i 4 o STUDIES IN FOSSIL BOTANY 



of still greater dimensions than the largest known 

 specimens of L. Harcourtii. 



Williamson sometimes said that it was an unfortunate 

 chance for fossil botany that the first Lepidodendroid 

 stem, of which the structure was investigated, happened 

 to be that of L. Harcourtii. The absence of secondary 

 wood in this species, which we now know to have been 

 quite exceptional, led Brongniart to believe that the 

 Lepidodendra generally were without exogenous growth. 

 On this ground, he removed Sigillaria, in which the 

 secondary tissues were discovered early, both from the 

 Lycopodiaceae and from the Cryptogams, and classed 

 the genus among Gymnospermous Phanerogams. Even 

 when Lepidodendra with secondary wood began to be 

 discovered, there was for a long time a strong tendency, 

 on the part of the French school of palaeobotanists, to 

 regard all such specimens as really Sigillariae, and so to 

 keep up the supposed distinction. We now know, chiefly 

 as the result of Williamson's researches, that most of 

 the Carboniferous Lepidodendra agreed essentially with 

 Sigillaria in their anatomy, and that the two genera were 

 closely allied members of the Lycopodiales. The contro- 

 versy, however, proved valuable as a stimulus to research. 



Lepidodendron Wunschianum} a species which is 

 abundant in the volcanic beds of Arran, belonging to 

 the oldest part of the Carboniferous formation, essenti- 

 ally resembles L. Harcourtii in its primary structure, 

 and needs no detailed description here. 2 Secondary 

 wood has only been found in stems with a primary 



1 Possibly referable to the genus Lepidophloios. 



2 Seward and Hill are inclined to identify these two types, but I do not 

 find sufficient evidence to justify us in uniting them. 



