In the lowest of these three horizons the shells are found, as already stated, in a thin bed of concretionary clay of dark grey colour, though associated with reddish beds. It contains _Zonites priscus_ as well, though this is very rare, and there are a few valves of _Cythere_ and shells of _Naiadites_ as well as carbonaceous fragments, fronds of ferns, _Trigonocarpa_, etc. The _Pupæ_ are mostly adult, but many very young shells also occur, as well as fragments of broken shells. The bed is evidently a layer of mud deposited in a pond or creek, or at the mouth of a small stream. In modern swamps multitudes of fresh-water shells occur in such places, and it is remarkable that in this case the only Gasteropods are land shells, and these very plentiful, though only in one bed about an inch in thickness. This would seem to imply an absence of fresh-water Pulmonifera. In the erect _Sigillariæ_ of the second horizon the shells occur either in a sandy matrix, more or less darkened with vegetable matter, or in a carbonaceous mass composed mainly of vegetable _débris_. Except when crushed or flattened, the shells in these repositories are usually filled with brownish calcite. From this I infer that most of them were alive when imbedded, or at least that they contained the bodies of the animals; and it is not improbable that they sheltered themselves in the hollow trees, as is the habit of many similar animals in modern forests. Their residence in these trees, as well as the characters of their embryology, are illustrated by the occurrence of their mature ova. One of those, which I have considered worth figuring, has been broken in such a way as to show the embryo shell. They may also have formed part of the food of the reptilian animals whose remains occur with them. In illustration of this I have elsewhere stated that I have found as many as eleven unbroken shells of _Physa heterostropha_ in the stomach of a modern _Menobranchus_. I think it certain, however, that both the shells and the reptiles occurring in these trees must have been strictly terrestrial in their habits, as they could not have found admission to the erect trees unless the ground had been sufficiently dry to allow several feet of the imbedded hollow trunks to be free from water. In the highest of the three horizons the shells occurred in an erect tree, but without any other fossils, and they had apparently been washed in along with a greyish mud.[138] [138] The discovery of the shells in this tree was made by Albert I. Hill, C.E. The tree is in Group XXVI. of Division 4 of my Joggins section. The original reptiliferous trees are in Group XV., and the lowest bed in Group VIII.