overflowing springs. No possible head to this overflow could exist; and, in general, this class of springs flowing out of the mountain-tops can not be materially influenced by the rainfall. There is no land above them from which such torrents could flow in such constant abundance. The White and Adirondack Mountains are also full of similar cases.
"Scribner's Monthly" (vol. xi., p. 784) has a very interesting article by Martin A. Howell, Jr., entitled "Is there a Subterranean Outlet to the Upper Lake Region?" While we are sorry to quarrel with Mr. Howell as to his conclusions, we are very happy to accept his facts. He speculates upon the premise that, because "an area of some 400,000 square miles is drained by the tributaries of Lake Winnipeg alone," a certain amount of this accumulation of waters which do not find an exit "toward the Polar Sea and through the Mississippi Valley" may pass by subterranean channels into Lakes Superior and Huron; and he says that "while Lakes Superior and Huron are supplied largely through such subterranean channels on the one hand, they suffer severely through losses by similar channels at some point in their vast expanse." He shows by a map the track of this supposed underground current to be down the valley of the Illinois from Lake Superior to the valley of the Mississippi. The facts he gives tend strongly to support his novel theory of underground flow southward from the lakes, however it may be as to the amount of it. He gives no evidence that it comes from the direction of Lake Winnipeg, but, on the contrary, the balance of his evidence goes to prove that the Northern lakes are nothing more or less than great, overflowing springs. "That there exist channels of communication with some of these lakes," Mr. Howell says, "has long been believed and admitted by many"; and then, having shown that Lake Superior at its surface is 600 feet above the Atlantic and at its bottom 573, and Ontario to be 235 feet above, with the same depth as Superior, he proceeds to make the following significant statement, which is not at all conclusive as to the intercommunication between the lakes, but is unanswerable as proof that these lakes are overflowing springs:
"And that a great subterranean influx into the upper lakes exists there is little doubt, as a comparison of the discharge through the mighty St. Lawrence with the limited supply from the country bordering on the upper lakes clearly demonstrates, leaving the problem to be settled in the mind as to where this volume does come from in its course to the ocean. Again, the discharge through the St. Lawrence is equal to double the volume poured into Ontario through the Niagara, or into Erie through the St. Clair; suggesting that from the shallowness of Lake Erie and the great depth of Superior and Huron a subterranean channel may connect Superior and Huron with Ontario, giving to the latter, through this source, to be discharged by the St. Lawrence, a greater volume than is given through St. Clair. It is also
