GEOLOGICAL SURVEY OF THE TERRITORIES. 295 



saline matter, and at many points on each bowlders are quite plenty. So 

 nuieli so is this the case with tlie Coteau of the Jlissouri tliat Mr. 

 Sliiimer remarks, in regard to the section where his Hue crossed it, that 

 " it is very stony." This elevated phiin will furnish us with the explan- 

 ation of the great bend of the Missouii in the northwest part of Dakota, 

 and its <lirection southeast from that point. 



Pembina Mountain, which is situated within the belt now under con- 

 sideration, and near the international boundary-line, is nothing more 

 than an elevated plateau, similar in character to those already described, 

 but much smaller. 



1 may remark here that the country northward, even as far as the 

 Saskatchewan, along this meridional belt, preserves about the same 

 character that we see exhibited in Dakota. I was at first under the 

 impression that when we entered the valley of the Assiuiboine we would 

 find a section sup[)lied with more moisture and ranker vegetation tiian 

 further south, but from information received from persons who have 

 long lived in that region I am satisfied it is but a repetition of the tyi)e 

 we see along the northern border of Dakota. 



Eecurring again to the vertical topograi)hy, I would call attention to 

 the fact that after we leave the valleys of Ked and Minnesota Rivers we 

 notice a gradual westward ascent, and in order to obtain a correct idea 

 of these sloi)es I estimate by the lowest water-levels, whenever these 

 can be obtained, and where they are wanting I use the lowest land- 

 levels, or such as represent the average elevations of extensive plains 

 or valleys. Following this rule we find that the ascent along the North- 

 ern Pacific Railioad, west from Red River to the Missouri River, averages 

 a little over 4 feet to the mile. From Lake Traverse to the Missouri 

 Eiver, at Fort Sully, the average ascent is not more than 2 feet to the 

 mile, and from the bend of Minnesota River to the Missouri, at the 

 mouth of Big Sioux, it is not more than IJ feet to the mile : but in this 

 case the land ascent, until we reach Le Mars, is about 2 feet to the mile. 



The plains immediately east of Cheyenne River, and between it and 

 James River, and west of the latter to the Coteau of the Missouri, have 

 an average elevation of about 1,450 feet above the sea, and from DO to 

 150 feet above the streams which traverse this region. 



Devil's Lake appears to be situated on a plateau or swell forming the 

 divide between the Cheyenne River and the northern tributaries of Red 

 River. 



It has an elevation of 1,407 feet above the sea, or about 100 or 150 

 feet above the level of Cheyenne River at the point immediately oppo- 

 site. But it is a general rule, as I have before stated, that the lakes in 

 the prairie-portion of the iSTorthwest are situated on the divides between 

 the streams. Lake Eckelson, on the line of the Northern Pacitic, be- 

 tween Cheyenne and James Rivers, has an elevation of 1,418 feet above 

 the sea — very nearly the same as Devil's Lake— 188 feet above Chey- 

 enne River, and 25 feet above James River. 



The leading ridges and land-swells', so far as my observations in the 

 southern and northern part of the Territory have extended, ap[)ear to 

 run nearly north and south, the direction being slightly west of north 

 and east of south. But, as will be seen by reference to a map of the 

 country, this corresponds with the general direction of the streams; 

 and, as the drainage is southward from latitude 48°, we may infer that 

 there is a gradual descent southward as well as eastward. No sufficient 

 data have yet been obtained for determining accurately the rate of this 

 descent to the south, yet we may approximate to it from the following 

 items : James l>iver, at the crossing of the Northern Pacific, (latitude 



