4 PAL.S:ONTOLOGICAL BOTANY. 



only be distinguished by the fructification; and as this is rarely 

 seen in fossil ferns, it is often impossible to come to a decided 

 conclusion in regard to them. A leaf of Stangeria paradoxa 

 was considered by an eminent botanist as a barren fern frond, 

 but it ultimately proved to be the leaf of a Cycad. The leaf 

 of Cupania filicifolia, a Dicotyledon, might easily be mistaken 

 for that of a fern ; it resembles much the frond of a fossil fern 

 called Coniopteris. The diverse leaves of Sterculia diversi- 

 folia, if seen separately, might easily be referred to different 

 plants. In the same fern we meet also with different kinds of 

 venation in the fronds. Similar remarks may be made in 

 regard to other plants. Harvey has pointed out many diffi- 

 culties in regard to sea-weeds. 



As regards the materials for a fossil flora, the following 

 remarks of Hugh Miller deserve attention : — 



*^ The authors of Fossil Floras, however able or accom- 

 plished they may be, have often to found their genera and 

 species, and to frame their restorations, when they attempt 

 these, on very inadequate specimens. For, were they to pause 

 in their labours until better ones turned up, they would find 

 the longest life greatly too short for the completion of even a 

 small portion of their task. Much of their work must be of 

 necessity of a provisional character — so much so, that there 

 are few possessors of good collections who do not find them- 

 selves in circumstances to furnish both addenda and errata to 

 our most valuable works on Palasontology. And it is only by 

 the free communication of these addenda and errata that 

 geologists will be at length enabled adequately to conceive of 

 the by-past creations, and of that gorgeous Flora of the Car- 

 boniferous age, which seems to have been by far the most 

 luxuriant and wonderful which our emphatically ancient earth 

 ever saw." 



The bark of trees at the present day often exhibits 

 different kinds of markings in its layers. This may be illus- 



