January 10, 1896.] 



SCIENCE. 



55 



western Illiuois and eastern Iowa appears 

 to have been determined by the ice sheet. 

 The loess is apparently an apron of silt 

 spread out to the south by water issuing 

 from the ice sheet. It is loose textured at 

 the north and becomes finer textured to- 

 ward the south, showing a decrease in the 

 strength of depositing currents. The wide 

 extent of loess over the uplands has led to 

 a consideration of the influence of wind as 

 well as water in its distribution. It is 

 thought that wind- deposited loess may be 

 distinguished from that which is water de- 

 posited. The wide extent, however, ap- 

 pears to be due to water distribution rather 

 than wind. Wind action apparently came 

 into force subsequent to the water distribu- 

 tion and is of minor importance. 



G. K. Gilbert in discussion expressed 

 his gratification at hearing of ' loess ' the 

 rock, instead of exclusively of 'the loess,' 

 the peculiar geological formation. He 

 cited a case in eastern Colorado, along the 

 Missouri Pacific Railroad, where loess had 

 gathered on the leeward side of sand dunes. 

 B. K. Emerson spoke of the aqueous loess 

 of the Hadley meadows in Massachusetts 

 from the annual floods of the Connecticut 

 river, and the eolian loess on the neighbor- 

 ing hills. 



High-level Terraces of the Middle Ohio and 

 its Tributaries. G. Frederick Weight, 

 Oberlin, O. 



This paper embodies the results of the 

 writer's personal observations during the 

 summer and autumn of 1895 on the terraces 

 of the Ohio river, between Steubenville and 

 Marietta, and on the Kentucky river, be- 

 tween High Bridge and Boonetown. The 

 presence of beds of granitic gravel and of 

 isolated boulders of this rock, i. e., of a rock 

 that must have reached its resting place by 

 the agency of ice from the north, in the 

 country adjacent to the Ohio was remarked. 

 An elevated and extensive bed of sand on 



the southwest end of a large island be- 

 tween St. Mary's and Newport was in- 

 stanced as indicating peculiar and as yet 

 not well explained conditions of high water 

 and of a change in the river channel. 



I. C. White in discussion explained the 

 large island as in large part caused by a 

 preglacial channel of Middle Island creek, 

 which enters the Ohio at St. Mary's, di- 

 rectly athwart its course and through a 

 gorge that is continued in the abandoned 

 channel that now forms the island's north- 

 west side. He also stated that pebbles often 

 reached exceptional heights on the hills be- 

 cause the farmers use sand with some con- 

 tained gravel for bedding in their stables 

 and consequently scatter it over their fields 

 at all altitudes. President Shaler also cited 

 the custom among the Indians of cooking 

 with heated boulders, and as the local lime- 

 stones and sandstones were of no value for 

 this purpose they often brought granitic 

 boulders from a distance. Prof. Wright, 

 however, cited boulders of 4,000 pounds, 

 which manifestly could not be explained in 

 these ways. A. Heilprin then mentioned 

 the polished and grooved rocks of South 

 Africa which had been regarded as gla- 

 ciated. More careful investigation how- 

 ever has shown that the polishing is due to 

 the habit of elephants to formerly resort to 

 them and roll and scrape on them, and that 

 the grooves are due to the rubbing of their 

 tusks. F. Leverett corroborated the obser- 

 vations of Prof. Wright in the northern 

 part of the area. 



Four Cheat Kame Areas of Western New York. 



H. L. Faiechild, Kochester, N. Y. 



This paper described three kame areas 

 south of Irondequoit bay and one south of 

 Sodus bay. These are remarkable for ex- 

 tent and quantity of material, as well as for 

 location and altitude ; one of them having 

 gravel hills 400 feet high and furnishing the 

 highest altitude of ground in western New 



