Steverson ] iOU [June 18, 



stray logs which floated down to the marshes. From the Carboniferous 

 to the Trias, a great change is shown by the fossils, but we have no evi- 

 dence to prove that this change is a true exposition of the actual change. 

 For aught we can tell to the contrary, a flora closely allied to the one 

 termed Triassic may have existed during the Carboniferous. In the Cre- 

 taceous the condition is little better. In the lower portion, leaves of 

 dicotyledonous plants occur in prodigious numbers, but they are not of 

 plants growing where the leaves occur. For the most part they are single 

 leaves, washed in by streams from the land. Between this sandstone and 

 the Lignite Group, there is an interval mostly unrepresented at the East, 

 but at the West occupied by a mass of shales, limestones, and fine-grained 

 sandstones, one thousand to two thousand feet thick, and absolutely 

 barren of leaves everywhere. This was a long period, during which, 

 under the sea, nothing but fine-grained materials were deposited. In the 

 Lignite Group, leaves ai-e numerous, but so far as has fallen undei* ray 

 observation, they are in the same condition as those at the base of the 

 Cretaceous. 



Such is the record of plant-life — a record little better than a blank, with 

 here and there a few markings, many of which are too indistinct to be 

 deciphered. In each horizon which yields relics of plants by far the 

 greater portion of the area is barren — even in the Carboniferous age, how 

 small a proportion of the rocks are leaf-bearing in the most favorable 

 localities, while the whole vast area west from the Mississippi has yielded 

 but a beggarly array of specimens. At best, the specimens are fragment- 

 ary. The same frond on a fossil fern, when broken up into its pinnules, 

 may yield two or three genera and half a dozen species. When only 

 fragments are found, it is impossible for the palaeontologist to resist the 

 temptation to make species. Describing fossil ferns from fragments, is 

 almost as accui-ate work as making genera and species out of fossil teeth 

 of sharks. In the case of leaves of dicotyledonous plants, the matter is 

 evidently worse. The limit of variation of a species has never been ap- 

 proximately determined among living plants, where one has the whole 

 tree at hand. With only imperfect and separated leaves to study, it 

 would seem almost impossible to determine this matter respecting extinct 

 plants. 



Like vertebrate remains, vegetable relics may be made serviceable. 

 The character of the coal flora has been so carefully studied for many 

 years that it is quite well understood. Here, indeed, the matter in many 

 cases is quite simple, for the roof of a coal-bed as exposed in the tunnel 

 of a mine, not infrequently exhibits the material for the reconstruction 

 of an entire plant. Unfortunately, attempts at re-construction are not 

 common, and the investigator is usually satisfied to describe fragments 

 as species, in preference to carefully studying their relation. But the 

 horizon of these plants is now fixed, their general type is well understood, 

 and they can be used as evidence when the animal remains are absent. 

 The day may come when dicotyledonous plants will have been studied to 



