which relate to the first Lake Algonquin. So far as they go, however, the latter are almost as good. Only the southern parts of the Algonquin beach are now left, those at the north having been washed over and obliterated by the later Warren Gulf. All the facts now known relating to the first Lake Algonquin are found south of the Petoskey node line.

In the Michigan Basin.-The history of Lake Algonquin in this basin was very simple in both epochs. More than half of the deformation which is found in the Algonquin beach was produced by the great Champlain uplift which so recently deformed the much younger Nipissing beach.

As nearly as I have been able to determine, the node line of the Chippewa and Algonquin planes passes east and west through Petoskey, as shown on the map by the line AA. North of this line the highest beach throughout is the marine beach of Warren Gulf and rises rapidly northward. South of the line is the true Algonquin beach. In the field the two so far as yet observed appear to be physically continuous as one beach. But taking the plane as a whole there is a decided break near the line AA, and this I take to be the node. This break in the plane is clear on both shores of Lake Michigan, but is less noticeable in the highlands east of Georgian Bay. Northward from Petoskey the Chippewa beach rises to Mackinac about three feet per mile, while southward from the same place the Algonquin beach declines about seven or eight inches per mile. The altitude of the Algonquin beach at Traverse City was not determined accurately. But terraces have been reported at Frankfort on the Michigan shore farther south, corresponding very closely to the extended Algonquin plane.

On the west side of the lake the Chippewa beach rises northward from Burnt Bluff to Cook's Mill at the rate of about two and one-third feet per mile. The Algonquin beach was found to pass under present lake level at Two Rivers, Wisconsin. Southward from that place it appears as submerged terraces off shore at Manitowoc and Sheboygan. Northward from Two Rivers it was found at Kewaunee, Green Bay, Sturgeon Bay, and Birch Creek north of Menominee. Its heights at all these places fall into an even plane which rises northward eight inches per mile. If this plane be produced northward to Cedar river on the west shore of Green Bay about 25 miles north of Menominee its altitude at that place would be about 60 feet above the lake. The Nipissing beach is calculated to pass under the lake at about the same place. Hence the Algonquin beach is there about 60 feet above the Nipissing. Cedar river is also near the node line AA, and is 120 miles straight west from Petoskey. At Petoskey the Nipissing

beach is 25 feet above the lake, and the Algonquin 100 feet. From these data it appears that in 120 miles from west to east the Nipissing beach rises 25 feet and the Algonquin beach 40 feet, or at rates of two inches and three and one-half inches per mile respectively. On the east side the Algonquin plane projected southward, passes under the lake in the vicinity of Ludington or Pentwater. On the west side the plane carried from Two Rivers to Chicago 150 miles south strikes more than 100 feet below the lake.

Mr. Upham supposes his glacial Lake Warren to have had its outlet at Chicago, and he refers to my third paper to show that the highest beach of the Green Bay region continues southward about at present lake level. He may have been misled by a misprint in the table of altitudes in that paper.* But I state distinctly in the concluding paragraph of the article that the beach passes under the lake southward from Two Rivers. If he accepts the present facts without offering others to controvert the conclusions which they plainly indicate, Mr. Upham will have to grant that neither the waters which made the Algonquin beach nor those which made the Chippewa beach had an outlet at Chicago. There are beaches converging to the Chicago outlet, as is well proved by the observations of Andrews, Bannister and Leverett. But they do not extend far north, and all are in planes which lie above the Algonquin, Chippewa and Nipissing planes. These latter planes all pass down under the lake long before they reach Chicago. The Algonquin reaches farthest south above lake level, and its plane produced strikes higher at the south end of the lake than either of the others.

It is interesting to note the relation of this part of the Algonquin plane to the two outlets which served the second lake Algonquin. From Petoskey to Port Huron is about 175 miles south and 125 miles east, and the plane projected from Lake Michigan passes 40 feet above the calculated place of Algonquin beach off Sarnia. From Petoskey, North Bay is about 260 miles east and 60 miles north, and the plane produced passes 210 feet above Georgian Bay, or about 50 feet above the Nipissing beach at the old outlet at North Bay. Subtracting 40 feet for post-Nipissing local elevation from the altitude of the pass at North Bay, the Algonquin plane produced strikes 90 feet above the outlet. Considering the fact that the Algonquin beach is the oldest of the three it is not surprising that its plane produced from so great a distance comes no nearer to these outlets.

*Third paper, second line in table on page 325, the altitude of the "Beach southward from Two Rivers, Wis.," was intended to read "-582" feet, not "582" feet.

The altitudes of the Algonquin beach at the different points of observation are given in the following table in feet above Lake Huron, which is 582 feet above sea level:

160

Kirkfield (Spencer)

Gravenhurst (Taylor) approx.,

Bracebridge

Huntsville

66

Water level over the Nipissing pass...

Prof. Spencer's heights were nearly all measured by leveling, those of the writer by aneroid barometer.

In the Huron-Georgian Bay Basin.-The eastern part of the Algonquin beach has been greatly deformed. But its relation to the St. Clair river points strongly to its origin as a lake beach. It has an eastward rise of about one foot per mile throughout. The northward rise is less at the south. At Grand Bend, which is about 40 miles northeast of Sarnia, the beach is 18 feet above the lake. The place of the beach (submerged 20 feet) off Sarnia was calculated on these data. It probably passes under the lake near Cape Ipperwash, which is about 25 miles northeast of Sarnia. On the opposite sides of Saginaw Bay there are two points, Sand point on the

east side and Gravelly point on the west, which may also be, as Spencer makes them conjecturally, nodal points. But this has not yet been established as a fact by observation.

From 20 feet below lake level at Sarnia to the Nipissing pass is a rise of 140 feet (40 feet, as before, deducted from altitude of Nipissing pass for recent local uplift). The northward distance being about 235 miles, the rate of rise is about seven inches per mile. Southampton is 125 miles north of Sarnia and 50 miles east. Kirkfield is 140 miles east and six miles north of Southampton. The Algonquin beach rises northward from Lorneville to Kirkfield 60 feet in 15 miles, and it descends from Kirkfield to Southampton 161 feet. Taking off 24 feet for the greater distance of Kirkfield north makes 137 feet descent westward in 140 miles to Southampton. This descent continued 50 miles farther west to the meridian of Sarnia passes 82 feet above the lake, whereas the SarniaNipissing plane is 19 feet lower-63 feet above the lake at the same point. As compared with this plane the Algonquin beach is 69 feet too high at Southampton, 130 feet at Lorneville, 190 feet at Kirkfield, and 334 feet at Huntsville. The same comparison shows that the beach at Petoskey is only five or six feet higher, and that at Cedar river on the west shore of Green Bay is about 34 feet lower. The Algonquin plane produced from Cedar river and Petoskey to Huntsville, 275 miles east, would be 90 feet above the Sarnia-Nipissing plane. It seems certain that the Algonquin beach was originally formed very nearly in the latter plane, and the present departure from it is the measure of the amount of subsequent unequal deformation.

These facts show how great is the eastward element of deformation east of Lake Huron and Georgian Bay. The northward uplift appears also to be irregular, especially farther north, as shown by the Chippewa beach. All this deformation, less that due to the later Champlain uplift, appears to have occurred after the Chippewa beach and before the Nipissing. This is the present sum of the facts which bear directly upon the first Lake Algonquin, so far as relates to the evidence in the upper lake basins, and if this were all the existence of the lake as one with a two-outlet stage like the second Lake Algonquin would remain in some doubt. But there are facts connected with the Niagara gorge and Warren Gulf which define the conditions which existed just before and just after the supposed first Lake Algonquin so clearly that theoretical considerations show that this lake must have existed in the interval. These will be presented briefly later.

AM. JOUR. SCI.-THIRD SERIES, VOL. XLIX, No. 292.-APRIL, 1895.

The Niagara Gorge.

In his admirable paper on the "Duration of Niagara Falls," Prof. Spencer presents five cross sections of the Niagara gorge.

[300

200 Niagara

Clinton'

100

O Medina

-80

A

2.

500

B

300

2.00

Niagara

100

o Medina

C

300

1000

500 1000

500 1000

500

7000

500 1000

O

A close comparison of them is so instructive that I have reproduced them in a single cut. Some of the details of his drawings are omitted. The significance of the difference between C and D and E was sufficiently explained in the paper on the second Lake Algonquin. The conclusion then reached was, that considering the uniformity of the geological structure it must be admitted that if D and E were made by the great cataract, C could not have been, but must have been made by a stream of much less volume. That stream was the Erigan river, draining only Lake Erie, which did not then receive the discharge of the upper lakes. Prof. Spencer's explanation, that C was made while the great cataract had a sheer fall of 420 feet was shown to be inconsistent with the obvious fact that the Erigan section of the gorge is much shallower than the wider secotion above.

If this explanation of the origin. of the Erigan gorge, which is represented by C, is valid, then the same argument applies to the lower section of the gorge which is represented by B and A. These sections must have been made by the great cataract and not by the Erigan Fall. For A and B are almost exactly the

NOTE TO FIG. 2.-These cross-sections are situated as follows beginning at the north or lower end of the gorge. A, half a mile above lower end; B, at Foster's Flat; C, at Whirlpool rapids just below the railroad bridges; D, at Johnson's Ridge about one mile above the bridges; E, at the Horseshoe Fall. The top of the sections is hard Niagara limestone. Then comes softer, shaley layers with the harder Medina below. 00, level of Lake Ontario; rr, river level; ii, level of Iroquois marine water during middle of Erigan epoch; bb, probable bottom of gorge (about 75 feet below rr). Both scales in feet. The strata are substantially the same on both sides of the gorge. These sections are copied from Prof. J. W. Spencer's cuts in his paper in this Journal for December 1894, with some omissions and also with some additions. Note the shallowness of C (bottom about at bb) and narrowness at top as compared with A, B, D or E.