DAVIS. — ROCKY MOUNTAIN TERTIARIES. 353 



pends on the accidental compensation of several imperfectly related 

 processes ; namely, marginal and central deposition, marginal and cen- 

 tral deformation, and lowering of level by erosion at the outlet, or in 

 interior basins by evaporation. It may be pointed out that the pres- 

 ervation of open fluviatile plains during the slow deformation of the floor 

 on which they are accumulating does not involve the balance of conditions 

 or processes so little related as in the case of extensive shallow lakes ; 

 for the rate of fluviatile deposition responds most delicately to any 

 deformation of the fluviatile plain. If slow local depression occurs, the 

 rate of aggradation increases and the swinging rivers build up the de- 

 pressed area about as fast as it sinks. If slow elevation occurs, the 

 swinging rivers tend either to degrade the uplifted area, or to con- 

 centrate' their deposits on the districts that are not uplifted and thus 

 maintain a generally even surface. 



If a shallow lake of large extent should happen to be formed, it would 

 probably soon change by depression to a deep lake or to an arm of the 

 sea ; or by elevation, marginal deposition, climatic change, and erosion 

 of outlet to a land surface. Hence the deposits of shallow lakes are 

 likely to be followed either by deep lacustrine deposits or by marine 

 deposits, each recognizable as such ; or by fluviatile and other sub- 

 aerial deposits, particularly around the margin of their basin. The 

 converse of these statements does not seem to hold true; for an open 

 fluviatile plain is more likely to maintain itself as such than a shallow 

 lake is to persist as a broad and thin sheet of water. The deposits 

 of a fluviatile plain are therefore not likely to be associated with those 

 lakes, unless in interior basins. 



If a review is now made of the descriptions of the so-called lacustrine 

 formations in the Rocky mountain region, frequent mention will be 

 found of strata that are strikingly unlike those which might be expected 

 fi'om the observation of existing large lakes and from the reasonable 

 extension of principles based on such observations ; but before giving 

 examples of this kind, it should be remarked on the other hand that 

 some of the deposits are entirely consistent with a lacustrine origin. 

 For example, King reports that there are even-bedded " paper shales " 

 of very fine texture and with numerous fossil fish in the Green river 

 basin, "giving general evidence of accumulation in still, rather deep 

 water" (40 Par. Surv., i , 447). Russell writes regarding the basin 

 of Lake John Day in the northwest : " Before the Columbia lava was 

 broken and tilted ... its surface over the whole of central Washington 

 and probably far into Idaho and Oregon, was covered by the waters of a 

 VOL. XXXV. — 23 



