2 WATSON, Two Species of Lepidodendron Harcourtii. 



He reviews the work of Witham ('32), Lindley and 

 Hutton ('33), Brongniart ('37), and Bertrand ('91). 



Williamson states that he has only recently ('93) 

 obtained a section of the true L. Harcourtii, and that the 

 specimens so far figured as L. Harcourtii belong to a new 

 species, L.fuliginosum, Will. (Williamson, '87). William- 

 son states that he has now ('93) obtained specimens of 

 Lepidodendron Hartourtii, Witham, from Dulesgate still 

 retaining their leaves, which were lacking in Bertrand's 

 specimens. 



These specimens from Dulesgate are not L. Harcourtii, 

 Witham, but L. Hickii, sp. nov. 



In one section only fin. in diameter the cells of the 

 " medulla" are actively dividing. 



Williamson suggests that this involves an increase in 

 diameter of the primary wood. This suggestion will be 

 considered later. 



Williamson then states that he is unable to accept 

 all Bertrand's conclusions with regard to the corona of 

 L. Harcourtii, Witham, but acknowledges the value of 

 his work. 



He then states that in "young specimens " the pro- 

 jections from the margin of the wood are " scarcely 

 visible," but that the leaf traces are much more prominent 

 objects. 



From this idea he draws the conclusion that the state 

 of the corona is of inconsiderable taxonomic value. 



The importance of the conclusion as to the corona 

 which Williamson drew from these specimens necessitates 

 an investigation of the evidence of the identity of these 

 branches with the true large stems of L. Harcourtii ; this 

 is undertaken below ; for the present, it will suffice to say 

 that Williamson's chief evidence seems to have been the 

 " double " leaf traces, which are now known to occur in 



