﻿Manchester Memoirs, Vol. xlix. (1905), No. % ^$. 3 



hamium. This plant has played a most important part 

 in the growth of our knowledge of the Carboniferous 

 Spermophytes, and it is from Williamson's early investiga- 

 tions of its structure, and that of an allied genus, that the 

 first advance towards our present position must be dated. 

 (Williamson, '73-) 



The stem of Lyginodendrou — the part first discovered — 

 shows a structure which at first sight appears anything but 

 .fern-like (Williamson and Scott, '95). There is a large 

 pith surrounded by a zone of wood and bast, with a 

 recognizable layer of cambium between the two ; the 

 greater part of this zone has its elements arranged in 

 radial series, and manifestly grew indefinitely in thickness, 

 as in any Gymnospermous stem ; the structure of the 

 wood and bast generally is in fact much like that of the 

 •corresponding zones in a recent Cycad. Around the pith, 

 'however, several strands of primary wood may be dis- 

 tinguished, a structure not represented in the vegetative 

 stem of any recent Gymnosperm. The development of 

 these circummedullary strands was for the most part 

 centripetal i.e. advancing towards the pith, (as shown by 

 the position of the spiral elements, which are always the first 

 to be formed). The primary strands pass out through 

 the secondary wood into the cortex, where they form part 

 of the vascular bundles which supply the leaves. These 

 outgoing bundles have precisely the structure of the foliar 

 bundles in recent Cycads. The presence of the primary 

 strands of mainly centripetal wood around the pith 

 suggested a search for corresponding structures in the 

 axial organs of Cycads, and they have been found in 

 certain genera in the peduncles of the cones, organs which 

 in other respects also appear to have retained primitive 

 characters (Scott, '97). 



The evidence afforded by the stem-structure of 



