ishes of Sabine and Natchitoches, where limestone and sandstone ridges also exist, is a question still open. In the latter case, this outline would conform to the general shore lines of the great cretaceous and tertiary Mediterranean,

In Mississippi, the Grand Gulf series is mostly overlaid by the Orange Sand, deposited on a deeply eroded surface, and bearing itself the evidence of its formation by fresh water in a state of violent flow. The southern outline of the main body of the Orange Sand runs southward of Opelousas, toward the mouth of the Sabine, whence, according to reliable information, a broad band of shingle extends toward Harrisonburg, Catahoula parish. This belt represents, probably, the most westerly bayou of the great Orange Sand Delta; while, as heretofore stated, the most easterly one extends from the neighborhood of Cairo along the western shore of the Tennessee river, down the valley of the Warrior toward the coast of Alabama. The middle and main pebble-stream evidently follows in general the course of the Mississippi river; but leaving it at the point where that river suffers its remarkable deflection eastward, we find the remnants of its ancient "bar" in the chain of the "Fire Islands," which lie directly across the shortest line by which the Mississippi could reach the Gulf, and no doubt have had their share in causing this deflection.

Both the size of the pebbles carried by this middle bayou, and their character proving transportation from high northern latitudes, show it to have been the main channel during the Orange Sand epoch. It is not surprising, therefore, that in the direction of its course the Orange Sand formation should extend farther south than anywhere else. The pebble-beds are now overlaid by fine sandy materials, proving a diminished velocity, owing, doubtless, to a general depression, but greater at the north than at the south.

While the lateral bayous descending through Louisiana and Alabama were closed at the end of the Orange Sand epoch, it is evident that the central channel continued open; inasmuch as the next succeeding deposit, viz: the Loess, lies in a troughshaped depression of the Orange Sand materials, the line of contact being always conformable and devoid of any trace of atmospheric denudation. The perfect peroxydation of the materials of the Orange Sand would seem, nevertheless, to point to a certain period of exposure to atmospheric agencies, caused by a temporary diminution of the influx of northern waters, through the cessation of subsidence, perhaps.

During this epoch of quiet might have begun the formation of those extensive swamp and lagoon deposits, the lower mem

* Am. Jour. Science, May, 1866; Miss. Rep. 1860, p. 26 and ff.

bers of the Port Hudson series, whose floor stratum, with its superimposed generations of cypress stumps, indicate a slow secular subsidence. The velocity of the latter seems gradually to have increased until the growth of old trees became impossible, and finally, in stratum No. 3 of the Port Hudson profile, we again meet the evidences of currents moving sand, pebbles, and drift-wood.

Then follows the Loess proper, a deposit utterly devoid, in Mississippi and Louisiana, of any evidences of fluviatile action -a uniform silt even in profiles of 80 feet, with scarcely a vestige of stratification, and none but terrestrial fossils.

The precise circumstances under which such a deposit could be formed, are perhaps a little obscure. There must have been such a depression of the whole country as to transform the immediate valley of the Mississippi, as far as Keokuk, as well as the valleys of the larger tributaries, into estuaries of the Gulf of Mexico, containing a mass of water too great to be sensibly affected by the variations now causing the annual overflows of those rivers (for otherwise the deposits must have shown lines of deposition), yet possessing a gentle flow above (since the materials of the bluff formation of Missouri and Indiana exhibit signs of fluviatile action); quite fresh in its upper portions (where fluviatile shells are found), but rendered unfit for the life of either a fresh or salt-water fauna by an admixture of sea-water, in its lower and almost stagnant portion, at tide level; and deriving its vestiges of animal life only from the "offscourings" of the adjoining unsubmerged lands. Sir Chas. Lyell inclines to consider the Loess as the product of "successive inundations of a great river," the absence of stratification from such deposits having, apparently, an analogue in the alluvial deposits of the Nile. But the case is far from being analogous ; for the same phenomena are still observed in the modern deposits of the Nile, and are clearly attributable to the peculiarities of the hydrographic basin of that river; whereas, in the modern alluvium of the Mississippi, it is exceedingly difficult to find a uniform stratum two feet in thickness. The Nile mud is each year derived from the same rivers of Abyssinia, and equalized by intermixture and subsidence during at least 1,500 miles of its course. the Mississippi, on the contrary, the deposits of different annual inundations are readily distinguished by the inhabitants for years afterward, according as the Illinois, the Missouri, Ohio, Arkansas or Red river happen to have furnished the main influx. The absence of any such differences from the Loess can only be explained on the assumption that the mass of

* Principles of Geology, 10th edition, p. 464.

On

water filling the channels was too great to be sensibly affected by such causes, the more so as the continental surface was sensibly diminished in consequence of a depression which, as far south as Fort Adams, cannot have been less than 400 feet, and on the coast not less than 200, but more probably the same as farther above.

The existence of the elevations on the Louisiana coast, above described, renders it necessary to assume that at the end of the period of depression-the "Champlain epoch "-the entire delta-plain (so-called) west of the Mississippi was covered by the deposits of the Orange Sand and Port Hudson series to an equal height; and that during the succeeding "Terrace epoch" of elevation, the veritable Mississippi-our Mississippi-swept away these deposits in excavating its present valley. At first it might sweep over or through the pebble ridge, but would finally turn to the direction of least resistance, leaving the "Five Islands" high and dry.

It would thus seem that, unlike other large rivers of the world which have from the outset added to the land by bringing down the materials to form their alluvial plain, the Mississippi has first formed by denudation the plain which it was subsequently to cover with its alluvial deposits to a comparatively inconsiderable depth. The western and southern limits of this denuding action would seem to be marked by the Grand Côteau des Opelousas, the Côte Gêlée and the Five Islands; and the materials swept away from this area doubtless contributed largely to form the foundations of the truly alluvial plain extending south and southeastward of lakes Maurepas, Pontchartrain and Borgne.

It is obvious how futile must be all attempts to estimate the age of the Mississippi river in absolute measure, by a comparison of the advance of its present delta into the Gulf, with the distance of its mouth from the divergence of bayous Plaquemine or Manchac. When the broad flood of the Terrace epoch contracted into the present Mississippi, that stream emptied into a sea rendered shallow by the deposition, within a comparatively short period, of a huge amount of material. Within such a sea its channel would be likely to change about, somewhat like those of the great rivers of China. Now, it is advancing into the deep water of the Gulf of Mexico, but at a very different rate, and by a very different process from that of simple alluvion. But the questions pertaining to this portion of the subject, together with the results of my observations in the Delta proper, I propose to discuss at a future time.

ART. VIII.-Notes on the recent volcanic disturbances of Hawaii; by Rev. TITUS COAN;* (from a letter to J. D. Dana, dated Sept. 1, 1868).

I HAVE recently returned from a tour of eighteen days in Puna and Kau, during which time I visited all the important points of interest, took measurements, and made careful observations. I take the liberty to give you a few facts and remarks from my notes.

I left Hilo on the 4th of August and took the shore road through Puna. From Hilo to the east cape, Kapoho, the disturbances were not remarkable. Walls and rocks had been, more or less, shaken down, and fractures appeared, here and there, in the earth. From the east cape to the south point of the island, Kalae, the disturbance had been great. The whole coast line appears to have undergone a subsidence of unequal depth. At Keahialaka, about seven miles southwest of Kapoho, where there has been a small pool of brackish water, passed by a causeway of stones, the water now stands, at high tide, three to four feet deep, and spreads out among cocoa-nut groves where water was never seen before. At Kaimu, the sea has driven in a heavy beach of lava sand for two hundred feet into a beautiful young cocoa-nut grove, and spoiled a delightful lawn. At Kalapana, the sand beach has been forced into groves of Pandanus and cocoa-nut, where the trees now stand six to eight feet deep in sand, and the shore line has been pressed back the distance of 100 feet. The old stone church is nearly buried in sand, and the tide rises and falls within its walls. t formerly stood 200 feet from the water. The plain of Kala* Mr Coan also sends us the following by way of correction of our newspaper extracts in the July number of this Journal: "Article XIII of the Journal of Science for July has just come to hand. I regret that several important errors have been published and very widely circulated. But this could hardly be avoided in times of such terrible excitement, when exaggerated reports filled the air, and the honest eye and ear sometimes failed to report the exact truth. I am surprised to find myself reporting 2,500 to 2,6001 head of cattle destroyed. Did I state that? If so, I stated just what was reported to me. The truth now appears to be, that on Kahuka, where 200 cattle were said to have been roasted in a holocaust, by a puff of Pele, only 35 perished, as I learn from the owner. And on the Kapapala ranch, one of the partners sets the number at 1,500, the other at 500, while some of the neighbors would reduce the loss to 200. But on a ranch of 5,000 cattle, spread over all the southeast and eastern flanks of Mauna Loa, it is impossible to tell the exact number destroyed by the landslide. The statements, also, that a vast lava stream went, under ground, into the sea at Punaluu, and that another broke out at Hiilea, were optical illusions of Mr. Lyman and many others. They spoke and wrote what, at a distance, appeared to them as facts, but which proved to be errors on careful inspection. So also the report, that the sea-waves rolled over the tops of cocoa-nut trees, sixty feet high, was an error by two-thirds.

'In my letter published in the Missionary Herald for July, you will see that I. say 500 to 600 head of cattle, which is nearer the truth."

pana has sunk four to six feet, and many acres of once dry land are now covered with water three and four feet deep. Bathing coves, also, once having spaces of three, four and six feet between the water and the roof, are now full or nearly so. At Kealakomo, the salt works are all destroyed, and the spring of cold water among the rocks near the shore is sunk. At Apua, the most western village in Puna, all the houses, the crescent sandbeach, and the beautiful little canoe harbor are obliteratedthe fishermen mourn. Water now stands where the village once stood. The same is true of Keauhou, the most eastern village of Kau. The subsidence there is seven feet by measurement upon cocoa-nut trees. This place has an ample and safe anchorage, and here was the station where large quantities of pulu were dried, pressed and shipped for market. The influx of the sea destroyed the buildings, but the proprietors have built again on higher rocks.

A, Apua. H, Honuapu. Ka, Kaaluala. Ke, Kalae. Kh, Kapoho. Kk, Kahuka. Kl, Kealakeakua. Ko, Kealakomo. Kp, Kalapana. Kw, Keaiwa. N, Ninole. Po, Pohue. Pu, Punaluu. W, Waiohinu.

In passing from Puna to Kau we took the high trail from Panau to Kapapala, a distance of about thirty miles. This whole route is a dreary wilderness, without a human habitation, a drop of water, or any thing to refresh man or beast. The track at first lies over pahoehoe, sand and aa, sprinkled with small trees and bushes; but the latter half of the way is upon open fields of lava and sand, and under a burning sun, with nothing to break his fierce rays. The road flanks Kilauea at a