the river with his barometer, the agency of the periodical barometric changes would also be perceived by their effects upon his determinations of elevation. I am supposing in this case of imagined barometric heights, that such an observer would have no corresponding observations at the sea level, and that the height of the barometer at the mouth of the Amazon would be considered a constant.

The mean monthly heights of the barometer at Para, as observed by Dewey in 1846, 1847, and 1848, and till May, 1849, showed an extreme range of only 0.41 in., viz.: from 30-02, which was the monthly mean for July, 1846, to 2961, which was the monthly mean for September, 1846.

The fluctuations arising from the monthly barometric changes might give the line of descent along the Amazon, as determined in this way, a wave-like appearance, amounting, perhaps, to 300 or 400 feet at most. But in the case before us, the change actually amounted to something like 2,000 feet. For after Herndon had descended the river 707 miles, and approached with its current the sea level 571 feet, he was then 1,589 feet higher than he was when he set off.

We cannot, therefore, well conceive how we could find from such a source as daily or monthly changes in the uniform barometric pressure of Para, such anomalies in barometric determinations, heights, and pressures as were observed.

If the suggestion, that the high boiling point of Herndon and the high barometer of Castlenau at Nauta were caused by the pressure of the Trade Winds against the Andes, should turn out correct, and the barometric observations on the head waters of the Amazon, both of Humboldt and Condamine, tend to confirm it, will not the Andes be converted into an immense anemometer, by which the force of the Trade Winds may be determined; and if their force, consequently their velocity also?

ART. XXXIX.-Impressions (chiefly Tracks) on Alluvial Clay, in Hadley, Mass.; by CHARLES H. HITCHCOCK of Amherst College.

In the summer of 1852, in company with Mr. E. C. Bolles of Hartford, I accidentally discovered various impressions on a clay bed situated upon the east bank of the Connecticut river directly south of Hadley Centre, and a short distance north of Shepard's Island. The bed lies beneath about twenty feet of alluvial sand, which abounds in ferruginous tubular concretions. By the action of freshets, a large amount of the sand lying upon the clay has been removed, leaving about two or three acres of level surface exposed. The bed itself is close by the place described

in my father's (Pres. Hitchcock's) Final Report on the Geology of Massachusetts, as abounding in claystones of remarkable forms.

The impressions are principally found upon muddy deposits made by rains in the irregularities of the surface. The circumstances in which these impressions occur, afford an admirable illustration of the manner in which similar appearances were produced upon what is now solid rock. It is to be lamented that recent tracks were not more studied at the time when it was doubted by men of science whether ichnolites were originally made by animals. In those days of discussion at least one of such doubters was convinced that the impressions were foot-marks, by noticing a piece of clay in the cabinet having on it a few tracks of a snipe. Though skeptics are now few on this subject, additional confirmation of the facts and deductions of Ichuolithology may still be of value.

Impressions of thirteen different kinds of animals have been noticed at this locality: viz., of man, four species of birds, two of quadrupeds, one batrachian, snails and annelids, besides two or three of a doubtful character.

The human imprint is one of the most interesting. It is a single impression of a boy's foot, and occurs with two of a crow. Raindrop impressions had been made on the spot before the others had been formed, and were not entirely obliterated by the foot of the boy. All the striæ and lines upon the sole of the foot appear distinctly on the specimens, particularly the fine striæ and ridges. The phalangeal impressions and papillæ of the crow's foot are also strongly marked. The difference between the integuments of the foot of man and birds is finely exhibited in the former the lines are much finer, and parallel to one another, running mostly across the foot; while in the latter the papillæ cover the whole phalanx with dots, scattered irregularly.

The tracks which I have referred to the crow, may have been made by some other bird. It is a curious fact, that these tracks are frequently more difficult to refer to the true animal than Triassic* impressions. The most common track at this locality is that of the common snipe, Tringa minuta. It is four-toed, and about an inch in length. In some places on the bed I have seen

*Prof. Henry D. Rogers (see this volume, page 123) has advanced reasons for supposing that the sandstone of the Connecticut valley belongs to the Jurassic rather than the Triassic series. My brother, also, (Dr. Edward Hitchcock, Jr., of E. Hampton, Mass.,) has recently discovered in this formation fine specimens of the Clathropteris, a genus of ferns confined in Europe exclusively to Liassic sandstone. He proposes to describe it in the following number of this Journal. But strong as the probability is that the Liassic sandstone exists in this valley, I have thought it best to use the name Trias. The recent measurement of two sections across the valley by my father, shows a thickness of rock four times greater than either the Lias or Trias of Europe, and leads him to conclude that probably several formations may exist here, which in time may be distinguished.

perfect tracks of this species in rows several yards long, and some have been found in relief, by splitting open the layers of clay. A few thousand years might so petrify the impressions. and their casts, that they could not be distinguished from fossil foot-marks, except so far as difference in species is concerned. The tracks of two other species of birds, one of them, perhaps a Tringa, have been found; but with no additional character of importance.

Of the two genera of quadrupedal tracks noticed, those of the dog only are recognised. In some specimens, because of the toughness of the clay, the claws merely are impressed. A careless observer might suppose that these canine impressions were made by a biped, because he sees only an apparently right and left foot-mark. But a close examination shows that every track is double: that is, the hind-foot steps into the tracks of each forefoot. In other cases, instead of being double, the two tracks are placed side by side very near to each other. The marks of hair may be seen on some specimens. The papillæ of the dog and crow are quite similar.

At another place there were impressions in two long rows which were possibly produced by a young bird-the progress of Tringa minuta, perhaps. Others were made by a frog at rest. Of these last, the two largest were made by the hind feet; they are toed in. An oval spot between the tracks was probably made by the body of the animal.

Another class of impressions forms irregularly curved lines. These may have been made by the smaller Conchifers and Mollusks, Unio and Paludina: of course, different species which were of the same size could not be distinguished from each other by their tracks.

The trackway of an Annelid is very distinct. It is a continuous fimbriated trail, depressed in the centre throughout its whole extent and elevated along its edges. At least two species have been noticed. Of these daguerreotypes have been taken as well as of some others; and we find this an admirable, though somewhat expensive, mode of preserving them.

Raindrop impressions and air-vesicles have been found in abundance. They are more fully noticed beyond.

From the facts obtained at this locality we derive a very clear idea of the manner in which foot-marks on stone were produced.

The surface of the denuded clay is somewhat uneven. Every time it rains the depressions become filled with water, thus making small pools of muddy fluid, which form a deposit admirably fitted for receiving impressions. Upon this soft stratum, after the water has nearly dried up, animals tread, leaving the imprints of their feet. Subsequently the deposit becomes hardened by the SECOND SERIES, Vol XIX, No. 57.—May, 1855.

50

heat of the sun, and the tracks of course remain upon the clay as distinct as when first formed. Another storm arises, and a new stratum is brought on, covering up the former layer, without obliterating the tracks. When the water has retired, this second deposit is trod upon, and hardened in the same manner as the first. At the same time the new layer fills up the recently made tracks. Thus an exact cast of the forms of the tracks is produced. A third stratum protects the impressions upon the second deposit, and in this way the process may go on indefinitely. Having dug up and cleared the successive layers, the tracks in relief, as well as the corresponding depressed ones, have been found, as was stated in the description of the foot-marks made by the Tringa minuta.

Now let us consider the condition in Triassic times, when the Connecticut valley was an estuary. Upon its shore after a plentiful rain, birds and other animals were scattered along in search of accustomed food. The heat of a tropical sun quickly hardens the first stratum impressed with tracks. Another shower comes, or more probably the tide rises, and a second layer is brought on, covering up the impressions just formed. This too in turn receives the impress of feet and becomes hardened, as may also others in succession till the sandstone of the whole valley above the tracks, at least a mile thick, was deposited. And since animals seek their food chiefly along the margin of the water, we find the rock, usually over only a few feet in width, impressed with tracks.

The correspondence between the alluvial and Triassic impressions extends to every minute point. In both formations appears the same alternate order of right and left foot. They are found in relief in each instance. The phalangeal impresssons are analogous in both periods. The print of the claws, also, is quite distinct upon stone as well as clay.

Tetrapodal ichnolites find alluvial counterparts at this locality. The form of feet and manner of gait of the Dog and Anisopus Deweyanus somewhat resemble each other. The fore-feet however, in the fossil species, are much smaller than the hind. But the impressions, instead of being situated at an equal distance from each other, are arranged by twos-a large and a small track together. This peculiar arrangement arises from the structure of quadrupeds, and the rapidity with which the animal moves. The swifter the movement the more nearly the hind foot is brought into the place of the fore one.

The tracks of a frog correspond well with the ichnolites of the Anomapus scambus. There are specimens of the impressions of this animal in the cabinet of Amherst College, proving it to be a huge batrachian in a sitting posture, and like the specimen on clay. The impressions of the Conchifers and Mollusks

upon the clay, illustrate the Herpystezoum Marshii and H minimum of the fossil foot-marks. No difference in character between the Triassic and alluvial specimens has been noticed. The trails of Annelids have been found upon stone. Refererence is made to those upon the Clinton group of the New York Silurian rocks, described in the 2d volume of the Palæontology of New York, page 30, 31, figures 13 and 14. Some differences exist between the impressions occurring in the different formations. But much of the diversity may be referred to the disparity of size in the animals, and the partial filling up of the fossil trails with sand.

These tracks afford a presumption respecting the time of man's appearance upon the globe. Among these impressions in Hadley as already mentioned, are those of man. Now, if he had lived when ichnolites were formed, it is probable that he would have left similar traces of his existence on stone: but in the whole series of rocks below the alluvial, no tracks made by human feet have been discovered.

The phenomenon of raindrops on stone receives a beautiful illustration at the Hadley locality. These are preserved most perfectly when it barely sprinkles. In a heavy or long continued shower, so many impressions are made, that they coalesce, and leave no distinct trace of their separate existence. The surface in this case resembles a chopped sea. Specimens of raindrop impressions denoting every variety of shower, have been found upon stone and upon this alluvial clay. In no particular are they dissimilar. On page 502 of the Final Report on Geol. Mass., it is stated that some specimens of raindrop impressions denoted the direction of the wind when the mud was sprinkled. The alluvial depressions indicate the same fact, though less perfectly; owing of course to the want of strength in the wind.

The Hadley clay furnishes another appearance so similar to raindrops as to be mistaken for them, and which has led some geologists to doubt the existence of fossil rainmarks. Air-vesicles are sometimes found scattered like raindrops over the surface. These are of course swelled upwards. By this mark they can always, when the surface has been undisturbed, be distinguished from raindrop impressions. But when the clay dries, the air escapes, and the vesicles collapse, and may even sink below the surface. The gas in the vesicles probably proceeds from the partial decomposition of organic matter in the clay. That organic matter also mixed with the clay, forms a thin film at its surface, which is raised by the gas into pustules.

Another kind of impression common to stone and this clay is that of ripple marks. These are formed by the action of waves upon the bottom, arranging small quantities of sand and clay in continuous ridges. Specimens of ripple marks upon stone and