﻿Manchester Memoirs, Vol. xlix. (1905). No. 12. 23 



which show similar characters. The great trees of the 

 genus Pitys, with wood of the Araucarian type, found by 

 Witham of Lartington, on the Tweed, and at Craigleith, 

 near Edinburgh, in the early days of fossil botany, have 

 proved on re-investigation to possess strands of primary 

 wood in their pith, which appear to be homologous with 

 those found in Lyginodendron or Poroxylon. There is 

 ever)' reason to believe that Pitys represents the stems of 

 ancient members of the Cordaitean stock. Thus it appears 

 that there is a series of forms connecting the stems of the 

 Lyginodendron type with those of typical Cordaiteae as 

 investigated by M. Renault (Scott, '02 L and '02"). In fact, 

 in the case of some of the lower carboniferous stems, the 

 characters exactly hold the balance between the two 

 groups, and it is only when evidence as to the character 

 of the foliage is available that one can say to which they 

 belonged. 



Thus, taking into consideration both the seed- 

 characters and those of the vegetative anatomy, there is 

 every reason to believe that a real affinity existed between 

 the Cordaitese and the Pteridospermeae. The Cordaiteae 

 are so ancient that their origin lies further back than our 

 records extend. But the evidence distinctly points to the 

 conclusion that at some remote period they sprang from 

 the same stock to which the carboniferous Pteridospermeae 

 belong, and consequently that they too, like the Cycado- 

 phyta, were ultimately derived from cryptogams allied to 

 the ferns. Considering, however, the enormous antiquity 

 of the Cordaiteae as compared with the comparatively 

 late appearance of true Cycadophyta, it seems probable 

 that the two classes originated separately, and at very 

 different periods, from the plexus of primitive, fern-like 

 seed plants, which played so important a part both in 

 Devonian and carboniferous times. While certain families 



