LINXEAX SOCIETY OF LONDON". 75 



when raicrophyllous, are only so by reduction, and consequently 

 that the anatomical characters correlated with microphylly are 

 not essential to the group. On the other hand, as we have seen, 

 the Lycopods stand apart as a genuine microphyllous class, 

 unconnected by any known transitions with the large-leaved phyla. 

 At the same time one cannot admit any very close relationship 

 between the Articulattc and the Ferns ; tlieir ancestors, though in 

 all probability megaphyllous, may have been much less like Ferns 

 than any of the known Primohlices. For these reasons, which I 

 cannot now develop at greater length, it seems to nie clear that 

 the attempt to divide Yasculares into two main series only must 

 be given up, at least for the present. 



I have proposed a threefold division, into Pteropsida (Ferns and 

 all Spermophyta), Sphenopsida (Equisetales, Sphenophyllales, and 

 Psilotales), and Lycopsida (Lycopods alone). The isolation of the 

 Lycopods while the Psilotales are put in ISphenopsida has been 

 criticised, and justly so — the position of the little family Psilotales 

 is a great difficulty, and I do not think we are yet in a position to 

 solve it, in the absence of all geological evidence of their history. 

 The group has certain definite characters in common with the 

 tSphenophylls, namely the nature of the sporangial apparatus and 

 the anatomy : for these reasons some modern authors have united 

 them in one class. On the other hand, the Psilotales have other, 

 less definite characters in common with the Lycopods — the dicho- 

 tomous branching, the alternate leaves, and to some extent tlie 

 habit (in the case of Tmeslptpris). The older writers always put 

 them in this class, but at that time the Sphenophylls were prac- 

 tically unknown. We are not at present able to reconcile the 

 two apparent directions of affinity. It is best to emphasize the 

 Sphenophyll relation as the more definite, and otherwise to reserve 

 judgment. 



It should be mentioned here that Prof. Lignier gives the 

 Psilotales quite a dift'erent position, regarding them as the most 

 primitive of the Lycopod series, and consequently of all living- 

 vascular plants. He believes that in the earliest Pteridophyta 

 the sporangia were terminal on certain cauloids (derived from 

 thallus-branches), and that the Psilotales only differ from this 

 type in having the sporangia grouped on special short branches. 

 In arriving at this opinion the author allows himself to be too 

 nnich influenced by the very problematic Devonian fossil Psilo- 

 phytoh, of which we really know nothing definite. I cannot accept 

 a view which i^^nores the points of agreement between the Psilotales 

 and the Sphenophyllales, and the probability that the former have 

 suffered some reduction in organization in consequence of their 

 epiphytic habit. 



Prof. Lignier is of opinion that all his Phyllinese (^. e. all 

 Yasculares except Lycopods and Psilotales) are descended from 

 the Primofilices. 



As regards the Articulata% I have already suggested that if we 

 accept this view we must take Primofilices in an extremely wide 



