216 D. H. Scott. 



Eenault, wliich has an essentially Coniferous, and more particularly 

 an Araucarian structure, the whole of the wood havinjç been centri- 

 fugally developed. In Poroxylon, however, which in habit, and general 

 structure, appears to agree most nearly with Cordaiteae ^), there are 

 strands of centripetal wood surrounding the pith, the stem-structure 

 closely approaching that of Lyghiodendron. As we now know, the 

 Poroxijlon tj'pe of stem extends back to the Lower Coal-Messures. 

 Still more ancient stems {CaJamopitys and Pitys) of Lower Carboni- 

 ferous age, likewise have a structure intermediate between that of the 

 Cordaiteae and Lyginodendreae (Scott, 1902). The leaf of Cordaiteae 

 and of Poroxylon, shows, in its mesarch or exarch bundles, a distinctly 

 Pteridospermous character. On grounds of anatomy and seed-structure 

 it thus appears to be established that the Cordaiteae and the Pterido- 

 spermeae were related, though the data do not yet enable us to 

 determine their common point of origin. 



I do not propose to enter, in the present article, upon any dis- 

 cussion of the relations of the Couiferae, a subject on which my 

 views have been repeatedly expressed.-) My friend Mr. Seward, in 

 his valuable monograph of the Araucariales, just published -^i, has 

 stated, with much ingenuitj', the case for the derivation of these 

 Conifers from a Lycopodineous ancestry. It would unduly extend this 

 article to enter into the controversy here, and I am absolved from 

 doing so by the consideration that the Palaeozoic Coniferae aiford, as 

 yet, scarcely any data for the discussion. 



In the foregoing sketch of the present position of Palaeozoic Botany, 

 I have dealt with the subject almost wholly from a morphological 

 and evolutionary point of view. It is important, as was pointed out 

 early in the article, to bear in mind the very special and limited 

 character of the floras from which our evidence is derived. Although 

 the fossil remains of Carboniferous age are extremely abundant com- 

 pared with those of other periods, it is for the most part only a 

 special kind of flora which is preserved. The petrified specimens, 

 on which so much depends, are not, after all, of very numerous types 

 and are limited to few localities. The opening up of new sources of 

 material showing structure is urgently to be desired, and it may be 



1) As M. Graud'Eury has recently shown, it is probable that Rhahdocarpus, 

 Broug-niart, was the seed of Poroxylon. Comptes Rendus, T. 140, 1905, p. 995. 



'■*) Studies in Fossil Botany, 1900, p. 521. Early History of Seed-bearing Plants, 

 1905, p. 24. The Fern-like Seed-plants of the Carboniferous Flora. Congrès Internat, 

 de Botanique, 1905, p. 294. The "Origin of Gymnosperms" at the Linnean Society, 

 New Phytologist. June, 1906. 



*) Phil.^Trans. Royal Soc, Ser. B, 190ß. 



