Valley Loess of Lansing, Kan. — Uphani. 29 
and all the others of this liasin were much bigger and much 
muddier. Though flowing with slackened currents on account 
of their decreased slope, they carried the fine particles of the 
loess hundreds of miles down the ^Missouri and Mississippi 
vallevs, building up wide and thick floodplains as far south- 
ward as to Crowley's Ridge in Arkansas, and even onward 
past A'lcksburg and Xatchez, thinning out on the shores of the 
Gulf. 
During the summers of each year the floods pouring along 
the valleys from the ice melting and rains added a little to the 
surface of the whole floodplain ; but in the autumn, winter, 
and spring, the diminished rivers flowed in comparatively 
narrow channels, probably permitting the main part of the 
floodplain to become more or less covered by grass and other 
vegetation and to be inhabited by air-breathing mollusks. 
Hence, even in the valleys, the fossil shells of the loess are 
mostly of terrestrial species. If a quantitative estimate of the 
average rate of deposition of the loess in the vallevs be at- 
tempted, I think two or three inches of added thickness in 
the three or four summer months of the river floods in each 
vear to be perhaps a reasonable supposition, which would re- 
quire a thousand }'ears, approximately, for the accumulation 
of the depths of 150 to 250 feet. 
From the verv extensive tracts of the floodplains, several 
or many miles broad and reaching along the valleys hundreds 
of miles, the winds in the cooler part of the year, when these 
great expanses were bare, blew away much of the Fme loess 
dust and spread it far and wide over the interfluvial higher 
lands. Received there on a grashv surface, like that of our 
present prairies, the loess settled from the gales and mantled 
somewhat uniformly all the land, regardless of its contour or 
altitude. In these great areas of eolian loess only terrestrial 
shells are found. Here again we may essay an estimate of the 
rate of deposition, assuming that the wind action might bring 
on the average a sixth or a quarter, or even a third, ot an inch 
in thickness vearly, so that Vv'ithin the same thousand years of 
the valley loess formation the observed interfluvial loess sheet, 
mostlv TO to 25 feet thick, would be distributed in its surpris- 
ing uniformity over the high and low lands alike. 
Nor \xeve the winds limited to the river floodplains for the 
derivation of their loess dust. On the drift-envelopcrl border 
