Mode and Period of the Production of the Trap Tufa. Such are the facts in regard to this formation. I now come to the mode and period of its production.

Two positions in relation to this subject, I think will be easy to prove. The first is, that this rock must have been of contemporaneous production with the sandstone in which its beds are interstratified. The very fact of their being interstratified, shows that they could not have been subsequently injected. They are divided into layers, if not as thin, yet often as distinct, as the sandstone. Consequently, after the deposition of the sandstone on which they rest, they must in some similar manner have been deposited, before the layers of sandstone, now above them, were formed over them.

The second position is, that these rocks must have been the joint product of igneous and aqueous agency, The stratification and mechanical structure of some of the varieties clearly prove the agency of water, while the vesicular and concretionary structure of other varieties, and the entire resemblance of others still to unstratified trap, where the deposit is thick, as well as the volcanic character of the cement, even of the most decidedly mechanical layers, are equally conclusive proofs of the agency of heat. Indeed there is no part of the trap formation in this valley that presents so decidedly a volcanic aspect, as some of the weathered and disintegrated hummocks of this rock. It is impossible, then, to explain the production of this rock without calling in the joint action of fire and water. But how and when did this joint action take place? I adopt, almost without modification, the views advanced by Mr. Murchison, to account for the origin of similar rocks, called by him volcanic grits, and "bedded and contemporaneous trap," connected with the older fossiliferous rocks of Great Britain. He supposes them to have resulted from early and minor volcanic outbursts at the bottom of the ocean, at different intervals, previous to the period when the principal ranges of amorphous trap were protruded along the same lines. These preparatory eruptions would of course throw out abundance of scoriaceous matter, finely levigated, with some melted matter, which would spread over the bottom of the sea, and mix with the sand, gravel, and mud, there accumulated; so as to form the tufaceous conglomerates above described: while, if the quantity of liquid matter was large, a portion of it might cool slowly, and neither mix with the sand and gravel, nor come in contact with the water, except at the surface and there the sudden refrigeration and commotion produced by the vapor and gases, would form scoriaceous and amygdaloidal rock. If in the bottom of the ocean this mixture of ashes, scoria, and lava, should envelope the remains of animals or plants, the heat might not be

sufficient to destroy all their organic structure, yet great enough to convert them into a scoriaceous mass; or if the organic matter had been driven off, to fill up the space with such matter rendered vesicular, (as in the particular case which I have described,) by the gas resulting from the decomposition of the organic compounds. Thus we might have organic remains in nearly melted rock, and thus might volcanic matter be made to take a stratified structure; but when quiet was again restored on the ocean's bed, new deposits of sandstone and shale would take place. After all this, the principal eruption of trappean matter might take place along nearly the same line, and the principal ridges of unstratified trap be protruded, and the sandstone and bedded trap, at least in the vicinity, be tilted up.

Now in respect to the sandstone and tufaceous rock of the valley of the Connecticut, we have the most decided evidence that those parts of these strata near the principal amorphous ridges of trap, have been elevated since their deposition. I have stated that the general dip of the sandstone in the valley is 15° or 20° easterly, both below and above (or west and east of) the trap ridges; and the fact that it is nearly the same below as above, shows conclusively that this general dip was not produced by the protrusion of the greenstone: for how could that elevate strata lying beneath it? It must, therefore, have been originally deposited with its present dip, or elevated by some other agency than the trap. But on the upper side of the trap ridges, we find numerous examples where the strata have been elevated very much more than the general dip; which greater slope, however, dies away as we recede from the trap. Take for an example, the section given on fig. 2; which crosses Mount Holyoke near its east end from Amherst southeasterly into Belchertown. Beneath the trap the dip is as usual; but on passing over the ridge, we find the strata dipping pretty uniformly as much as 50°, and running from N.E. to S.W. as if turned aside 45° from the usual course. The deposit of trap tufa, which is crossed by the section, shares in the increased dip. Now it is not possible that sandstone, most of it rather coarse, should have been deposited on a slope of 50°. It must have been tilted up after its deposition, and consequently after the deposition of the tufaceous rock. At the other beds of this rock, the dip is not much greater on the upper than on the lower side of the principal trap range. But along the south side of Holyoke generally, the dip is southeasterly, although not usually more than 200; and I see not how this can be explained except by imputing it to the protrusion of the trap range in a direction at right angles to the usual strike of the strata of sandstone.

Near the northern extremity of the sandstone formation in this valley, at Turner's falls, we have another example, where the strata, composed of fine shales and sandstones, are raised for a

considerable distance from the ridge of trap, so as to dip 45°. But at this place is another fact still more conclusive to prove the elevation of the strata since their deposition. Those strata, dipping about 45°, are often covered with the most perfect footmarks in connection with raindrops, and not one of them have I ever seen in the least distorted, as if the animal had walked on a slope. The mud was so delicate as to retain the impressions of each phalanx of the foot most perfectly, and yet it has not yielded at all laterally. No one at all familiar with the tracks of living or extinct animals, can doubt that the surface must have been nearly level when these markings were made. The same is essentially the case at the Horse Race, three miles farther up the stream, and indeed at almost every locality of footmarks the slope of the strata appears to me too great not to have shown the slidings of the animal, if he walked upon it at the same dip which it now has. For I have some examples where the effect of walking on an inclined surface is manifest; and yet at that locality the dip rarely amounted to 10°. On the west side of Connecticut river in Northampton, where are numerous footprints of the huge Brontozoum giganteum, a majority of the rows of tracks cross the strata at right angles to the strike; so that if the strata had their present dip at that time-which is about 12°-either the animals must have walked so high up the bank that no tide or other rise of water could have covered them, as must have been done to bring over a layer of mud, or the strata must have had a less inclination than at present; since some of these rows are several rods long. The latter supposition seems to me to be the true one; and the general fact may here be stated, that, with one exception just discovered, (July, 1847,) it is only on the east or upper side of the trap ranges that the footmarks occur, doubtless because the water was too deep when the rock was soft. My theory is this: That the earlier outbursts of the trap, which perhaps formed the trap tufa, lifted the bottom of what was then the ocean, so as to form a low mud or sand beach, whither the birds resorted, and over which the water frequently flowed, loaded with silt. As successive masses of trap were protruded, successive deposits would come over it, on which other tracks might be formed; and it seems to me that in this way we might explain their occurrence upon successive layers, to a considerable thickness, without resorting, as Mr. Lyell does, to subsidences. And yet when volcanic agency was so common as we know it was during the volcanic period, vertical movements of the surface must have been

a common occurrence.

But I have wandered somewhat from the subject in hand; and to return, it seems to me we have decisive proof that the protrusion of the principal ranges of trap constituting Holyoke and Tom, must have been the last of the changes that brought the

rocks of the valley into their present position; and that the tufaceous traps had previously been protruded and consolidated. It is possible that the principal ranges of trap may have been, at an earlier date than above supposed, partially elevated, and then ultimately have been thrust forward in a solid state, as we know to have been done in some modern examples. And, indeed, the fact that a portion of the turfaceous conglomerates contains really rounded masses of solid trap, shows at least that a good deal of that rock existed in order to furnish these materials. And it must have been more or less above the water in order that the currents should tear up and wear its fragments. This might in part have been done by igneous agency, as the masses were driven upwards through the fissure; yet they appear to me for the most part, to have been acted upon by water. And it is a curious fact that the ripple marks, in some places, show that the current was southerly, as it now is.

Tilting of the Sandstone Formation Generally.

We are conducted, then, as it seems to me, by these facts and reasonings, to a probable solution of the question concerning the relative age of the trap and sandstone of the Connecticut valley. It appears that all the thick bedded sandstone below the trap, constituting at least one-half, was deposited before the protrusion of any trap. Then succeeded some small outbursts, as precursors of the principal ones, and the bedded or contemporaneous trap was produced at intervals, while the upper beds of sandstone were being deposited; and last of all, the principal trap ranges emerged with considerable disturbance of the sandstone. But I have shown that the general dip of that rock, extending through the whole formation, was not the result of the protrusion of the trap. Was it then the result of original deposition, as some maintain, or of some other upheaving force? I incline to the last supposition for the following reasons:

In the first place, if originally deposited with its present dip, the west side of the valley must have formed the shore of the ocean, and the materials have been brought in from the west side of the valley, and have gradually extended entirely across it, with a very uniform easterly dip. But most of the materials composing the sandstone and conglomerates, so far as it is possible to determine, appear to have been derived from the north.

In the second place, as already stated, we have some evidence in the ripple marks, that the current was from the north when the sandstone was deposited. I will mention one striking example. At the footmark locality in Northampton, the ripple marks on the fine sandstone are at right angles to the present course of Connecticut river; and as I was looking at them one

day, close at the water's edge, I saw that they appeared to be continued beneath the stream. But on close examination I found quite a deposit of mud above the rock, and on this mud the current had formed ripple marks very similar to those upon the stone. The rounded masses of trap, which enter into the composition of the tufaceous conglomerate, also, must have been swept southerly.

In the third place, the eastern side of the valley appears to have been elevated earlier than the west side, and to nearly the same height, as I have endeavored to show in my Final Report on the Geology of Massachusetts. If so, it must have presented currents moving from the west shore to the eastward, so as to carry materials for rocks across the entire valley, even if each shore had not been entirely above the waters. Yet we find in fact that the coarsest materials were deposited latest; that is, on the east side of the valley; but as the sides of the valley rose higher and higher, the easterly current, if it existed, must have been weaker and weaker; whereas, that elevation might have made a southerly current more violent, as we find must have been the case.

In the fourth place, a dip from 10° to 20° is too great to be the result of deposition, where the materials are very fine, as in the present case, and where they are spread with great evenness over large surfaces. Unless in peculiar circumstances, plastic materials cannot be retained on such a slope.

Finally, in the elevation and plication of the Hoosic or Green Mountain range, lying west of the valley, it may be by a lateral force, we have an adequate cause for the elevation of the sandstone. This will explain the uniformity of its dip through the whole valley, and its existence even on the eastern border of the valley. It appears to me, therefore, more in accordance with sound geological reasoning, to regard these sandstone strata as tilted up by some of those later movements, probably by lateral pressure, which folded up the Green, Hoosic, and Appalachian

ranges.

ART. XVIII.-On Terrestrial Magnetism; by WILLIAM A. NORTON, Professor of Mathematics and Natural Philosophy in Delaware College.

(Concluded from p. 12.)

Declination of the Magnetic Needle.

THE first computation was of the declination for London, assuming the poles of greatest cold to be coincident with the magnetic poles. The result was a declination a number of degrees too small. After two or three trials I found that by placing hy