MODIFICATION OF GREAT LAKES BY EARTH MOVEMENT. 353 



the glacial lakes ami lias given much study to related problems, is of 

 opinion that the movements are not complete, and predicts that they 

 will result in the restoration of the Chicago outlet of Lake Michigan 

 aud the drying of Niagara. 1 



The importance of testing this question by actual measurements was 

 impressed upon me several years ago, and 1 endeavored to secure the 

 institution of an elaborate sei of observations to that end. Failing in 

 this, I undertook a less expensive investigation, which began with 

 the examination of existing records of lake height as recorded by 

 gage readings, and was continued by the establishment of a number of 

 gage stations in 1896. To understand fully the nature of this investi- 

 gation it is necessary to consider the difficulties that arise from the 

 multifarious motions to which the lake water is subject. 



If the volume of a lake were invariable, and if its water were in per- 

 fect equilibrium uuder gravity, its surface would be constant aud level, 

 and any variation due to changes in the height of the land could be 

 directly determined by observations on the position of the water sur- 

 face with reference to the land; but these conditions are never realized 

 in the case of the Great Lakes, where the volume continually changes 

 aud the water is always in motion. The investigator therefore has to 

 arrange his measurements so as to eliminate the effect of such changes. 



Consider first the influence of wind. The friction of the wind on the 

 water produces waves. These are temporary aud practically cease in 

 periods of calm ; the perpetual ground swell of the ocean is not known on 

 the lakes. The friction of the wind on the water also drives the water for- 

 ward, producing currents. The water thus driven against the lee shores 

 returns in undercurrents, but the internal friction of the water resists and 

 delays the return, and there is consequently a heaping of water against 

 lee shores and a corresponding lowering of its level on other shores. 

 During great storms these differences amount to several feet, reaching 

 a maximum in Lake Erie; in October, 1886, a westerly gale is reported 

 to have raised the water 8 feet at Buffalo and depressed it 8 feet at 

 Toledo. 2 For light winds the changes of level are much smaller, but 

 they are nevertheless appreciable, and they have even been detected 

 in the case of the gentle " land and sea" breezes which in calm weather 

 are created by the diurnal cycle of temperature change on land. 



The water is also sensitive to atmospheric pressure. If the air pressed 

 equally on all parts of the lake surface, the equilibrium of the water 

 would not be disturbed; but its pressure is never uniform. As shown 

 by the isobars on the daily weather map, there are notable differences 

 of pressure from point to point, and within the length of one of the 

 Great Lakes these often amount to several tenths of a barometric inch. 



1 Proc. Am. Ass. Adv. Sci., Vol. LIIL 1894, p. 246. 



2 Science, Vol. VIII, pp. 34, 391. The effect of a storm in October, 1893, is ably dis- 

 cussed by William T. Blount, in Ann. Rept. Cbief of Engineers, U. S. A., for 1894, 

 part 6, pp. 3431-3435. 



SM 98 23 



