traces within the cortex is very characteristic 

 of Sphenophyllum" z As regards the structure of the collateral strands of 

 the leaf, those of the larger forms show similarity to the Eusporangiate 

 Ferns, the smaller correspond rather to those of the larger-leaved strobiloid 

 Pteridophytes. 



The occurrence of secondary thickening in the Ophioglossaceae is 

 occasional rather than typical of them. In Helminthostachys it is absent : in 

 Ophioglossum a feeble growth has been occasionally seen ; but in Botrychium 

 it is a marked feature, and extends from the axis onwards to the base 

 of the root. This inconstant occurrence of secondary activity, sometimes 

 feebly shown, has its parallel in other affinities, both of Filicales on the 

 one hand, as in the Marattiaceae, 4 and of the Psilotaceae, where a develop- 

 ment very like that of Ophioglossum has been seen in Psilotum? The 



J It will be seen below that certain Ferns, for instance the Botryopterideae, are not 

 phyllosiphonic : thus the anatomical distinction of Jeffrey breaks down on both sides. 



- The only known exceptions to this are in O. pendtdum and simplex, where the leaf- 

 trace is inserted on the cauline system as several distinct strands. These species belong, 

 however, to a section of the genus believed to be highly specialised rather than primitive 

 types : and this character itself must, by analogy with the similar cases in the Ferns, be held 

 to be derivative (see Ann. of Bot., xviii., pp. 209, 215). 



3 Scott, Studies, p. 91. 4 Farmer, Ann. of Bot., xiii., p. 440. 



5 Boodle, I.e. Scott (Journ. Roy. Micr. Soc., 1906, pp. 519-521) has described 

 under the name of Botrychioxylon a new genus from the Lower Coal Measures, with 

 "radially seriated wood, apparently of secondary character." It is related anatomically 

 to Zygopteris somewhat as Botrychium is to Ophioglossum. 



