we think referable to this genus. It differs from T. fibratum in the following particulars: the tubes are not as uniformly foursided, nor are they arranged with equal regularity; the walls are more strongly rugose; the lamellæ appear to have been more delicate, and are generally not to be seen; traces of them however can, in most instances, be found upon close examination. The four-sided character of the tubes is sufficiently well marked to justify this reference, in connection with the fact that traces of the lamellæ can often be detected.

This species is associated with the last, and occurs, in addition, lower in the series, with Columnaria alveolata Hall. It is a common fossil in our Central Basin.

3. T. apertum Safford-Tubes isolated or fasciculated, or else united in linear series which often intersect, forming irregular reticulations; breadth of tubes about half a line; lamellæ as in T. fibratum.

This species includes certain open, loosely constructed corals which belong to this genus. Two varieties may be designated. These appear to run into each other in some specimens, though it may be found necessary hereafter to separate them.

(a) Masses composed of separate tubes occasionally united by their sides. These forms often resemble Syringopora.

(b) Masses composed of tubes arranged in linear series, the latter intersecting and forming masses like those of Halysites catenulatus Linn.

Should it be found necessary to separate these varieties, the first may be designated T. laxum and the second T. apertum.

We have observed no characters, with the exception of the open mode of growth which separate this species from T. fibra

tum.

The first variety is abundant in the middle part of the Lower Silurian series of Middle Tennessee. The second is found in the

upper half as well as near the base. same species in Kentucky.

We have observed the

4. T. minus Safford-We include in this species massive specimens, (generally small,) the tubes of which are only from 4th to d of a line in breadth. The tubes in some specimens are quite regular, in others, though generally four-sided, are more or less irregular and have the aspect on the upper surface of Chætetes. Lamellæ as in T. fibratum.

We have occasionally seen this species in the upper division of the Lower Silurian series in Middle Tennessee, as well as in Kentucky.

ART. XIX-A new Fossil Shell in the Connecticut River Sandstone; by E. HITCHCOCK, Jr.

I HAVE lately found in the coarse sandstone of Mount Tom, (Easthampton, Mass.,) a shell of a mollusk, the first I believe that has been discovered in the sandstone of the Connecticut Valley. It is preserved and not petrified, and a considerable part of it has disappeared. Enough remains however to enable us to refer it to a family if not to a genus of shells. It is represented in the annexed diagram of the natural size as it lies in

the rock. The upper part is gone, leaving an oval opening about an inch and three quarters in one diameter and an inch and one quarter in the other. It extends downwards, tapering somewhat rapidly nearly an inch and a half, and is left without a bottom, the lower opening being about an inch wide. The walls are very thick, in some places nearly half an inch, and made up of several concentric layers.

From the resemblance of this shell to a model of the lower valve of the Sphærulites calceoloides in the Cabinet of Amherst College, it seems probable that it may be referred to that family of Brachiopods denominated Rudista by Lamarck.

Its lower parts as well as the lower valve are missing, but what remains approaches nearer to the genus Sphærulites than to any other of the Rudiste of which I have seen specimens or figures.

The geological position of this fossil will be readily understood by referring to the description of Clathropteris rectiusculus

as described in vol. xx, p. 22 of this Journal. The shell is found in the same coarse grit as the Clathropteris, immediately beneath the trap (see section in the paper just referred to).

By referring to Bronn's Lethæa Geognostica, I find that the Rudista with the exception of the genera-Orbicula and Crania, are confined almost wholly to the Chalk Formation, and the shell from Mount Tom certainly comes nearer to the genus Sphærulites, Radiolites and Hippurites, than to Crania.

This specimen is too imperfect to allow of a specific or generic description, but if there be no mistake in associating it with the above genera, it seems to lend additional strength to the inference derived from the discovery of the Clathropteris, that the upper part of the Sandstone of the Connecticut Valley is as high at least as the Liassic or Jurassic series. It might seem even to carry us higher in the series, but it would be premature to draw such an inference from a single imperfect specimen, even though its true analogies be ascertained. The specimen now belongs to Amherst College Cabinet.

ART. XX.-On the Eruption at Hawaii; by Rev. TITUS COAN.*

ERE this you may have seen my letter of Nov. 16th to Mr. Lyman, giving an account of a visit to the end of the lava stream in the forests of Hilo. Since that date I have made four trips to the fire, making six in all. The great fire fountain is still in eruption, and the terminus of the stream is only about five miles from the shore. A track for horses has been cut to the fire, so that we can now ride up with ease and return in half a day. The lava moves slowly along on the surface of the ground, and at points where the quantity of lava is small, we dip it up with an iron spoon held in the hand. During the last three weeks the stream has made no progress towards Hilo, and we begin to hope that the supply at the summit fountain has diminished. There is, however, still much smoke at the terminal crater; and while the lower end of the stream is hardened for two miles above its terminus, thus checking the flow in the forest, the fusion is by hydrostatic pressure, gushing up vertically above this line, and creeping, like fiery serpents, in a thousand gory looking rills, over the smouldering masses of lava, long since deposited. These repeated and numerous up-gushings of the fusion through cracks, holes and fissures in the superincumbent masses of recently solidified lava, are caused by the sudden hardening of the end of the stream, thus obstructing the passage and causing the incandescent material, flowing under cover from regions above, to force

* From a letter to J. D. Dana, dated Hilo, March 7, 1856.

lateral outlets, or burst again to the surface by raising the superincumbent crust into ten thousand tumuli, cracking it in every direction and tilting it at every angle. In this way, the hardened stream becomes an irregularly laminated mass of unequal thickness, with a surface rolling in ridges, raised in blisters, cones, hillocks and domes, depressed into valleys, indented with pits, rent with yawning fissures, frowning with precipices, and bristling with crags. The process is somewhat like that of a superabundant quantity of water forcing its way into too small or obstructed channels under vast fields of ice; allowing, of course, for the great difference in consistency. You will understand, that the molten flood is all poured out of the fissures on the summit and for a few miles down the slope of the mountain. At first, this disgorgement flowed down and spread wide on the surface of the mountain as blood flows down a punctured limb. This phenomenon continued until the stream had swept down some thirty miles, which it did in about two days. It now came upon a plane where the angle of slope was small, say 1°. Here its progress became slow, it spread more widely, and refrigeration was more rapid. The surface, of course, hardened first. But this refrigerating process went deeper and deeper like the congelation of water, and extended higher and higher up the mountain, until at length all the lava was covered, except at occasional vents-as heretofore described-for the escape of steam and gases. Meanwhile the molten river careered unseen under the enormous mural ceiling which had been formed of its own substance, in a continuous longitudinal stream-showing itself in fiery lines, points, rills and capes, as it gushed out from under the black crust at the terminus of the stream. Here we could deliberately note its movements, as it pushed sullenly along over the rocks, through the jungle and into the mud, the pools, and water courses. The process of breaking up vertically and spreading out afresh upon the hardened crust, was occasioned by obstructions at the end of the stream, damming up the liquid, and thus obliging the accumulating lavas to force new passages and outlets for disgorgement. In this way the stream was widened by lateral outgushings, divided into several channels, swayed to the right and left, and raised to great heights by pushing up from below, and heaping mass after mass upon what had been its upper straOften when the stream had been flowing briskly and brilliantly at the end, it would suddenly harden and cool, and for several days remain inactive. At length, however, immense areas of the solidified lava, four, five or six miles above the end of the stream, are seen in motion-cones are uncapped-domes crack-hills and ridges of scoria move and clink-immense slabs of lava are raised vertically or tilted in every direction, while a

tum.

SECOND SERIES, VOL. XXII, NO. 65.-SEPT., 1856.

low, sullen crash, is heard from below, as if infernal spirits had risen to the surface of their fiery abyss and were there struggling to burst the adamantine ceiling of their prison and breathe the air of mortals. While you gaze in mute amazement, and feel the solid masses of rock-often 30, 50 or 70 feet thickmoving under your feet, the struggling lava oozes out, through ten thousand orifices and fissures, over a field of some four or five square miles. More than once have I been on such a field, and heard, and seen and felt more than is here or can be described. And yet the action of the lava is so slow-in the conditions described-that there is no fear, and little danger to one well acquainted with such phenomena. While the timid novitiate would flee for miles before such a scene, without looking back, and without consciousness of breathing, the experienced explorer will walk deliberately among the fiery pools, and rills, pry off the caps of bursting tumuli, and dip up spoils from the incandescent rocks.

When the lava becomes obstructed so that it ceases, for a time, to flow from the end of the stream, then the process which has been described takes place at some point above, and the molten mass coming up at many points, and accumulating on the surface, moves down in a superincumbent stream or streams, covering up the hardened masses below, deepening the lava, and at length reaching the terminus of the former flow, pushes on into the standing forests, and continues its progress towards Hilo perhaps a mile or so, when this hardens and stops, and at length the process is repeated. Here you see the reason why Hilo has not long since been buried.

Several large tributaries of the Wailuku-the stream which empties into our bay-are blotted out, and the water of the Wailuku is greatly reduced and rendered for the present unfit for use.

Scenes of terrible splendor have been witnessed in some of our river channels, as the molten flood moved resistlessly down, displacing the water, leaping the precipices, and lighting up the banks with immense bonfires of flaming jungle. I have witnessed two scenes of the kind of inexpressible brilliancy. One on the night of the 29th of January, and the other on the 12th of February. During the former night, the molten stream poured continuously over a precipice of 50 feet, into a deep, dry basin, half filled with flood-wood. The angle down which this fire-cataract flowed, was about 75°: the lava was divided into two, three, and sometimes four channels, from one to four yards wide, and two or three feet deep. The flow was continuous down the face of this precipice from 2 P. M. on the 19th until 10 A. M. on the 80th, when we left. During the night the immense basin under the fall was filled, the precipice converted into an inclined plane