In the description it is said: "Muscular impressions in the ventral valve, four; one pair in front of the beak, near the middle or in the upper half of the shell." The pair here alluded to are the laterals. Their upper and lower extremities are sometimes not visible, and what remains occupies the middle portion of the length of the shell. The expression "or in the upper half," I can thus explain: I had the dorsal valve of O. crassa, from Troy, which I then supposed to be a ventral valve. In this the laterals are in the "upper half." The transverse scars were not then observed and hence four scars instead of six. It must be borne in mind that fourteen years ago nothing was known of the internal characters of these shells. The materials were imperfect and consequently so was the description. It is now certain that the genus is a good one and that all of the three species on which it was founded belonged to it.

The described species which I consider to be truly within the genus are: 0. chromatica, O. polita, O. crassa, O. nana, and O. gemma. They all, so far as is yet known, are confined to the Potsdam Epoch. A number of other species have been referred to the genus, but they are all more or less doubtful.

The specimens which have furnished the above additional details of the structure of 0. chromatica were collected at L'Anse au Loup, the only place where the species has been found, in 1863, by T. C. Weston of our Survey, and by him very skilfully worked out of the matrix.

Art. XXI.-On the Damming of Streams by drift ice during the melting of the great Glacier; by J. D. DANA.

WHEN treating of the overflows of the flooded Connecticut, in the Supplementary December Number of this Journal, (p. 497,) I suggested, in view of the fact that the terraces in the Farmington Valley about Tarifville and Simsbury are at least 50 feet higher than those a mile eastward in the parallel Connecticut valley-that the gorge through the Divide Range, by which the Farmington river there passes into the Connecticut valley, had been closed by drift and so remained until the flood had reached its height.

I allude to this subject again to add that the events connected with the opening, in the Spring, of many of our modern icecovered streams afford abundant reason for believing that, during the breaking up of the long Glacial winter, when the melting was going forward, the gaps, gorges or narrows, along the river courses, would have been liable to obstruction by floating ice.

(a) Such obstructions would have been of all grades, from that which could simply impede the free flow of the waters, to the nearly perfect dam.

(b) The obstructions in particular cases might have existed. for a very long era, instead of for a few weeks such as happens after a modern winter.

(c) Again, the slackened or suspended flow of the water, caused by such ice-obstructions, would have favored the deposition and accumulation about them of drift, and some may have thus been converted into complete dams. This process might occasionally have wholly filled with earthy material a gorge or narrow valley, so as to block up and divert the course of the stream.-The well-known case of Niagara River may be an example of this.

In view of these possible results, or rather these probable conditions of many river-valleys in the era of the Glacial flood, we are required to consider whether the height of the upper terraces above the narrows on the several rivers, the Thames below Norwich, the Connecticut below Middletown, the Housatonic below Derby, Westfield River below Westfield, and Farmington River east of Tarifville-was not partly owing, in each case, to the existence of ice-obstructions at the narrows.

It seems to be very probable that this was so. The height of modern spring floods in the Connecticut at Middletown and Hartford is now often due in part to this very cause.

It appears to be certain, that if such obstructions existed in the Thames, Connecticut and Housatonic valleys, they were only partial obstructions; for, in the case of each, the terrace of the valley below the narrows declines quite gradually in height from the level above the narrows, instead of abruptly. Had the waters been held back, up to the height of the high upper terrace, by a close dam, they would have fallen over the dam with a plunge to a lower level; and this abrupt fall would have been registered by means of an abrupt fall in the level of the terrace. Instead of this, the terrrace on passing the narrows southward falls off at a rate not exceeding 10 feet a mile, varying in rate only with the varying width of the valley: a fact that seems to testify to the vastness of the flood as its cause, and not mainly to obstructions. Moreover, the material of the terraces below the narrows is like that above: the same in the prevalence of sands below and coarse gravel at top,—though having the latter of greater coarseness because of the more rapid flow of the stream along a narrower valley.

Further evidence with reference to the existence of such icebarriers is to be looked for in a distribution of gravel and large bowlders across the valley just above the gorge or narrows, where the ice-masses had been brought to a stop and piled together;

for much of the floating ice would have been loaded with bowlders. I have as yet observed no satisfactory evidence of this kind, but think the question needs more investigation. Even if this evidence fails, we can hardly assert that no aid was afforded by ice in producing the great height of the flood-waters above the narrows, or doubt that ice-barriers made of drift ice had much to do with the height and extent of the upper terraces in portions of many other valleys.

There are two questions which should have here a word.

1. May not the obstructions or dams have been made by the Glacier itself? On this point we observe that the extent of the terrace formations along the valleys,-sometimes a score of miles in width even in New England-show that water swept in immense streams over the surface; and thus they seem to prove that the glacier was already out of the lower part of the valleys, and hence too far away to have obstructed the flow except through the pieces set afloat by its dissolution.

2. Were not the dams due to rocky barriers at the narrows, or to the non-excavation of the valley from the narrows southward? The features of the region about the narrows on each of the rivers mentioned, and of the valleys below, suggest decidedly that the valleys had nearly the same depth and extent then as now. The gradual decline in the height of the terrace on going from the narrows southward to the Sound shows that all was one valley, the part above the narrows and its continuation below. The terraces below the narrows, moreover, are built up in general from the present bottom of the valley, or from a lower depth, and this points to a depth for the valley as great as now or greater. It cannot be urged that the lower portions of the terraces were made after the upper. Wherever the hills on one side, at the narrows, retreat so as to give a chance for high terrace deposits, there these deposits are usually found, and sometimes the beds rise abruptly from the water's edge to the level of the highest terrace; and on the Connecticut, in a place of this kind above Middle Haddam, the bottom layers are of claylike the lower lavers in much of the stratified drift on the river.

In fact, the conditions of the terrace deposits of the valley, as well as the features of the valley itself, are explicable only on the view that the part of each valley below the narrows, like the rest of it, the narrows included, had been made before the Champlain period opened. The Glacial period was the era of valley excavation rather than the Champlain period.

ART. XXII-Sliding Friction on an Inclined Plane; by A. S. KIMBALL, Professor of Physics in the Worcester (Mass.) Institute of Industrial Science.

THE following investigation was undertaken with a desire to demonstrate, if possible, by a laboratory experiment, that the law which affirms that the coefficient of sliding friction is constant for all velocities is not strictly true.

Our results seem to establish the point, at least in the case of bodies sliding down an inclined plane. I am aware that the truth of this law has been questioned; indeed the opinion of very many practical mechanics is directly opposed to it. Long ago Prof. Playfair remarked, as the result of some observations made at the slide of Alpnach, that it would appear that friction is neither proportioned to the pressure nor independent of the velocity. Later observations made at the launching of the Raritan and the Princeton (Jour. Frank. Inst., 3d, VII, 108) showed that the coefficient of friction just before the vessel left the ways was much less than during the first five seconds of its motion. More recent still are the experiments of M. Bochet (Comptes Rendus, April 26, 1858,) upon the friction of railway carriages and brakes, which point to the same conclusion; indeed the author goes so far as to give the form of the function which expresses the variation of the coefficient of friction with the velocity, and gives approximate values to its constants for the case of railway trains. His formula is copied by Weisbach with a caution.

Opposed to these views are the careful experiments of Coulomb and Morin, upon which the statements of our textbooks are founded.

The apparatus used in our experiments was simple, but it seems capable of giving very sharp and reliable results. A smooth pine plank 10'x12"x2" was firmly placed at a measured angle with the horizon and supported throughout by stout beams. Upon this plank was a weight box with pine runners, having a bearing surface of 24 square inches. The cover of the box was about six feet in length, and upon it were placed slips of smoked glass. Firmly fixed above the glass, to an independent support, was a verified tuning fork of 435 complete vibrations per second, carrying a style which lightly touched the glass surface beneath it. The weight box was supported in position at the upper end of the inclined plane by a cord fastened to a screw which served to give the box a very slow upward motion. At the proper time the screw was turned, the fork vibrated, the cord cut or burned off, and the box allowed to slide to the bottom of the plane. The style of the fork at the same time

would trace upon the smoked glass a waved line, which would be a perfect autographic register of the experiment. The time of sliding, the velocity at any point, the distance passed over in any unit of time, could all be measured or counted directly from the smoked glass.

The graphical method of working up the experiment was employed, as follows: The bottom of a sheet of section paper was made a "time line" (5 of a sec. = a unit). At various points on this line the corresponding velocities were erected as ordinates. The equation of a line connecting the upper extremities of these ordinates would express the law of the motion studied.

It is evident that this line would have been straight if the acceleration of the slide had been uniform, like that of a body falling in vacuo. If, however, a variable resistance be opposed to the motion of the slide, the acceleration will no longer be uniform, and the line will become curved, concave toward the axis of abscissas, if the resistance is increasing, convex if the resistance diminishes. The acceleration of such a motion at any time will be proportional to the tangent of the angle which the direction of the curve at that point makes with the time line. It is also evident that such acceleration may at once be measured from the paper, since it is the difference between the velocities for two successive units of time. The curve constructed as above, from every experiment made, was decidedly convex toward the time line, showing a constantly decreasing resistance to the motion of the slide as the velocity increased. If we assume that this increase in acceleration was due to a diminished coefficient of friction, the value of the coefficient for any time may be found in the following manner:

Let a, b, and h= the altitude, base, and length of the inclined plane.

W= weight of the slide and contents.

W' normal pressure on the plane, W.

b

= h®

a

g= acceleration of a body falling freely. g'= theoretical acceleration of the slide =9.7° g"= the observed acceleration at any time.

Then the resistance of friction = F=

efficient of friction= =

-(g'-g′), and the co

The following tables give the results obtained from a series