1853.] DAWSON — COAL-MEASURKS, NOVA SCOTIA. 33 



heart-shaped, Uke those of S. remformis. At the hase of the trunk 

 the ribs are obliterated, and there are distant rows of double scars, 

 like those of ^S". alternans (figs. 9, 10, 11}. These differences must 

 have been produced by expansion of the trunk ; and it would thus 

 seem, that ribbed trees of this genus were adapted to a remarkably 

 regular kind of exogenous growth, the arrangement of these surface- 

 markings being, in one modification or another, preserved even in 

 the lower part of large trunks. 



From the great abundance of Stigmarice in some beds, it may be 

 inferred that the trees to which they belonged formed thick groves. 

 In other cases, however, they grew singly or sparsely on mud-flats. 

 They could flourish on a variety of soils, from poor sand to the richest 

 vegetable mould, but in all cases they seem to have required a 

 swampy or water-soaked locality. In no instance did we find leaves 

 attached to the trunks of Sigillaria. In some of the beds, however, 

 the trunks are associated with great numbers of leaves of Poacltes, 

 and slender grass-like or needle-like leaves like those oi Lepidodendron 

 and Cyperites. 



The preservation of so many trunks of Sigillaria in an erect 

 position is in part due to the very durable nature of their bark as 

 compared with their woody portion. This is well illustrated by the 

 occurrence of an inclined trunk in one of the sandstones, and a 

 prostrate one in the roof of the main coal, with shells of Spii'orbis 

 attached to the inner side of the bark, showing that the bark 

 remained for some time under water and periectly hollow, without 

 losing its form. In both these instances the species is a narrow- 

 ribbed one, resembling S. Organum. This difference in the durability 

 of the bark and wood may be observed in many modern trees ; for 

 example the canoe birch, Betula jyapyracea, the tough bark of 

 which often remains as a hollow cylinder long after the disappearance 

 of the wood, and may sometimes be seen in swamps, filled with 

 mud, while still apparently sound and fresh. When the casts of 

 erect trees are of hard sandstone, they sometimes project from the 

 cliif like columns or pilasters; but where they are filled with soft 

 material, they waste with the surface of the cliff, and but for some 

 slight difference of colour, and the coaly bark passing up at each 

 side, across the beds, might easily escape observation. I may remark 

 here that the section includes the erect trees observed both in 18.52 

 (by Sir C. Lyell and myself) and 1853 (by myself alone). Many of 

 them were not seen in 1852, and others seen in 1852 had quite dis- 

 appeared before the summer of 1853. I have no doubt that a series 

 of observations continued through several years would greatly increase 

 the number of beds with erect plants. 



Among the most common species oi Sigillaria at the South Joggins 

 are S. Organum, catenulata, reniformis, scutellata, and alternans, or 

 species closely resembling these. Prostrate trunks of three species 

 of Favularia, of two species of Ulodendron, apparently the U. majus 

 and minus, and of three or four species of Lepidodendron are found 

 along with the Sigillaria, and probably grew in the same swamps. 

 We had no means of ascertaining whether the trees of these last- 



VOL. X. — PART I. D 



