RESEARCHES ON THE CARBONIFEROUS FLORA. 263 



In his later works Williamson placed Sphenophyllum in 

 the family Calamarieae, following Weiss in this, but he fully, 

 recognised that their systematic position is still an open 

 question. A remarkable, newly discovered fructification, 

 lately described by Solms-Laubach under the name of 

 Bowmanites Romeri, appears likely to throw some further 

 light on the affinities of the group (45). 



The next family which we propose to consider in illus- 

 tration of Williamson's work, is one of which he may fairly 

 be called the discoverer. This is the group Lyginodendreae. 

 Although a few specimens had been previously observed, 

 the investigation and interpretation of their structure has 

 been practically in our author's own hands. The first full 

 account of the genera Lyginodendron and Heterangium oc- 

 cupies the fourth memoir of the series in the Philosophical 

 Transactions. At this time the main points of the anatomy 

 of the stem in both oenera were determined. Williamson, 

 with his usual acumen in such matters, at once saw that the 

 two plants were closely related, though there is a striking 

 difference between them, Lyginodendron, at least in the 

 normal form, having a large pith with a ring of distinct 

 bundles around it, while in Heterangium the wood extends 

 to the centre, and there is no pith at all. At that time 

 Williamson already suspected that the leaves which he then 

 called Edraxylon belonged to Lyginodendron, and " then," 

 he said, "strange as it may appear, Lyginodendron will 

 become an undoubted arborescent Fern ". This conclusion 

 may well have appeared strange, for the anatomy of the 

 stem of Lyginodendron is much like that of a Cycad, 

 especially as regards its secondary tissues. 



Restorations of the stems of both oenera are fioured in 

 the memoir of 1872. They are extraordinarily good, that 

 of Heterangium especially being so exact that it could not 

 be improved upon at the present day after all the additional 

 work that has been done. 



The foliage was fully described in 1874 (mem. vi.) under 

 the name of Rachiopteris aspera, and shown to be of the 

 Sphenopteris type. The proof that it really belonged to 

 Lyginodendron was, however, reserved for a much later 



