SUMMARY 493 



some degree of real affinity, and strongly confirms the initial comparison 

 of the spike with the sporangiophore of the Sphenophyllales. Lastly, the 

 anatomical comparison of the Psilotaceae with the Ophioglossaceae has 

 shown not only the interesting transition from the cladosiphonic to the 

 phyllosiphonic structure, but also that in the upper region the wood of 

 Tmesipteris is mesoxylic, as it is also in ffelminthostackys, while feeble 

 secondary development, analogous to that in Ophioglossnin and stronger in 

 Botrychiuni, is seen both in Tmesipteris and in Psilotum. These several 

 characters form a cumulative body of evidence, confirming the comparison 

 of the shoot and of the sporangiophore in the Sphenophyllales with those 

 of the Ophioglossales : the nearest approach among living plants being 

 between the Psilotaceae and Ophioglossum^ 



It would thus seem probable that the Ophioglossaceae sprang from 

 some offshoot of the sporangiophoric Pteridophytes, allied in some degree 

 to the Sphenophyllales, and possessing early a saprophytic habit of the 

 underground prothallus. That this encouraged a peculiar specialisation of 

 the sporophyte, which shared occasionally, though not generally, in the 

 mycorhizic habit, but not so far as to lead to the cessation of self- 

 nutrition. That the exigencies of the underground habit were met by an 

 enlargement of the leaves, culminating finally to the monophyllous state. 

 A parallel enlargement of the sporangiophore with that of the leaf was a 

 natural consequence, since in homosporous forms, as comparison shows, 

 the spore-output usually marches with the vegetative development. If this 

 were so, then the spike would never in its descent have been anything 

 other than it is now normally seen to be, viz. a spore-producing part, 

 originally of the nature of a sporangiophore, and seated in a median position 

 on the adaxial face of the sporophyll. 



Referring in conclusion to the theory of the strobilus, the Ophio- 

 glossaceae readily conform to it. The shoot, with its rare dichotomous 

 branching, appears as a simple strobilus, while the indeterminate position 

 of the root in the embryo bespeaks the accessory nature of that part 

 upon it. The axis bears leaves, which are of one order only. The 

 spore-producing parts appear earlier in the individual life than in any 

 other group of Pteridophytes, and this indicates a probability that all the 



1 1 wish to state quite explicitly that the homology of the I'silotaceous synangium 

 with the Ophioglossaceous spike is no new opinion on my part, though additional and 

 more detailed evidence is here adduced to support it. It was accepted by me in 1891 

 (Proc. Roy. Soc., p. 270) and more fully stated in 1893, on the basis of developmental 

 evidence (Proc. Roy. Soc., vol. liii., p. 22): this view has never been relinquished. I 

 emphasize this here because a passage recently published appears to suggest that I do 

 not uphold that homology (Scott, Progressus Rei Bolanicae, i., p. 163). My position is 

 unchanged, except in so far as I now include the Sphenophylleae also in the comparison : 

 the suggestion of this came from Dr. Scott (On Cheirostrobus, Phil. Trans., vol. clxxxix., 

 1897, p. 27), and it greatly strengthens the comparison originally drawn by Celakovsky. 

 There may be differences of opinion as to what morphological rank these parts hold, 

 or how ultimately they came into being : these are, however, separate questions from the 

 recognition of their homology. 



