CHANNELS OF TTIK EARLIER STAGE 465 



Creek to Weigand Creek. It forms a distinct terrace in places and there 

 are also numerous gravel-topped knobs. South of Herrick a portion of 

 it has either been let down by the undermining of the Tertiary sands 

 below or been rearranged at a lower stage of the stream. 



Some years ago^ I regarded these gravels as remnants of a high ter- 

 race of the Missouri and not part of the Coleridge channel, an interpre- 

 tation based on a record of elevation at Coleridge which was given 100 

 feet too high. Later when the gravel was traced past Hartington its 

 relations were made clear. In places where the deposit has been cut 

 through by later drainage the gravel remains as a cap on adjacent bluffs 

 and knobs somewhat resembling moraines/^ and they were so regarded 

 by Aughey.* A view of such is given in figure 1, plate 26. 



South of Santee agency the gravel is underlain by a stratum of vol- 

 canic ash several feet thick, which lies on laminated clay containing 

 fresh-water shells.^ 



This ash is shown in figure 2, plate 26, from a photograph taken at 

 D on the map, plate 25. 



At Coleridge, where the Coleridge-Hartington channel passes through 

 the divide at an altitude of 1,550 feet, it is represented by a valley 3 to 4 

 miles wide, with its floor about 150 feet below the adjoining loess plain. 

 This wide valley includes two or three subordinate channels or intervales 

 20 to 30 feet deep, with broad, flat ridges intervening which may have 

 been bars or islands in the ancient streams. A view across the valley at 

 Coleridge is given in figure 3, plate 26. Some features of these old 

 channel gravels suggest that they may have had a similar history to that 

 of the Aftonian gravels, lately described by Professor Shimek,® as exten- 

 sively developed in western Iowa. Xo correlation can be offered, but the 

 difference in altitude appears to show that they are not contemporaneous. 

 It is likely that at the time of deposition of the gravels in the Coleridge- 

 Tlartington diannel the drainage to tlie south was farther west than th(j 

 present Missouri I-Jiver, and ])r()l)ably ]>assed into a lake or series of lakes 

 in eastern Nebraska which may transiently have drained southward into 

 Big Blue Biver." It seems not impossible that both formations are to 

 be connected with the Kansan ice-sheet, and that the thin deposit of till, 

 which is found overlying the Aftonian in places, may belong to the 



2 Bull. U. S. Geol. Survey, No. 158. 



3 See also U. S. Geol. Survey, Geol. Atlas V. S.. Elk Point folio. No. l.ir.. p. o. 

 * Physical Geography and Geology of Xebraslca. p. 2.50. 



■5 Bull. U. S. Geol. Survey. No. 158, p. 70. 



"Iowa Geol. Survey Rejiorts. vol. 20, and Bull. Geol. Soc. Anu'rica. vol. 21, p. SI. 



' See Trans. Kansas Acad. Sci.. vol. 22, p. 107. 



