Ohio upon the unusual rise of any of its upper tributaries. At Pittsburg thousands of coal barges collect during low water to take advantage of these waves of translation, and move forward upon them with their valuable freight like a vast army to supply the great cities of the Mississippi Valley with fuel. But, as with

FIG. 1.-JASPER CONGLOMERATE BOWLDER, THREE FEET IN DIAMETER, FROM NORTH OF LAKE HURON. Found near Union, Boone County, Ky. (See Map II.) From photograph by the author, reproduced in The Ice Age of North America, p. 328.

everything else, the best gifts of Nature are those which come in moderation. Enough is better than more. Excessive floods interfere with navigation as effectually as does a lack of water.

With these facts in mind, while surveying, in the year 1882, the glacial boundary across the Mississippi Valley, I reached Cincinnati, having traced the border line to the river twenty-five or thirty miles above the city. Upon crossing to the general level of the hills in Kentucky, I found various indubitable evidences that the ice had extended across the trough of the Ohio, and left its marks several miles south of the river over the northern part of Boone County, and up to an elevation of more than five hundred feet above low-water mark. This was along the watershed between the Licking and Ohio Rivers, which was continuous at this height to the central part of Kentucky. Among other evidences one of the most conspicuous was a bowlder of jasper conglomerate, three feet in diameter, found near Union, in Boone County, which was subsequently transported to Chicago as a part of the Ohio glacial exhibit at the Columbian Exposition. Its right to have a place in an Ohio exhibit was due partly to the fact that it was discovered by an Ohio man, but chiefly from the fact that, at

the snail's pace at which a glacier moves, this bowlder must have been in the territory of Ohio for an enormous period of time, long enough for even a bowlder to become naturalized. If, however, the Canadians should claim it as a fugitive from justice, they would have a prior right, for the ledges from which it was derived are near Thessalon, in Ontario, north of Lake Huron. In searching for bowlders in southern Ohio, I was accustomed to hear them referred to as "niggerheads." In the progress of discovery it was found that the numerous articles of that description which in recent times Kentucky had furnished to Canada were in payment of a debt under which the Dominion had placed the southern commonwealth long ages before.

It is important to note that my discovery of Canadian bowlders on the hills of Kentucky was not the first which had been made there. As far back as 1845 Prof. Locke had noted the post-glacial conglomerate called Split Rock, below Woolpert's Creek, opposite

FIG. 2.-SPLIT ROCK, NEAR MOUTH OF WOOLPERT'S CREEK, KY. This is part of an extensive deposit of bowlders and gravel with some Canadian pebbles, all cemented together by infiltrated carbonate of lime. From photograph by the author, reproduced in The Ice Age of North America, p. 345.

Aurora, Ind., but had regarded this as the remnants of local strata which had been nearly worn away. In 1872 also, Mr. Robert B. Warder had suggested that this was possibly a terminal moraine. Still later Dr. Sutton, in 1876, and Prof. Cox, in 1878, had noted similar deposits near the summit of the Kentucky hills, on Middle Creek opposite Aurora, and had attributed them correctly to glacial action during the maximum stage of the great

Ice period. But because of the imperfect knowledge of the glacial geology of the valley possessed at that time, these discoveries attracted little attention. Various causes, however, conspired to give a somewhat extraordinary notoriety to the facts as they were presented at the meeting of the American Association for the Advancement of Science at Minneapolis in 1883. At that time a systematic exploration of the glacial boundary had been conducted from the Atlantic Ocean to Cincinnati, showing that the Ohio River lay for the most part considerably south of the farthest extension of the ice. Also attention was then first called to the full extent to which the ice had crossed the river in that vicinity. For a distance of nearly one hundred miles it was now demonstrated that the ice came down to the north margin of the trough of the river, and for much of that distance crossed it and mounted the hills upon the opposite side, reaching at one point fully ten miles upon the high land beyond the river. This could not well help suggesting the formation of an ice dam at Cincinnati which would set the water back up the Ohio and its tributaries to the level of the watershed between the Licking and the Ohio, thus forming a narrow and tortuous lake several hundred miles long, which would be five hundred feet deep above Cincinnati and two hundred and fifty feet deep at Pittsburg. (See Map I.)

Finally, some of the geologists who had been engaged upon the survey of western Pennsylvania at once came forward and affirmed that such an obstruction as this supposed at Cincinnati helped to explain a great number of facts respecting certain highlevel gravel terraces characterizing the Alleghany and Monongahela Rivers, which were surprisingly near the level of the water of the supposed glacial lake. At the meeting at Minneapolis Prof. Lesley, under whose vigilant eye the recent geological survey of Pennsylvania has been conducted, declared that he had for some time been expecting the discovery of a local obstruction to the drainage of the Ohio River which would account for the gravel terraces on the Alleghany and Monongahela to which reference has been made, and now, says he, Providence has provided it, and Wright's dam clears up the whole problem, or words to that effect.

Such was the boom with which the theory of the Cincinnati ice dam was brought before the public in 1883. During the ten years which have since elapsed, the hypothesis has been subject to much criticism, so that the faith of some has been shaken, and the theory itself is thought by many to be left in rather a damaged condition. The fullness with which the main facts have been already presented makes it possible to tell the remaining part of the story and state the present condition of the theory in few words.

So complicated are the forces of Nature that one discovery is sure to lead to another, and the man of science soon learns that he never exhausts attainable knowledge even in respect to the simplest subject, and the student has made little true advancement if he has not acquired ability to hold his mind wide open

Map II, showing the partly filled preglacial channel of the Ohio, extending from Cincinnati to the Big Miami at Hamilton. The figures show elevations above the sea.

for the reception of any and all additional facts which may modify and enlarge his theories. In the present case most interesting additions to our knowledge of the facts were made by Prof. Joseph F. James, who called attention to the breadth and depth of the valley running northward from Cincinnati to Hamilton, on the Great Miami River, and to the comparative narrowness and

shallowness of the present rocky gorge of the Ohio between Cincinnati and the mouth of the Great Miami. The relative narrowness also of the latter opening between the rocky escarpments is readily visible to the transient traveler, Mill Creek Valley being about twice as wide as that of the Ohio for fifteen or twenty miles below the mouth of the creek; while a low passage joins Mill Creek at Ludlow Grove which sweeps around north of Walnut Hills, and enters the Ohio through the valley of the Little Miami-Walnut Hills, Mount Auburn, and Mount Lookout, the principal residence portions of the city, being upon a high, rocky pedestal completely surrounded by a depression which has at some time been produced by river erosion.

This valley from Cincinnati to Hamilton is now filled with gravel and clay to a great depth. Upon inquiring for the extent to which the old channel had been filled, it was found by the wells which had been sunk in it that the rock bottom descends from Cincinnati to Hamilton, and is considerably lower than the rock bottom of the present Ohio below Mill Creek. Near Ludlow Grove the bed rock is at least sixty feet below present low water in the Ohio. A few miles farther north, at Ivorydale, on Mill Creek, the bed rock where reached was found to be thirty-four feet below low-water mark in the Ohio, while there was nothing to show that in other portions of the valley the gravel was not still deeper. At Hamilton the bed rock was found to be at least ninety-one feet below the bottom of the Ohio River, showing that there is a deeply buried channel through Mill Creek Valley from Cincinnati to Hamilton; while, according to the inspector, Mr. C. J. Bates, upon building the piers for the great bridge of the Cincinnati Southern Railroad, which crosses the Ohio River near the west end of the city, it was found that the rock bottom was everywhere within a few feet of the low-water mark; thus fully justifying the inference of Prof. James, which can best be given in his own words:

66

". . . Previous to the Glacial period a barrier of land extended from Price Hill on the north to the mouth of the Licking River on the south, preventing the westward flow of the Ohio, and forcing it north and northwest along the channels of Mill Creek and Duck Creek. These met at Ludlow Grove (near Cummingsville) and together continued north to Hamilton. Here entered the Great Miami, and the united streams continued in great volume southward to the present channel of the Ohio, at Lawrenceburg.

"At the coming on of the Glacial period a tongue of ice projecting down the valley from the north and surrounding the Cincinnati Island, as we may call that high land now covered with

* See Map II.