624 
DR. J. W. DAWSON ON ERECT TREES CONTAINING ANIMAL 
to the encroachments of the sea. East of this the coast recedes, still presenting a 
high cliff in which the beds crop out one by one to the surface. In front of the point 
the great sandstone, No. 1, and the minor bed, No. 8, extend in reefs seaward, and 
owing to the great rise and fall of the tide in Cumberland Bay a length of about 
330 yards of these reefs is exposed at low tide (see map, section, and view of the coast, 
Plates 46 and 47). 
From information obtained by Henry Poole, Esq., it would appear that the cliff 
recedes at the rate of about 25 feet in 40 years, so that trees are rarely exposed in the 
bank, and those existing on the reef, and which are The most accessible, represent the 
effects of denudation extending over at least five centuries. 
The relative positions of all the trees observed are seen in the plan and section, from 
which it would appear that they occur in groups. There is reason to believe, however, 
that these erect trees are only survivors of a much more dense forest, of which the 
weaker and more perishable trees had been overthrown. The positions of the trees 
extracted before 1878 could only be indicated approximately. 
The manner of accumulation of the beds enclosing the erect trees is rendered evident 
by their character and contents, and has been noticed in my ‘Acadian Geology,’ p. 190, 
et seq. The details ascertained by the recent excavations may be stated as follows : - - 
The underclay, No. 13, represents a loamy soil on which Sigillcirice must have 
flourished for a long time, filling it with their roots and rootlets. The result was the 
accumulation of the coal, No. 12, which is filled with flattened and carbonised bark of 
these trees, as is the black shale constituting its roof, Bark of Lepidojloyos k is however 
largely associated with that of Sigillctria in these beds. Leaves of Cordaites also 
occur in this coal, and its mineral charcoal shows under the microscope bast fibres of 
the inner bark of Sigillctria, with scalariform, uniporous, and reticulated tissues, 
probably belonging to the wood of Sigillctria, Lepidojloyos, CoJctmites, and Cordaites. t 
Leaves of the latter genus, as I have shown (op. cit.), constitute a large part of some 
thin coals at the Joggins, and some portion of the ligneous matter, which in former 
papers I have referred to other genera, may, since the discoveries of Grynd-’Eury, 
possibly be referred to Cordaites. The more compact portion of the coal when sliced 
shows shreds of epidermal tissue with a few rounded bodies, probably spores of ferns 
or iycopods.J 
* TJlodendron of some English, palseobotanists. 
f ‘Acadian Geology,’ p. 168; and paper on “ Coal Accumulation,” Jour. Geo. Soc. 
+ I cannot admit that the large trunks of silicified and calcified wood of the genus Dadoxylon 
(Araucafioxylori) , so abundant in the coal-formation of Nova Scotia, belonged to Cordaites. Their foliage 
is more probably represented by the leafy twigs of Walchia or Arauca/ntes found with them. It is also 
true that some ribbed trees with the markings of Sigillana have wood of the structure attributed by 
Gbvnd-’Euey and Renault to Cordaites. Where the dividing lines between Sigillaria, Cordaites, and 
JDadoxylon will finally be fixed remains somewhat uncertain. As I have elsewhere argued, however, it is 
evident that under the names Sigillana and Cordaites are included cryptogamous and gymnospermous 
trees of very different grades. 
