projection about seven or eight inches long, and half an inch
wide; probably the bone of an animal, perhaps a clavicle of the
Brontozoum Giganteum.

The vestiges of fishes are very numerous in the sandstone rocks of
Connecticut River. We have not less than two dozen specimens from this
locality; a number equal to all the other specimens in our collection.
These impressions of fishes are generally from three to six inches
long, and three or four inches wide. They are of the grand division
denominated by Professor Agassiz "heterocercal," having their tails
unequally bilobed, from the partial prolongation of the dorsal spine;
and they are considered to be of lower antiquity than the fishes which
are entirely heterocercal. The most remarkable of the fish-specimens
in our collection is a CEPHALASPIS (?): this fish is found in the
specimen containing tracks of the Brontozoum Gracillimum, and traces
of a turtle or tortoise. This fossil was discovered in the upper layer
of the old red sandstone of Scotland, and had been mistaken by some
for a trilobite: to us it appeared to be a Limulus, but further
observation leads us to believe it to be a _Cephalaspis_. It exhibits
a convex disc, four inches across, by two inches from above downwards,
and a tail at right angles with the disc, the uncovered part of which
is three inches long. The animal has been described by Professor
Agassiz as being composed of a strong buckler, with a pointed horn at
either termination of the crescent, and an angular tail.

To the vegetable impressions discovered among the sandstone rocks a
peculiar name has not yet been assigned. When, however, we consider
the strong probability that many impressions of stalks, leaves,
fruits, and other parts of vegetables, may be hereafter discovered in
these rocks, it will be found convenient to have a distinctive
denomination. Vast numbers of vegetable impressions of a distinct and
beautiful appearance, and in great variety, have been found in the
coal-formation, which is nearly allied to the sandstone: such are the
Sigillaria, Stigmaria, Equisetaceae, Lycopodiaceae, Coniferae, Cycadeae,
&c. It is sufficient to say that the number of these has been already
swelled to many hundreds: we must also believe, that some of the
impressions in sandstone rocks which have been assigned to other
substances ought to be attributed to vegetables. We may, therefore,
venture to call the vegetable impressions "phytological."

A numb