Stevenson.! ^^ [August CO, 



The enormous erosion on the plains and in the tributary canons must have 

 antedated the Turkey Mountain eruption. The basalt rests on the Dakota 

 sandstone between the crater and the Mora canon ; it has flowed into the trib- 

 utary canons, where it still remains ; the stream is almost wholly continu- 

 ous across the plain ; the surface of the sheet is scarcely scarred in the side 

 canons ; while on the terraces of the Mora caiion and on the broad terrace 

 between that and the Canadian canon, the surface is as regular as though the 

 flow had occurred less than 100 years ago. Such facts make it sufficiently 

 clear that since the Turkey Mountain eruption, the rainfall has been com- 

 paratively insignificant. But this season of dryness must have been pre- 

 ceded by one of much greater rainfall, unless, indeed, we conceive that the 

 erosion had been continuing for an indefinite period. But a moment's 

 consideration will show that such a continuance is impossible ; remnants of 

 the Colorado shales remain on the plains as mesas nearly 1000 feet high, 

 which have been protected from erosion by a thick plate of basalt. This 

 plate, which covers the Canadian hills east from the Turkey mountains and 

 is an outlier of a vast basaltic area, marks the occurrence of a previous 

 outpouring, not earlier than the beginning of the Pliocene. The enor- 

 mous erosion, then, must have been performed during the interval between 

 the eruptions, which, in a geological sense, could not have been very long. 

 There seems, therefore, no room fo» doubting that a great cliraatal change 

 passed over this region ; that during the interval between the eruptions the 

 rainfall was great, sufficing for the removal of the thick Colorado group and 

 the digging out of the many imposing arroyos or side-canons, but that 

 since the later Pliocene the region has been arid, the corrasion of the canons 

 having been done by water drained from the mountains. 



But the new canons do not coincide with the old ones. A small bit of basalt 

 caps a hill in the Canadian canon, at somewhat more than 2 miles above the 

 mouth of Mora river ; within a short distance lower, the canon bends west- 

 ward, and the thick bed of basalt is well exposed, covering a terrace on 

 the south wall ; soon the gorge curves southward and becomes very narrow 

 as it passes through the basalt, which caps both walls, its base being nearly 

 200 feet above the stream. But within one-fourth of a mile, a slight bend 

 eastward carries the caiion beyond the basalt, which now lies only on the 

 western wall and covers a bench, stretching thence to the canon of Mora 

 river. A fragment of basalt remains in the former canon on an isolated 

 hill almost immediately below the mouth of Mora. 



The present canon of Mora river has the basalt altogether on its north- 

 easterly side for a distance of nearly a mile and a half, and a fragment of 

 the old westerly wall still remains as an irregular conical hill, rising above 

 the basalt terrace. At a little way above this, the basalt has been wholly 

 eroded from both sides of the canon, excepting only an insignificant and 

 badly broken patch on the south side; but immediately above this, the 

 gorge becomes close, and the basalt is shown on both sides. 



It is evident, then, tlrat the caiions have been re-eroded partly along the 

 original line or through the basalt, and partly alongside of the ancient 



