1895.] NATURAL SCIENCES OF PHILADELPHIA. 155 



ness of sands and clays were laid down. Another point not satis- 

 factorily explained by this theory is the general homogeneity of 

 structure exhibited by these beds and their freedom from interlami- 

 nations of or interstitial clays. Vegetable deposits due to drift ma- 

 terial would naturally contain more or less clay intermixed with the 

 deposit. And again we can hardly suppose the sea to have been 

 devoid of either vegetable or animal life. 



The theory that these lignite beds were formed from materials 

 growing in situ while probably it does not explain all the difficulties 

 attached to the question comes much nearer doing so than any other. 

 As far back as 1828 A. Bronghiart held the opinion that coal was 

 formed of plants growing in situ and in 1853 Le Conte taught that 

 coal was formed as in the peat swamps of the present day, but these 

 swamps occurred at the mouths of large rivers and were subject to 

 overflows by the rivers and occasional inundations by the sea. 82 The 

 absence of river mud he ascribes to a straining operation of the 

 plants along the margins of the swamp and quotes Mr. Lyell as the 

 authority for the statement that "although the peat swarnps of the 

 Mississippi are annually flooded by river water they are entirely un- 

 touched by river mud. These favored spots are surrounded, par- 

 tially on the side next the river, by dense vegetation, which, acting 

 as a sieve, completely strains the water of its mud before it reaches 

 the peat swamp. The water of these swamps is, therefore, pure, 

 and pure peat has been quietly depositing there for ages. 



According to Giimbel coal should be considered an inland deposit 

 formed in wide, flat depressions of continents and also on low grounds 

 along the sea coast. Undisturbed growth of marsh vegetation 

 alternated with floods during a long continued subsidence might reach 

 locally a thickness of several thousand yards of successive beds of 

 coal, sandstone and shale. 83 



As already suggested the sea had reached to a considerable dis- 

 tance south of the lignitic areas and a broad but low sand bar, 

 broken in places, protected the sandy plains from marine invasion. 

 Erosion had begun work on this sandy waste and the different streams 

 were being gathered into the courses they were subsequently to 

 occupy. This cause of destruction operated rapidly, as it does now, 



82 Smithsonian Institution Report of 1853, pp. 136-137. 



83 Beitrage zur Kenntniss der Texturverhiilt d. Mineralkohlen, 1883. 



