The means employed for knowing the capacity of a machine, as well as for the measure of work it can accomplish, and, also, the measure of the work it has to overcome, is an instrument, usually small enough to be carried in the pocket, but sometimes of very large dimensions. They are made to measure very feeble currents of such small electro-motive as the one millionth part of a volt, or higher than the currents generated by a Brush machine. Such instruments, as you are probably aware, are called galvanometers. They are sometimes called, when used as a measure of quantity of the electric generators under discussion, current meters. Some of them are very complex, and require a knowledge of the higher mathematics to understand the principles upon which they are based, while others are so simple as to require no more than a knowledge of simple ratios. Of the latter class is that known as Deprez's galvanometer, a very crude sample of which has been brought here. This instrument is based upon the ratio of two opposing magnetic forces, one of which is obtained from a permanent magnet, which is a part of the instrument, and the other is created by the current to to be measured. In form, it consists of a horse-shoe magnet, made up of five layers of flat steel, bent in the form of the letter U, and fastened to a board which serves as a base. In the strongest part

of the field, between the poles, is fastened a rectangular box, which just fills the space between the limbs of the magnet, around which is wound, for the measurement of currents of intensity, a very few turns of coarse wire, or a single band of sheet copper. In the inside

of the box, upon two V shaped pieces of brass, one of which is fasttened to each end, rests a small thin bar of iron with its two ends V shaped, cut along each edge like the teeth of a very coarse comb to deduce friction, and wide enough just to clear the inside of the box. At the end of this bar next to the binding posts, is inserted an upright piece of brass wire, whose movements, to right or left, are indicated and measured upon a quadrant, placed just behind it, and arching the ends of the magnet. Two binding posts, with screws, standing in front of the quadrant, to which are attached the ends of the wire encircling the box, complete the machine.

This instrument has been given the name of absolute galvanometer, from its power of measuring the intensity and electro-motive force of a current. Its action is as follows:-The permanent magnet

ism, contained in the magnet, induces opposite magnetism in the ends of the needle, attracting them, so that they lie in a horizontal plane, with their ends toward either pole of the magnet. Their position causes the brass pointer before mentioned, to take up a vertical position, opposite a zero mark, on the quadrant. To test the current of any machine, the conductors are attached to the binding posts. On the armature being set in motion, the passage of the current around the box sets up a magnetic whirl, which causes the needles to take up a new position, dependent upon its intensity, which is indicated by the pointer on the quadrant, which is the measure of the intensity of the current. Its present arrangement is intended as a measure of intensity, or quantity only. To measure electro-motive force, it is only necessary to substitute for the present box, one wound with many turns of fine wire, offering a resistence of 300 or 400 ohms. The needle will then indicate, as before, the electro-motive force of the current, It is necessary, before using the instrument, to compare it with some well known standard of resistance, in order to be able to measure the intensity or electro-motive force of any current.

SUPERFICIAL GEOLOGY OF DUNDAS VALLEY AND WESTERN ANCASTER.

BY WM. KENNEDY.

In 1879, Mr. J. F Carll, of the second Geological Survey of Pennsylvania, presented a report on the pre-glacial and post-glacial drainage of the Lake Erie country, in which it was shown that a great number of the streams of the northern part of Pennsylvania, in pre-glacial times, flowed into Lake Erie.*

The great difficulty with Mr. Carll's deductions was the finding a necessary outlet for Lake Erie.

On 18th March, 1881, Professor Spencer, of King's College, Windsor, Nova Scotia, read a paper on "The discovery of the preglacial outlet of the basin of Lake Erie into Lake Ontario," before the American Philosophical Society, and on the 8th December last, Professor Spencer read a paper on the same subject, before the Hamilton Association. On the formation of the Geological Section of this Association, it was understood that the Geology of the dis* Report Second Geological Survey of Pennsylvania, III. page 330 et Seq.

trict, and indeed the whole of Wentworth County, should be worked out by the Section and the report presented to a full meeting of the Association.

The following pages are intended to form a part of this report, when the Geological Section has completed its work. The question primarily discussed, is the Superficial Geology of that part of the country lying in Dundas Valley and the parts of Ancaster Township around the head of the valley.

Dundas Valley lies at the western end of Lake Ontario, in the form of a rude triangle, having for its base the beach, spanning the mouth of Burlington Bay, and for its two sides the Niagara escarpment. The valley may be divided into three parts. First, the lower portion, occupied by Burlington Bay, a deep body of water, bordered by a low sandy shore, much broken by inlets on the southern side, and a shore rising almost precipitously to the plain above, on the northern side. (This north shore consists of sand and other drift materials.) The bay is enclosed from Lake Ontario by a low semicircular beach of sand and gravel, and is separated from the second or middle third, by Burlington Heights, an old beach containing fossils of the Hudson River period. Second, the middle division, or lower portion of the valley proper, extending from Burlington Heights to within the vicinity of the town of Dundas. And the third or upper portion of the valley, comprising all that broken and hilly region at the head of the valley, and extending from Dundas to the village of Copetown, where the valley proper ends.

In this present paper, I shall not take into consideration the first or lower division, containing Burlington Bay, but confine myself to the two upper divisions of the valley. The division containing Burlington Bay had other and later causes at work in its formation, than those concerned in the construction of the upper two divisions of the valley.

Beginning at Burlington Heights, we have then a narrow canyon shaped channel, lying in a position about N. 70, E., and a little more than eight miles long, cut out between the two walls of rock forming the escarpments. This channel is about four miles wide at the lower end, and gradually narrowing until at Binkley's Corner, on the Hamilton and Ancaster road, the valley is three miles wide, a width it maintains for more than two miles, or until after passing Dun

das, when the more westerly escarpments turns slightly east, to the village of Copetown, where it approaches within a mile and a quarter of the eastern side of the valley. From Copetown this western escarpment turns in a westerly direction and disappears. These measurements are from a map made by Mr. T. C. Keefer, in 1859. This map is on a scale of two inches to the mile.

Where, or at what elevation, these escarpments join each other, has not yet been determined. Sir William Logan in his Geology of Canada, published in 1863, says, " It is not, however, certain where it (the Niagara Formation) folds over the Dundas anticlinal, there being no exposures whatever upon the axis. The most western appearance of the upper part of the formation, on the south side of the anticlinal occurs in the vicinity of Ancaster; the most western on the opposite side, about two miles north of Ancaster, on the third lot of the first range of Flamboro' West. It may be inferred from the trend of the formation on each side, and from the general shape of the country, that its summit would fold over the axis of the anticlinal on the line between the townships of Ancaster and Beverly, at about the thirty-fourth lot.

The portion of the valley from Burlington Heights to the town of Dundas, is to a considerable extent, occupied by Dundas Marsh. Between the marsh and the detritus at the foot of the escarpment on both sides, there is a tract of raised level country lying at a general elevation of about eighty feet above the level of the lake on the eastern side, and, a perhaps somewhat higher elevation on the western side. The level plain on the eastern side is here and there cut through to the blue Erie clay, by streams of recent origin. On the western side, the country rises by broad successive steps to the foot of the escarpment. The western side is also peculiar in the absence of streams of any size, and also their fewness in number. The composition of this level plain appears to be chiefly beds of clay and silt in alternate layers, with patches of conglomerate in places.

The division from Dundas to Copetown, lies on a much higher plain, rising by steps to the summit of the valley.

Passing up the valley we come to the second elevation about a quarter of a mile beyond Binkley's Corner, on the Hamilton and Ancaster road. This elevation, which is about twenty feet higher than the general level of the second division, stretches in a semicircular

form with its concave side looking down the valley, from the Hamilton or eastern side, crossing the public road, passing through Mr. Hatt's farm and coming to an end in the heavy clay beds near Dundas. This level is very much cut up by deep ravines; showing on their sides in many places, gravel terraces or levels of resting places of the waters of the lake in former places. From this point to the upper end of the valley the district rises in quick stages. This district towards the head of the valley, is much cut up by streams, and showing a generally broken surface. The hills between the streams lie in positions so that their long axis points in the direction of the long axis of the valley. Several, and indeed most of these hills, show distinct traces of two or more terraces or old beaches, and being in every case rounded on the top. Some of the hills are cone shaped, and this the more so, the nearer the head of the valley is approached. In composition, these hills, are for the most part clay of a whitish yellow color, lying upon beds of a stiff blue clay, or bluish sand. The yellow clay shows little or no signs of stratification, in any manThe cone, or rounded hills, near the head of the valley, consist to a great extent, of drift sand or silt, and some few being of fine gravel mixed with reddish colored silt. Many of them have all the characteristics of sand dunes, the sand being apparently blown sand. A number of these hills in this division contain beds of conglomerate

ner.

The height of land closing the head of the valley proper, and separating the drainage system of Dundas valley, from that of Fairchild's Creek, and the Grand River, is composed largely of coarse, washed or beach sand, with broken shales in some parts. On the road leading from Ancaster to Jerseyville, on the farm of Mr. J. Crysler, there is a fine exposure of these gravel or sand beds. Here, the sand is distinctly stratified, lying at a high angle and dipping eastward, or down the valley. The angle of beds, to the west, or towards the head of the ridge, being the highest (abcut 40 degrees). Passing east, the beds gradually assume a more horizontal position until they merge into the general level. Again, on the line of the next concession road to the north, and about a mile and a half, or two miles, further west, there is an exposure of beds of the same material dipping at a low angle to the west. This second exposure is on the southern border of a large swamp. On the northern border of this swamp, and still on the western side of the height of land the ridge