ig8 REMINISCENCES OF 



it was far more accurate than others which he 

 substituted at later periods of his life. The larger 

 quarto work was never completed. The first volume 

 appeared as well as one or two parts of the second. 

 He showed that nearly all the plants living at 

 the most ancient geological epochs were Cryptogams, 

 chiefly belonging to the great families of the Ferns, 

 the Equiseta, and the Lycopods, or club mosses ; but 

 that whilst most of those living now are small 

 herbaceous objects, during the carboniferous age 

 they were magnificent forest trees, whilst none of 

 the modern denizens of our forests were .to be found 

 among them. Of these ancient Cryptogams he de- 

 scribed 225 species, whilst in the entire range of 

 rocks he discovered about twenty-six that could be 

 regarded as representatives of the flowering plants, 

 and in twenty of these he was certainly mistaken. 

 Nevertheless, the above was a grand generalisation, 

 which, to a large extent, continues to be true, not- 

 withstanding the gigantic progress of research during 

 the last seventy years. No wonder that its author 

 at once took the position he so well deserved, of the 

 foremost palaeo-botanist of the age. 



Unhappily Brongniart was the first to strike a 

 serious blow at his own philosophy. During his 

 earlier years nothing was known of the internal 

 organisation of these fossil plants. We had nothing 

 beyond their external forms to guide us as to their 

 position in nature's scale. 



Many years later two specimens fell into Brong- 



