278 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



For these reasons the present paper includes the whole of northern 

 Europe. Of course it was not possible to make a complete study of 

 the Quaternary literature of Europe, which can not fall far short 

 of a hundred thousand papers, etc., in more than a dozen languages. 

 The working bibliography actually collected numbered several 

 thousand entries, but the references cited in this paper include most 

 publications of importance bearing on the subject. 



There are a few points of detail arising in connection with the 

 correlation which may be briefl}^ referred to. The first is the color 

 of deposits. W. O. Crosby (1)^ found that in the northern United 

 States and Canada the soils are almost universally brownish or yel- 

 lowish, but not red, except Avhere they result from the disintegration 

 of a red rock. On the other hand, in the southern United States, 

 the red color greatly predominates over browns and yellows, and in 

 the West Indies and South America the redness of the soil is even 

 more intense and universal; the red lateritic aspect of soils in the 

 Tropics is well known. The difference is more or less distinctly 

 observable in all longitudes and in both Northern and Southern 

 Hemispheres. The brown, yellow, and buff colors of northern soils 

 are due to the presence of yellow ferric hydrates like limonite; the 

 red color of southern soils, though usually attributed to hematite, is 

 probably mostly due to the red ferric hydrate turgite. This differ- 

 ence depends not on underlying rocks but on climate. Crosby con- 

 tinues : 



Ferric hydrate, the coloring agent of northern soils, is dehydrated at the 

 lemperature oi' boiling water, and it seems proljisble that a partial, if not com- 

 plete, dehydration may result at much lower temperatures, if unlimited or 

 geologically long time is allowed. 



In the Southern States the red color is only superficial, extending 

 to a depth of 2 to 10 feet, and passing through orange and gray to 

 the natural color of the rock. Thus the redness of a soil depends 

 both on its age and the temperature at which it was deposited, and in 

 this we find a cause of the red color of many very old bowlder clays, 

 noted and used as a means of correlation by F. Leverett (2), J. van 

 Baren (3), C. Gagel (4), and others in opposition to the prevailing 

 blue-gray color of later bowlder clays, weathered brown at the sur- 

 face. In extremeh' calcareous old clays like the chalky bowlder clay 

 of East Anglia this rule does not hold, but the disintegTation of the 

 granitic rocks forms an equally reliable index of age. Erratics of 

 granite in the chalky bowlder clay of Hertfordshire and Finchley 

 frequently fall to a granitic sand at a touch, though they must have 

 been sound when they were incorporated in the ground moraine. 

 The same state of decay has been noted in the granitic pebbles of the 



1 Numbers in parentheses refer to bibliography at end of paper, 



