No formula has yet been found which agrees with either of the percentages obtained. As the atomic weight of antimony is differently stated from 120 to 122, this is easily explained.

It is unfortunate that the atomic weight of antimony is yet so uncertain, but it is hoped that these results, agreeing so closely among themselves, will be accepted as proving the value of the process.

Arsenous Oxide. This was the next substance analyzed. The sulphide was precipitated in the usual manner, but it was noticed that by having the solution decidedly acid and passing through it a very rapid stream of sulphydric acid, that the sulphide was obtained in a more granular condition.

The precipitate can be dried with impunity at 120°. At 140° the lemon-yellow sulphide commences to change to red, and at 180° is completely converted into the liver-red variety. With ordinary commercial arsenous oxide, the following results were obtained.

(1) 17940 grms. gave 2.2260 grms. As,S=75.64

(2) 5770 (3) 8520

Found pr.cts. Theory pr. cts.

75.76 As.

66

Mr. Lincoln obtained as the result of two analyses:

Found pr. cts. Theory pr. cts.

3

66

=75:77

(1) 8926 grms. gave 1·1087 grms. As2S=75 76 (2) 8145

66

1.0128 Mean 75.76

These results show that this process renders the estimation of antimony and arsenic one of the simplest and most accurate operations in quantitative analysis.

The cones can be used repeatedly and can replace paper filters in every case. They will undoubtedly be found to be of great value in commercial work, for drying crystals and filtering corrosive liquids. As they will stand sudden changes of temperature without breaking, they can be substituted to advantage in many cases for crucibles.

In closing, I desire to return my sincere thanks to my kind teacher, Dr. Gibbs, who has furnished me with the material for these investigations, and has aided me by his counsels and

advice.

Cambridge, March 13th, 1871.

ART. L.-On certain Glacial and Post-glacial phenomena of the Maumee Valley; by G. K. GILBERT, A.M.

THE observations, upon which the following remarks are chiefly based, were made during the summers of 1869 and 1870 for the Ohio Geological Survey, and are here published, with permission, in advance of the official report.

I. The Maumee occupies a broad shallow valley with easy slopes and without definite boundaries. It is the westward prolongation of the trough occupied by Lake Erie, and includes the upper waters at least of the Wabash river as well as the entire basin of the Maumee. The highest point along the axis of the valley is at Fort Wayne, one hundred miles from Lake Erie, and is but 185 feet above the present level of that lake, so that the descent of the Maumee eastward is less than two feet per mile; the sides of the valley rise with a slope nearly as moderate to an altitude of 300 to 400 feet above the median line. So nearly does it approximate to a plain that the consequences of the earth's sphericity are not cancelled, and an observer on one side is prevented from seeing the opposite by the interference of the intervening land.

The rock surface is covered by a sheet of drift, that masks its minor irregularities and contributes to the general impression of a plain. Where the drift has been freshly removed by the wash of waves or currents, the surface of the underlying rock is shown to be planed, striated, and furrowed by glacial action. The general bearing of the striæ, subject to local deflection from details of contour, is parallel to that of the valley. At Sandusky City, and on Kelley's, South Bass, and West Sister Islands it is S. 80° W.; at Fremont and Genoa, east of Toledo, S. 65° W.; at various points in Lucas county, west of Toledo, S. 55° W.; and along the Auglaize river near Defiance, S. W. Farther south the bearing is still more southerly, being S. 35° W. at Lima, O., and S. 15° W. at Middlepoint near Van Wert. That the motion to which this system of striæ is due was southwestward-up the valley of the Maumee as it now lies,—and not in the opposite direction, is attested by some flint nodules contained in limestones of the Waterlime Group on West Sister island and at Monclova near Toledo. By reason of their superior hardness, they were able to resist, better than the limestone, the grinding action, and they project boldly from the planed surfaces, protecting and maintaining each upon its west or lee side, a train or ridge of undisturbed limestone.

The overlying drift is in chief part a blue clay-the Erie Clay of Sir William Logan-containing irregular, lenticular beds of sand and gravel, and abundant, but unequally dis

tributed, glaciated boulders of all sizes up to a diameter of twenty feet. It has already been recognized by Dr. Newberry as the sediment of a sea or ocean flanked by glaciers and floating rock-laden icebergs; and every phase that has fallen under my own observation has served to confirm the identification. Its more elevated portions, which have been exempt from lacustrine action, present a rolling surface with frequent, undrained hollows, occupied by lakelets or the deep swamps that have resulted from their slow filling; while at and below the level of 220 feet (above Lake Erie) this surface has been remodelled and levelled by the waves, currents and sediments of Lake Erie.

Premising thus much of the general character of the valley, I will now ask the reader to notice, upon the accompanying map, some peculiarities of the arrangement of the watercourses. North of the Maumee river, where the land slopes to the southeast, most of the rivulets flow in that direction; but the principal streams, the St. Joseph and Tiffin rivers, cross the system at right angles. Moreover, all the tributaries of the St. Joseph come from the northwest, while the country east of it is drained by branches of the Tiffin, the divide between the waters of the two streams running within five miles of the former. This feature is caused by a narrow step in the country,-a natural embankment, that carries the St. Joseph across the face of the slope, as a mill-race is carried along the bank of a stream: and that river may be said to result from the confluence of eastward-flowing creeks that have found a barrier in the ridge which forms its eastern bank. The slopes of this ridge are quite gentle and the height is moderate; west of Bryan, where it is crossed by a branch of the Lake Shore and Michigan Southern Railway, it rises to an altitude of fifty feet, while at the north line of Ohio it is barely distinguishable. Superficially it everywhere presents the same rolling gravelly clay as the surrounding country.

South of Fort Wayne the same phenomena are presented by the St. Marys river. The slope of the country is toward the northeast, yet its course is to the northwest. Its tributaries are from the south and west, while its right bank is drained by branches of the Auglaize. Where the dividing ridge has been measured, it has an altitude of 35 to 50 feet.

A correlated feature of the two streams-the St. Marys and St. Joseph-is that, forming the sides of a right angle, they unite at its apex, and the resultant river returns within the angle.

The page of history recorded in these phenomena is by no means ambiguous. The ridges, or, more properly, the ridge which determines the courses of the St. Joseph and St. Marys

rivers is a buried terminal moraine of the glacier that moved. southwestward through the Maumee valley. The overlying Erie Clay covers it from sight, but it is shadowed forth on the surface of that deposit, as the ground is pictured through a deep and even canopy of snow. Its irregularly curved outline

accords intimately with the configuration of the valley, and with the direction of the ice markings; its concavity is turned toward the source of motion; its greatest convexity is along the line of least resistance. The notch at Fort Wayne, through which the Maumee flows, was perhaps first cut by the river to which the retiring glacier gave rise, but has certainly been cleaned of the subsequent deposit of clay by a later current,

to be described farther on. While it affords no section of the moraine for examination, it yet confirms the general fact of its existence, by demonstrating that the superficial ridge has here no nucleus of rock in situ. With the exception of the Maumee, no stream crosses the moraine from near Hudson, Michigan, to Fort Amanda, near Lima, Ohio, a distance of 83 miles in a right line and 120 miles on the line of the moraine. Beyond these points I have not traced it, but an inspection of the map suggests that it may be found along the upper waters of the Tiffin and Scioto rivers, as dotted in the accompanying chart. This would increase its length and chord to 200 miles, and 120 miles, respectively.

The courses of the Tiffin and Auglaize rivers are determined by a more easterly, and so more recent, moraine of the same system, but its form is less conspicuous, as this portion of the valley has been greatly modified by lacustrine action. The same agency has obliterated all superficial traces, if any existed, of minor moraines.

South of the St. Marys river are other and numerous moraines accompanied by glacial striæ. Their character and courses have not yet been studied; but their presence carries the mind back to an epoch of the cold period, when the margin of the ice-field was farther south, and the glacier of the Maumee valley was marged in the general mass. As the mantle of ice grew shorter and, in fact, at every stage of its existence-its margin must have been variously notched and lobed in conformity with the contour of the country, the higher lands being first laid bare by the encroaching secular summer. Early in the history of this encroachment the glacier of the Maumee valley constituted one of these lobes, and has recorded its form in the two moraines that I have described.

II. That Lake Erie formerly had an outlet past Fort Wayne, Indiana, and down the Wabash valley, has for some years been recognized by local observers; but, as there is reason to believe that the fact is as yet unpublished, an account of it will not be without interest to the public. I was led to make an examination of this outlet by its relation to certain cotemporaneous beach lines that I had occasion to examine and trace in Ohio.

The relation of the waters of the Great Lakes to their shores has undergone a series of changes since it was first established, by the recession or drainage of the iceberg-bearing sea, at the close of the Drift period. Some of these changes appear to have been gradual and others comparatively sudden, while the

*The fact was suggested to me before I visited the locality, by Martin Arrowsmith, Esq., of Farmers' Centre, O., and was realized less perfectly by several gentlemen whom I met in Indiana.