S. V. Wood— Origin of the Loess. 393 
The Loess of the valleys of the Mississippi and its northern 
tributaries occupies a similar position, relatively to the area covered 
by the great ice-sheet from which the glaciation of the United States 
proceeded, that these beds of Prof. Kerr do; and there is, I think, 
a strong presumption of their identity. Although Pumpelly has 
advanced the dust theory for it, the opinion prevailing among 
American geologists has been that the Loess of America was of 
fluviatile, or lacustrine origin, during or subsequent to the Glacial 
period. As regards a river occupying these valleys up to the 
elevation to which the Loess extends (which is said to be 500 feet 
above the rivers), then, whether we imagine such to have arisen 
from the effluent water of the land-ice which overwhelmed the 
parting between the St. Lawrence and the Mississippi systems 
during the accretion of the ice, or whether we imagine it to have 
arisen from the annual flood during the dissolution of this ice, the 
objection presents itself that it would have been a permanent flood, 
during summer at least, and the area occupied by it could not have 
furnished vegetation for the land shells’ subsistence. Further, as 
neither the floods during the wane of the ice-sheet, nor the effluent , 
water from it during its accretion, have in New England, according 
to the opinion of American geologists (Dana and Upham) who have 
studied the question, produced anything but thick accumulations of 
sand and gravel, without any shells or other organic remains (save, 
perhaps, some of Reindeer) such as occur in the Loess, it is difficult 
to see how any similar action could have given rise to a formation 
so very distinct from this as is the Loess of the Mississippi drainage 
area. As regards its origin from a lake, then, irrespective of the 
physical difficulties besetting the existence of a lake maintained at 
the level up to which the Loess extends, while the lower part of the 
Mississippi Valley was depressed relatively to the upper as much as it 
is now (which Prof. Hilgard’s description of the Port Hudson group; 
and late gravel beds, in the States of Mississippi, Louisiana, and 
Texas seems to show), an unstratified formation in which land shells 
for the most part only occur could hardly have resulted from it ; 
for the shells of a lake would have been almost exclusively of fresh- 
water habit, and shells of the great American genera Unio and 
Anodonta would have abounded everywhere in it. Freshwater 
shells, such as Limneea, do occur occasionally; and adjoining the 
rivers the unstratified Loess with land shells is said to overlie and 
pass down into stratified beds with freshwater mollusea. Sir C. 
Lyell (Antiquity of Man, p. 237) states that by his own observation 
he found this to be the case at Natchez, and that it occurs also in 
the valley of the Rhine. He, also, considered the Loess of the 
Mississippi Valley ‘from its homogeneous nature, the absence of 
stratification, and its terrestrial and amphibious shells, to have been 
formed: by a great river which, like the Nile, inundated the wide 
plains bordering it on either side.’ The Nile flood does not; 
however, rise above the ordinary river stream to the tenth part of 
the height to which the Loess is said to rise above the streams 
of the northern affluents of the Mississippi; and if an unstratified 
