Septembeb 7, ISSo.] 



SCIENCE. 



317 



ing, Tioga, and Potter counties in Pennsj'lvania, and 

 Cattaraugus county, New Tork, reaching its most 

 northerly part about five miles norlli of Salamanca. 

 From here it runs through Warren, Venango, Butler, 

 Lawrence, and Beaver counties, to the Ohio line, 

 crossing Beaver creek at Chaintonn about fifteen 

 miles above the Ohio river. 



The boundary enters Ohio in the northern part of 

 Columbiana county, and proceeds nearly west to the 

 middle of Stark; then turns more to the south, 

 touching the corner of Tuscarawa.s, and dividing 

 Holmes into two nearly norlh-and-south sections. 

 Near the north-east corner of Knox, the line makes a 

 right angle, and runs south through Knox, Licking, 

 the north-west corner of Perry, Fairfield, Ross, High- 

 land, Adams, and Brown counties. Then it follows 

 the line of the Ohio river across Clermont, and enters 

 Kentucky near the boundary between Pendleton and 

 Campbell counties, and, after crossing the northern 

 part of Kenton and Boone counties, recrosses the 

 Ohio, entering Indiana a little below Aurora. 



In Indiana the line still continues to bear in 

 a southerly direction through Ohio and .Jefferson 

 counties, grazing the edge of Kentucky again oppo- 

 site Madison, and reaching its southernmost point 

 near Charleston in Clarke county, Ind. From here 

 it bears again to the north, through Scott and Jack- 

 son counties, to the line between Bartholomew and 

 Brown, and follows this to the north-east corner of 

 Brown. There ^ain it turns to the south-west, 

 touching the north-east corner of Jlonroe, where it 

 again bears north for ten miles, to near Martinsville 

 in Morgan county. Here again the line turns west 

 and south, passing diagonally through Owen and 

 Green counties, and in Knox as far as Harrison town- 

 ship, ten miles south-east of Vincennes. Beyond 

 this point, the author did not propose at present to 

 trace the line. 



The signs of glaciation cease where there is no 

 barrier to account for their cessation, and where 

 no barrier ever could have existed such as must be 

 supposed if the so-called gl.acial phenomena are the 

 product of floating ice. Of the correctness of this 

 inference, the different elevations at which the signs 

 of glacial .action cease are sufficient proof. For in- 

 stance, the line is near sea-level in New Jersey; in 

 Pennsylvania it rises over Mount Kittatinny to a 

 height of 1,200 feet, then descends SOO feet into a val- 

 ley, and, again rising, reaches the summits of moun- 

 tains 2. OCX) feet above sea-level. Crossing tbe valley 

 of the Susquehanna at an elevation of only 500 feet, 

 the line mounts the AUegbanies diagonally, and runs 

 over them at a height of 2,.")()0 feet. 



The paper proceeds to describe certain marked 

 features of glaciated are.is. South of New England, 

 the terminal line is characterized by a series of gla- 

 cial hills, 100 to 300 feet high. These are also ob- 

 servable in New Jersey, near Plainfield and Menio 

 Park. 



Among the most interesting results of the author's 

 survey in Ohio, was the demonstration of the exist- 

 ence of an ice-dam across the river at Cincinnati. 

 The line bounding two glacial accumulations crosses 



the Ohio river into Kentucky, near the boundary 

 between Campbell and Pendleton counties, about 

 twenty-five miles above Cinciun<iti, and recrosses it 

 near Aurora, Ind., about twenty-five miles below 

 Cincinnati, thus filling the channel for about fifty 

 miles of its course. The Ohio, it should be said, 

 occupies, throughout nearly its whole extent, a nar- 

 row valley of erosion, not often, more than a mile 

 wide, and from :j(X) to .")00 feet ile.'p. Emptying into 

 the main channel there are subordinate channels all 

 along, of smaller dimensions, but of nearly equal 

 depth. The proofs that tlie ice bodily crossed the 

 river at the point indicated arc, that till and granitic 

 bowlders are found in the Kentucky hills south of 

 the river to a certain distance, and not beyond it. 



To the question, Why is the boundary of the glacial 

 area so crooked? the author replied at some length; 

 •assigning as a principal cause, aside from differences 

 of level, the probability that unequal amounts of 

 snow fell over different regions of the north. The 

 effect of such diiterences of accumulating snowfall, 

 in determining the extension of the glacial outline, 

 is illustrated by supposing that two loads of s.and are 

 placed in one pile, and one load in an adjoining pile; 

 when the s.and will flow downward to unequal dis- 

 tances upon a level. 



A little reflectiim will show that the glacial theory 

 will not make extravijgant suppositions as to the 

 amount of ice required. The ice was indeed 600 feet 

 deep over New England, and, very likely, of an 

 equal depth over the area to the west; but it is not 

 necessary to suppose a great increase of this depth to 

 the north. All that is necessary is to suppose great 

 .accumulations of ice to the north of the granitic hills 

 of Canada, starting a movement past tliem to the 

 south. This movement may have been kept up to- 

 ward the margin by fresh accumulations of snow 

 upon the spreading glacier. An accumulation of 

 snow over the glacier in any part of it would spend 

 its effective force in giving impetus to the movement 

 of the front along the lines of least resistance. 



The discussion which followed the reading of this 

 paper took a wide range, as the paper itself contained 

 many points of interest. The opponents of the gla- 

 cial theory, or of the younger theories which have 

 sprung from its loins, based their criticisms chiefly 

 upon doubts of tlie evidences of glaciation. The 

 questions raised, as to the distinctive characteristics 

 of glacial and subaqueous deposits, gave tone to the 

 paper of the next speaker, which was delivered orally, 

 and was, at least in part, extemporaneous. 



The tenuixial moraine west of Ohio. 



BY T. C. CUAMBEKLIS OF BELOIT, WIS. 



This paper was introduced by a statement of the 

 author's views on some points that had been alluded 

 to in the discussion of Professor Wright's paper. Dr. 

 Chamberlin had himself observed the features of the 

 drift-bearing area west of the Kocky Mountains. 

 Certain of the drift clays are unquestionably glacial: 

 others have quite as certainly had a wholly different 

 origin. Be specified with great particularity the 



