May 14, 1897.] 



SCIENCE. 



769 



main steeams is similar to that between the 

 gray loamy clay and the ' loess ' of Missouri. 

 The upland deposit is similar in composition to 

 the typical loess of the valleys, except that it 

 contains a large constituent of fine clay parti- 

 cles which partially bind the rounded silt grains. 

 This deposit of unusually argillaceous loess, 

 besides mantling the upland ridges, descends 

 the slopes of the valleys and is present over the 

 thick terrace-like deposits of true loess. They 

 are separated by a smoothly undulating but 

 quite sharp line. This line is present through- 

 out northwestern Illinois and in Iowa at Du- 

 buque. Without discussing the significance of 

 this line, I will call attention to the fact that 

 Chamberlin and Salisbury, in making the study 

 of the loess of the ' Driftless Area, ' recognized 

 the distinction between the silt deposits along 

 the main streams and on the uplands remote 

 from the main water-courses, particularly on 

 the east side of the Mississippi river ; and in 

 reference to their origin they say, on'pages 306-7 

 of the 6th Annual Report of the U. S. Geolog- 

 ical Survey, " the loess-depositing waters were 

 neither true lakes nor true rivers, but assumed 

 an intermediate fluvio-lacustrine character, pos- 

 sessing sufficient onward flow to prevent the 

 deposition of a large proportion of the clayey 

 constituents of the silt they bore. * * * 

 The loess probably represents the debatable 

 ground between the two " (a lake and a river). 

 "The coarser stratified portion along the im- 

 mediate valleys of the great streams seems 

 clearly on the fluvial side of the line, while the 

 broad, loam-like mantle apparently lies on the 

 other side." It is this comparatively thin but 

 uniform mantle of argillaceous loess or loam, 

 on the uplands and in the valleys, for whose 

 origin I claim a purely lacustrine or possible 

 semi-marine character, largely because, while 

 clearly water-laid, it passes within one-half 

 mile and less from the top to the base of hills 

 100 and even 200 or more feet in height. It 

 mantles a region having a range in altitude of 

 600 or more feet. 



With special reference to the question of the 

 mode of formation of the gray loamy clay of 

 Missouri, which I regard as a portion of the 

 Loess formation, as also does Professor Todd, I 

 ■wish to make the following statements, which. 



I believe, can easily be demonstrated to be 

 facts : 



The drift sheet of northern Missouri was not 

 laid down on a perfectly level plain, but on an 

 undulating upland, dissected by valleys of 

 which the Missouri and Mississippi were ex- 

 amples. The Missouri valley, at least from 

 Boonville onward, is preglacial inage, and it is 

 hardly necessary for me to add that the Missis- 

 sippi valley along the entire east border of the 

 State is so also. These valleys do not now have 

 much drift because it has been removed from 

 them by subsequent erosion. 



The line of Professor Todd's supposed ' bar- 

 rier ' or ridge from the Osage-Gasconade 

 divide to Pike county, Illinois, being traversed 

 by two important preglacial valleys — the Mis- 

 sissippi and Missouri — loses its effectiveness to 

 account for the difference in altitude of the 

 loess plains of northern Missouri and southern 

 Illinois. 



The gray loamy clay or ' upland loess ' of 

 northern Missouri was laid down on an undu- 

 lating plain, some portions of which have a 

 decided slope. The ' fluvial ' theory of its 

 origin, which supposes that the area upon 

 which it was being formed was dry during a 

 large part of the year, and, therefore, that it 

 was deposited somewhat like the alluvial silts 

 of the Mississippi delta at the highest flood 

 stages, is opposed by the inequalities, when 

 considered over broad areas, of the surface of 

 the sheet of loess or loam. Rivers do not 

 wander about except on plains which have an 

 even and very slight slope. In order to uni- 

 formly sheet with water-laid silt even a very 

 slightly hilly region, a lake-like body of water 

 is required. That which covered northern 

 •Missouri at the time the ' upland loess ' was 

 deposited may not have exceeded 50 or 100 feet 

 in average depth, and may not have covered the 

 entire loess region at one time, but its nature 

 was decidedly different than that of a river. 



The failure to find beach ridges is far from 

 being positive evidence of the absence at any 

 former time of the shore line of a great lake, 

 or even of the sea. For instance, the ocean 

 waters in advancing over Florida to form the 

 Columbia sands which mantle the northern por- 

 tion of the peninsula, and later in withdrawing 



