58 PEOCEEDI^'GS OF THE 



these organs are of too advanced a type to throw light on the 

 probable derivation o£ the group. 



The anatomical characters clearl_y indicate an affinity with the 

 Fern-stock closer than that with any other Cryptogamic phylum. 

 The male organs, now beginning to be known, further support 

 this affinity. Hence we are led to conclude that the Pterido- 

 sperms had a common origin with the Perns, and may assume 

 that the common ancestors resembled Perns in being Cryptogamic 

 rather than Spermophytic. 



The difficulty arises from the surprising extension of the 

 Pteridosperms, which threaten to absorb every thmg Palaeozoic 

 that used to be called a Pern. But this idea is of the nature of a 

 " scare." Though the Perns were not the dominant group in the 

 Palseozoic that they were once supposed to be, there is as yet no 

 justification for the extreme view that they were non-existent. 

 The important position of the Botrj^opteridese as a synthetic family, 

 perhaps most nearly i-epresenting among known Palaeozoic plants 

 the common stock from which Perns and Pteridosperms may have 

 sprung, has already been indicated by a previous speaker. I agree 

 with him that the family is best regarded as a special type of a 

 much more extensive group. 



As regards the other great question under discussion — the 

 affinities of the Coniferse — the data are still very inadequate, for 

 we have little knowledge of the structure of early forms of 

 Coniferae. In some Permian plants referred to Araucarieae, the 

 structure of the wood was of the Araucarioxulon type, a wide- 

 spread form of wood, common to Cordaitese, Pteridosperraeae, and 

 even Botryopterideae, but almost wholly absent from the Lycopod 

 phylum. Whatever this character may be worth, it favours a 

 connnon origin of the Araucarian Conifers with the Cordaitean 

 and Pteridospermic series. 



The existence of the Cordaitese, offering clear points of agi-ee- 

 raent at once with Pteridosperms, Cycads, Giukgoales, and Conifers, 

 certainly suggests that all these groups ultimately had a common 

 oingin from the same great plexus of primitive Pilicineae. Ginlr/o 

 itself forms a bond of union between the Cordaitean phylum and 

 the Taxaceae among Conifers. 



The Lycopods attained a high development on their own lines, 

 producing seed-like organs in certain cases, and showing some 

 anatomical analogies with Conifers. A more exact comparison 

 appears to indicate that these characters are homoplastic, and not 

 indicative of any affinity with the higher plants. 



