garding this point are given, but from the great breadth of country upon which outcrops of one and the same group of the Ter tiary occur in Alabama, the fact is apparent enough as concerns the latter. Whether the same is true of any part of the Cretaceous, is doubtful; unless the great north and south width of the Ripley group, as exhibited on Chunnenugga ridge in Macon and Barbour counties (according to Tuomey's and Thornton's observations combined), should thus find its explanation. It seems doubtful, in fact, whether the true Rotten Limestone (if it exist there) comes to the surface at all. This is the more remarkable from the near proximity of the primary and metamorphic rocks of that portion of the state, whose original upthrusting would thus be proved to ante-date greatly the general Allegheny up

heaval.

The Grand Gulf group.-I have considered the older and welldefined eocene Tertiary apart from the two other groups described in my Report, whose age is doubtful and whose relation to the former is not well recognized. A glance at the map. nevertheless shows that so far as extent is concerned, the Grand Gulf group is perhaps the most important of the formations of the state of Mississippi, and that, judging by the trend of its outlines on the Mississippi river, it must be still more so in Louisiana; while in Alabama it rapidly contracts, and attracts so little attention that I find but two observers who, passim, advert to anything resembling this formation as it exists in Mississippi.

Conrad (this Jour., [2], vol. ii, 210) states that the bluffs of Vicksburg, Grand Gulf, Rodney and Natchez, have a similar geological origin; that their lower portion is of marine origin, and a member of the Eocene.

I am unable to refer to a prior publication, mentioned by Conrad, for the data upon which this determination is based, so far as the bluffs below Vicksburg are concerned. I have made detailed examinations of the profiles at Grand Gulf and at Fort Adams, at the extreme limits of the formation in Mississippi, and I may say, of all the important outcrops in the interior; but thus far, have failed to find even a trace of a marine fossil, and in fact, but a single specimen-a bone fragment as I take itlikely to prove of zoogene origin. Vestiges of vegetation are common, but only in one instance, so far, have I found any specimens likely to admit of exact determination. I refer to the deposit on the Chickasawhay, already referred to, which exhibits the trunks, stumps and roots of an ancient forest, inhabited, among other trees, by tree palms. But even here, scarcely any. thing beyond the most general outlines of a few leaves can be traced. It may be that in proximity to the (rare) lignite beds of this formation, better success might be had-as has been the case in the Lower Lignitic. In the sand- and claystones belonging

to this group, neither Wailes, who resided amongst them and gave the name of "Davion rock" to one variety of the former; nor myself who have delved in scores of exposures, have ever found a trace of any fossil whatsoever.

The Natchez bluff I have not visited; but Wailes, who resided within six miles of it, must have done so, and he is silent on the subject of any but the Loess fossils, although he mentions all other fossiliferous rocks occurring in the State. Thus, while I have seen mentioned in various places "marine strata at the foot of Natchez bluff," I cannot trace the report to any authentic source. I shall endeavor to settle the point as soon as possible, but meanwhile observe, that according to reliable information given me, the Rodney bluff is essentially a counterpart of that at Grand Gulf; a detailed profile of which, obtained at a medium stage of the river, is given in my Report (p. 148).

The extreme scarcity of fossils in this formation is the more remarkable, as from the regularity of its stratification it is manifest that it has been formed in quiet water, and it contains a great variety of materials suitable for the preservation of either fauna or flora. Even the strata containing carbonate of lime, however, seem to have had nothing to fossilize, save in the solitary instance of a doubtful fragment of cellular bone already mentioned. In some portions of it, we might imagine that the abundance of soluble salts (which pervade more or less the entire deposit) indicated the former existence of bitter lakes, incapable of harboring life; but this could by no means apply to the formation as a whole.

The only probable presumption in favor of referring it to the Eocene, so far as I know, arises from the lithological resemblance and transition of its strata, at its northern limit, to those of the Vicksburg group. The upper division of the latter group in the neighborhood of Brandon is undistinguishable from the materials of the Grand Gulf group at many points, and I so referred them until I found them overlaid by a string of limestone nodules containing Orbitoides, about forty-five feet above. the uppermost sands of the Vicksburg group. (See section, Miss. Rep., p. 140.) At a level about forty feet higher, the characteristic soft white sandstone of the Grand Gulf group crops out.

This, however, amounts to mere conjecture; and, per contra, toward the sea-coast the lithological transition into the materials of the "Coast pliocene" seems about equally cogent. The mere fact that tree palms are found in the formation, amounts to nothing, inasmuch as these grow at the present time in the same latitude in South Carolina.

The existence of this formation in Alabama appears from Mr. Thornton's notes (2d Rep. Ala., Appendix), in which he mentions similar materials as overlying the (Vicksburg) marine Ter

tiary at its southern limits. Moreover, Bigelow (this Jour., ii, 419) describes a sandstone formation in Baldwin county, Ala., which impresses me as though it might be Grand Gulf sandstone overlaid by the ferruginous sandstone with tubes, of the Orange Sand group (Miss. Rep., p. 9), "filled with variously colored sand." Bigelow states that in the lowest portion of this rock he has seen obscure impressions of shells; which, if my conjecture be correct, might offer an opportunity of determining the age of the group. The same rock is said to occur at Pensacola.

The "blue clay bottom" of the CoastCoast Pleiocene."— Finally, as regards the imperfectly known black clay formation of the coast (Miss. Rep., p. 154, ff.), a comparison of specimens of shells and borings obtained from the New Orleans artesian well,* in 1854, seems to show that it underlies the whole of the delta, perhaps as high up as Port Hudson, whose subterranean Cypress swamps, observed by Carpenter and by Lyell, may belong to this formation. No Eocene fossils have been brought up by the augur, even from the lowest shell-bed found, at the depth of 570 feet, (the greatest depth was 630); while from among the shells of the first bed struck, at 41 feet, I have thus far determined eighteen marine species, all now living in the Gulf. At 153 feet a trunk of cypress, with bark, was found. At 256 feet, some extinct, or if living, undescribed shells seem to occur; and at 480 a Gnathodon bed. I hope to be able to determine by microscopic comparison whether or not the Grand Gulf group has been passed through or reached in this bore, which from present appearances has penetrated both Post-pleiocene and Pleiocene marine deposits. This would parallelize more closely the Tertiary of the Atlantic coast and of the Gulf; though so far as I know, nothing apparently corresponding to the Grand Gulf era has been observed in the former series. Should the chain of the Antilles, after the close of the Eocene epoch have for some time cut off the Gulf of Mexico from the Atlantic, it seems possible that the deposits of the former might have changed their character to the extent required by the facts observed. A strong influx of fresh water-perhaps that pertaining to the Great Lignite era-from the continent might for the time being have extinguished the Eocene marine fauna without replacing it by another sufficiently numerous to be readily detected in the deposits of the period, which might thus correspond to the Atlan tic Miocene. Upon the subsequent irruption of the Gulf stream through the Antilles chain, the formation of normal marine deposits along the margin of the Gulf would be resumed.

University of Mississippi, July 26, 1866.

These specimens were furnished by Drs. Copes and S. S. Riddell, of New Orleans, to Maj. Gen. A. A. Humphreys, and by him referred to me for examination; which, however, is not yet completed.

AM. JOUR. SCI.-SECOND SERIES, VOL. XLIII, No. 127.—JAN., 1867.

Art. VI.-Evidences of the existence of ancient Local Glaciers in the White Mountain Valleys; by A. S. PACKARD, Jr., M.D.

THE following observations were made during the past autumn in the valleys of the tributaries of the Saco, and Androscoggin rivers. At Jackson, N. H., on Thorn mountain, which lies just south of Tin mountain, there are some well marked glacial scratches which point directly toward Mt. Washington, which stands at the head of the valley of the Ellis river; their course being N. 25° W. These were first noticed on quartz veins running over the ledges which have been polished smooth as porcelain and finely grooved. At other places on the same mountain part way up, and also upon the summit, upon removing the soil, similar striæ occurred running in the same direction. On this mountain and the neighboring hills occurred occasional boulders of a peculiar mica slate, enclosing crystals of staurotide, which had evidently been transported from near the summit of Mount Washington. The summit of Mt. Kearsage we found moulded by ice. Dr. C. T. Jackson in his report on the Geology of New Hampshire states that the drift scratches one half way up Mount Kearsage run N. 30° W. He also states that on Mount Chicorua they run N. 35° W. (S. 35° E.), which is the course of the Ossipee valley just below it.

On a hill just east of Goodrich's falls on the Ellis river are very distinct ice-marks, on polished surfaces, with striæ running N. 30° W., and lunoid furrows with their horns pointing up the valley in the same general direction as the grooves.

Crossing over the mountains into Chatham, and Stowe, Maine, into the valley of the Cold river, another tributary of the Saco, we find another set of striæ. The broad summit of Speckled mountain, opposite Mt. Royce, which two mountains guard the southern entrance of Evans' Notch, is glaciated both on the N.W. and N.E. flanks. Here also is a "col," down which the ice must have moved in both directions. Near the summit the grooves and lunoid furrows run N. 15° E., following the course of the valley at this place, and aiming at a higher peak to the north and east. On Mt. Baldface, 3600 feet high, three or four miles southward, the grooves are very clearly indicated both below and directly upon the summit. Here they run N. 10° W., and it might be mentioned that the Cold river valley turns more to the southeast at this point. On a shoulder of the mountain, perhaps 300 feet below the summit, the lunoid furrows are especially abundant.

On the summit of this mountain, which is made up of a light colored fine syenite, were a few boulders of a peculiar porphyritic syenite, with oblong crystals of albite. Following the N. 10° W. course, less than a quarter of a mile, we traced them to the

parent rock composing Peaked mountain, which is somewhat lower than Baldface, at least 100 feet.

Again, crossing the high range of mountains over into Gilead in the Androscoggin valley, glacial marks directed N.W. occurred on a high ledge near the river, indicating that the ice moved from the northwest, pursuing the general course of the valley at this point.

Thus

Here, then, are good proofs of distinct systems of glaciers radiating from a central mer de glace which capped the White mountains. This dome of ice must, so far as our slight observa. tions show, have been soon subdivided into local glaciers which pursued their route down the different valleys to the sea. following down the Androscoggin river, at Lewiston, the icemarks run nearly north and south, the course of the valley at that place, as we are informed by Mr. G. J. Varney of that town; and at Brunswick, on the seashore, there are deep furrows running in a N.W. direction, being the ancient course of the river where it undoubtedly entered the sea, up to a late period of the Terrace epoch.

ART. VII.-Experiments on the Electro-motive Force and the Resistance of a Galvanic Circuit; by HERMANN HAUG.

[Concluded from vol. xlii, p. 889.]

NATURALLY I first inquired whether or not the experiments of other observers would exhibit peculiarities of a similar character. I am not fortunate enough now to have a great choice of material at hand. But the experiments of J. Müller, cited on page 384, with six cells of Daniell's, if every possible combination of two different intensities is calculated, show decidedly a similar great increase of the internal resistance with the decrease of the combined intensities, viz:

As I have every reason to believe that in these experiments the circuit was really opened every time a greater resistance was to be introduced, I considered this circumstance, or the reverse, as not of any great account, and tried to determine at least the general character of its influence upon the results of all the other circumstances which may be regarded as important.