Glacial Geology in America. — Fairchild, \yj 
of englacial drift, except near the base (see page i68) and re- 
gard even the highest moraine hills as due to piling of sub- 
glacial or basal drift by the thinner ice edge overriding its own 
debris, 
Eskers. — The name "osar," borrowed from European lit- 
erature, was applied by Edward Hitchcock to masses of water- 
laid drifts as early as 1842, and appears frequently in subse- 
quent writings upon the drift. The word was generally used 
without close discrimination, however, being applied to irregu- 
lar mounds of gravel and sand as well as to ridges. An early 
identification was made by C. T. Jackson, in 1843, who said 
that the European osars were identical with the "horsebacks" 
of Maine. During subsequent years down to 1880 the word 
"kame" was generally employed instead of osar. The name 
"serpent kame" was applied by Shaler in 1888, in print. 
The term "esker" was first used by G. H. Kinahan in 1863, 
being applied to gravel ridges in Ireland, but in America it 
remained for W J McGee to make, in 1881, the needed dififer- 
entiation between eskers, applied to elongated ridges of gravel 
and sand, and kames, designating sand or gravel deposits of 
irregular or morainic topography. Gradually the word esker 
has displaced the less convenient word osar. although the lat- 
ter has priority. 
Since the acceptance of the glacial theory it has generally 
been recognized that these ridges of gravel, under whatever 
name, were formed by glacial streams. In 1872 N. H. Win- 
chell attributed the gravel ridges observed by him in Ohio and 
Minnesota to the work of streams in longitudinal crevasses, 
and a more definite explanation in America was made by War- 
ren Upham in 1876, when he referred those in New Hamp- 
shire to glacial rivers, either supraglacial or subglacial. 
■ The genetic distinction between eskers and kames, made 
by McGee in 1881. was emphasized and amplified by Cham- 
berlin in 1883 and 1884, referring the esker phenomena (using 
the term osar) to the class of axi-radiant or longitudinal 
stream drift, as distinguished from the peripheral kame de- 
posits. That eskers are constructive forms, deposited in the 
beds of glacial streams, in comparatively stagnant ice, seems 
to be universally admitted. There remains, however, some 
lack of unanimity as to the attitude of the streams, whether 
