4. THE ISOTHERMALS OF THE LAKE REGION. BY ALEXANDER WINCHELL, of Ann Arbor, Michigan.

IT may be remembered that four years ago, at the Buffalo meeting of this Association, I read a paper on the "Fruit Belt of Michigan," in which I presented some statistics, illustrating the influence of Lake Michigan upon the climate of the contiguous regions on the east side. More recently I have had occasion to continue my investigation of the climatology of the Lake Region, and to prosecute it to a much greater degree of thoroughness and detail. For this purpose I have accumulated all the meteorological observations ever published from within the limits of the State of Michigan, as well as many observations yet unpublished. For purposes of comparison, I have collected similar data respecting more than fifty selected localities lying outside of the State of Michigan. The Michigan observations aggregate two hundred and eighty-four years, and those of the other localities four hundred and ninety-three years. The result of this discussion is to establish from extensive inductive data the existence of very remarkable influences exerted by the great lakes upon the temperature of the regions adjacent. A general statement of these results is here presented.

For the purpose of exhibiting the thermometric generalizations to the eye, I have constructed nine isothermal charts, covering the area between the fortieth and forty-eighth parallels of latitude, and between the eightieth and ninety-seventh meridians. This embraces the region within the influence of Lakes Superior, Michigan, and Huron, and the valley of the Mississippi as far west as Kansas and Nebraska.

It is well known that these great bodies of water exert a cooling influence in summer, and a warming influence in winter. The isothermal charts for July and January, to which I direct your attention, present these influences in strong contrast. Turning our attention first to the chart for July, we are at once impressed by the magnitude of the deflections of the isothermals in passing the great lakes. These deflections are toward the south, in consequence of the cooling influence of the lakes. In the lower peninsula of Michigan the lines all form loops opening southward, showing that the mean temperature of July, in the interior, is much higher than along the lake borders. And yet, within the peninsula

of Michigan, the isothermals do not attain so high a northern limit as in the continental region west of Lake Superior. The isotherm of seventy degrees, for instance, first appears within the limits of the chart in the latitude of forty-eight degrees in the valley of the Red River of the North. Passing south-eastward and eastward to the valley of the Menominee River, it comes within the influence of Lake Michigan, and bends directly southward through Green Bay and Milwaukie to latitude 42° 40', and thence trends northward to Traverse City in latitude 44° 40'. Here it is deflected southward again under the influence of Lake Huron, and, passing Saginaw and Sanilac, finally bends north-eastward to attain its normal position, striking Penetanguishene on Georgian Bay of Lake Huron. West of Lake Michigan, this isotherm sweeps across a latitudinal belt of five and a half degrees. Within the peninsula of Michigan, it is deflected first northward two degrees, and then southward one and a half degrees.

Similar deflections are experienced by the isotherms between 67° and 72°. The isotherms of 73°, 74°, and 75°, appear to escape much of the influence of Lake Huron. The isotherm of 74° divides in Southern Michigan, one branch passing eastward through Northern Ohio, and the other southward through Central Indiana and Southern Ohio. The State of Ohio consequently constitutes an area of uniform temperature in July, which is identical with the mean temperature of Central Michigan to the limit of four and a half degrees of latitude, or three hundred miles, further north.

An area in the south-eastern part of the peninsula of Michigan seems to be an area of cold; since the temperature is two or three degrees colder than it is on either side. There exists a region in this part of the State which is topographically elevated about three hundred feet above the general level of the peninsula. It is the region of outcrop of the sandstones of the Marshall Group, but it is not entirely coincident with this area of cold. An area of warmth seems to be indicated in Northern Iowa. It will be observed that the cooling effect of Lake somewhat greater on the west side than on the east. the isotherms deflected from a higher latitude on the west side, but they likewise attain a somewhat lower latitude. The lowest deflection of the curve of 75°, for instance, is at Ottawa, Ill., to the west of the meridian of the lake. The curves of 71° and 72° are also somewhat more southern on the west side than on

Michigan is

Not only are

the east. This circumstance is undoubtedly accounted for by the slight preponderance, during July, of winds from the east of the meridian. Thus, at Chicago, this preponderance is as 60: 33= 1.82; at Milwaukie, as 48: 37=-1.30. But at Milwaukie, and further north, northerly and even north-westerly winds feel the influence of Green Bay.

Contrasting with these results those represented on the isothermal chart for January, we are at once struck with three phenomena: 1st, the great deflection of the isothermal lines; 2d, their northward deflection; and 3d, the exertion of an excessive amount of lake influence upon the east side. All this is illustrated by tracing the isotherm of 22°. Coming within the limits of the chart, a few miles south-west of Omaha, it pursues an undulating course eastward to Ottawa, in Illinois, when it bends abruptly northward, passing west of Chicago, and east of Milwaukie, to Northport, at the mouth of Grand Traverse Bay, whence it bends southward to Corunna, in the middle of the lower peninsula of Michigan, and northward again to Thunder Bay Island of Lake Huron, and thence east to Penetanguishene on Georgian Bay. The isotherm of 23° reaches almost as far north; but, in crossing the peninsula of Michigan, it strikes southward into Northern Indiana and Ohio, thence northward again almost to Thunder Bay Island. The sinuosities of this isotherm spread over a belt four and one-half degrees, or three hundred miles in width. In other words, the influence of the lakes is such that the mean temperature of January at Northport and Thunder Bay Island is identical with that of Omaha, Peoria, Chicago, and Fort Wayne. The January temperature of Mackinac and Marquette is the same as that of Green Bay and Fort Winnebago.

An island of cold is again indicated in the south-eastern part of the peninsula of Michigan. In this case its form and position correspond quite exactly with a region of elevation. The area in Northern Iowa, which in July is an island of warmth, appears to be in January an island of cold. A similar one exists in the elevated region of Southern Wisconsin, while a remarkable axis of cold stretches through Northern Wisconsin and Minnesota. This axis is not entirely coincident with the crest of the ridge dividing the tributaries of Lake Superior from those of the Mississippi; since the warming influence of Lake Superior crowds it about sixty miles southward.

One of the most striking phenomena exhibited by the chart for