Acciiiii iilntljni of iJi'iiiiil liii<. — I j)hiim. 339. 
obvious that no diroct comparison can be made, but while these 
results fail to show relationship between the amlier or fossil 
resin and the cannel coal, the}^ do not enable us to say that the 
latter did not have its origin in vegetable structure. It is (^uite 
possible that the greater degree of insolubility of the coal ma}- 
be the result of a greater degree of alteration. 
EXI'LAXATIOX OF THE PlGUKKS. I'l.ATK X. 
1. Coal showing rods in longitudinalsection, togetlier with granular 
matter lying between, X 6. 
2. Filament of coal in longitudinal section showing u snuill area 
occupied by transverse shrinkage cracks, X 70. 
CONDITIONS OF ACCUMULATION OF DRUMLINS.* 
By Wauuen Upiia.m, Soiiiorville, Mas^. 
CONTENTS. 
Page. 
Statement of the Problem 339 
Definition and Varieties of Drinnlins 339 
Oeographic Distribution 342 
Siibglacial and Rapid Deposition shown by Strncttire 34() 
Questions concerning tlie Action of the Ice-sheet 3.51 
Probable Accumulation of tlie Drumlins from Bnglacial Drift 3.51 
Objections to former Theories 351 
Transportation of Drift into the lower part of the Ice-8heet 353 
Ablation causing Enulacial Drift to become Supery;lacial 354 
Stratum of Superglaciiil Drift made again Englacial by increased Snowfall 
and by Advance of tlie thicker portion of the Ice-sheet 3.5.5 
Ice Currents massing tlie Knglacial Stratum into Drumlins 356 
Review of Objections to this Explanation 3fj8 
Comparison with Terminal Moraines, Rames, and Eskers 300 
View of the Ice Age as One Cilacial Epoch 361 
StATK.MEXT of THE PROBLEM. 
Definition and Varicfirs of Drninlinx. — The drift hills called 
drumlins consist, at least superficially and in most cases through- 
out their entire mass, of till or boulder-clay, being unstratified 
clay, sand, gravel, and boulders, mingled indiscriminately to- 
gether, which therefore must be attributed to deposition by ice 
without modification by the assorting and stratifying action of 
currents of water. The}' have usually an oval form, with smoothly 
*This pai)er, in preliminary outlines, was ])resented bel'oi'e the (leo- 
logical Society of America, Aug. 1(5, 1892, at the meeting in Kocliester, 
N. Y., as noted in thcOctober (ikoloijist (p. iMS). Furtlier study lia.s 
since led to tlie opinions given here, that t\w englacial drift of wliich, 
under this view, tlie drumlins apjx'ar to liavr hpenciiielly formed, had 
become superglacial i)y al)lation and was aftiM-ward enclosed as ji 
stratum within the ice-sheet, being thence aniasseci in these hills; 
and that, according lo such explanation of the origin of (he drumlins 
and their comparison with the terminal moraines, t he Ice age^may 
probably have comiirised only a single glacial epoch. 
