EXPLANATION OF JOGGINS SECTION. 191 



the longitiulinal section sliows only elongated cells, but is very badly 

 preserved. A tree of this description is not likely to have been more 

 perishable than the S/gillarice, which, in the same situation, remained 

 until nine feet of sandy mud had accumulated. I suspect, therefore, 

 that this stump may be the remains of a Coniferous forest, which pre- 

 ceded the Sigillarice in this locality, and of which only decaying 

 stumps remained at the time when tlie latter were buried by sediment. 

 This is the more likely, as the appearances indicate that this tree was 

 in a complete state of decay at the very commencement of the sandy 

 deposit. It is, however, possible that this older forest, represented by 

 coaly stumps, may have consisted of Sigillaroid trees. 



The history of this group will thus be as follows: — (1.) The Stig- 

 maria underclay shows the existence of a Sigillaria forest, on the soil 

 of which was collected sufficient vegetable matter to form six inches 

 of coal, which probably represents a peaty bog several feet in thickness. 

 (2.) On this peaty soil grew the trees represented by the stump of 

 mineral charcoal mentioned above. This tree, being about one foot 

 in diameter, may have required about fifty years for its growth to that 

 size. It was then killed, perhaps by the inundation of the bog. (3.) 

 During the decay of the tree last mentioned, Sigillarice grew around 

 it to the diameter of two feet, when they were overwhelmed by 

 sediment, which buried their roots to the depth of about eighteen 

 inches. At this level Calamites and other Sigillarice began to grow, 

 the former attaining a diameter of four inches, the latter a diameter 

 of about one foot. (4.) These plants were in their turn embedded in 

 somewhat coarser sediment, but so gradually that trees with stigmarian 

 roots grew at two higher levels before the accumulation of mud and 

 sand attained the depth of nine feet, at which depth the original large 

 Sigillarice that had grown immediately over the coal were broken 

 off, and their hollow trunks filled with sand. Before being filled with 

 sand, these trees, while hollow, must for some time have projected 

 from a swamp or terrestrial surface, such as that which immediately 

 succeeds them in ascending order ; and it is no doubt to this circum- 

 stance that we owe the occurrence, in some of them, of reptilian 

 remains and land shells, as well as many vegetable fragments, such 

 as Calamites^ C'urdaites, and a Lepidostrolus, with many of the fossil 

 fruits called Trigonocarpa, evidently introduced before the sedimen- 

 tary matter, and forming just such a mass as might be supposed likely 

 to fall into an open hole in a forest or swamp. (5.) The remaining 

 beds of this group evidence the continuation of swamp conditions for 

 a long time after the trees last noticed were completely buried. They 

 include, in a tliickness of twenty-eight feet, three underelays supporting 



