156 ANATOMY OF THE GYMNOSPERMS 



probable that there may have been some one or more intermediate 

 forms of which we have no present knowledge. Our present 

 studies, on the other hand, show clearly that we must bring into 

 this phylum two other genera of an obviously higher degree 

 of development, but which have commonly been ranked with 

 the Abietineas and which, according to Eichler (15), occupy the 

 highest position in the scale. This position is untenable upon 

 anatomical grounds which give us reason to believe that Dam- 

 mara and Araucaria (including, of course, Araucarioxylon) are 

 not only inferior to the Coniferales as a whole but that they are 

 also distinctly Cordaitean. Accepting this view and the fact 

 that Dammara is the inferior genus, the sequence would place 

 Cordaites at the base and Araucaria at the top, with Walchia as 

 the immediately ancestral form of the latter. This relation is not 

 only natural but it is justified on anatomical grounds. 



The tendency to segregation of the bordered pits, as exhibited 

 by Poroxylon, suggests the relation of this genus to others in 

 which such a feature is fully expressed, and it thereby forms the 

 basal member of another series. From the opposite point of view 

 it has been shown that the occurrence of two-seriate pits in Pinus 

 and others of the Coniferales, as well as in Gingko, points to a com- 

 mon origin for such genera in a type with multiseriate hexagonal 

 pits, and that both Dammara and Araucaria must likewise center 

 in the same generalized form. This gradual convergence is justi- 

 fied on other grounds, and the genus Poroxylon among known 

 forms most nearly fulfils the requirements of the case. We may 

 therefore look upon it as lying between the Cycadofilices and all 

 the higher gymnosperms, giving rise to two lines of descent, the 

 first of which embraces the Cordaitales, as already described, while 

 the second shortly divides once more. This secondary division 

 gives rise on the one side to the Gingkoales, and on the other to 

 the Coniferales. The anatomical data already discussed, when 

 viewed collectively, show that the general sequence within the lat- 

 ter would be (i) the Taxoideae, (2) the Taxodiinae, (3) the Cupres- 

 sineae, (4) Abies, (5) Tsuga, (6) Pseudotsuga, (7) Larix, (8) Picea, 

 and (9) Pinus, of which one division, (II), represents the highest 



