[ 237 ] 100 



By 80 observatious at noon, mean difference of level - 362 feet. 



Reduction for St. Louis to the Gulf - - - 382 " 



Altitude of the Mississippi, at low water, near St. Peter's, 



abovp. the Gulf of Mexico .... 744 <; 



The accordance of these detailed results appears to me satisfactory, 

 when account is taken of the different circumstances under which these 

 elements have been obtained. The observations of the month of October 

 present the greatest discrepancy. The autumnal season was then too far 

 advanced, during which the oscillations of the barometer begin to be 

 rapid. It is, therefore, very rarely that 1 have made use of barometric 

 observations taken at this period of the year, except when the state of the 

 atmosphere (as is sometimes the case in the west during the fall) was ex- 

 tremely favorabl<?. 



But, leaving this point, 1 consider that these two principal stations pre- 

 sent results in altitude sufficiently exact to answer the cedls of geography, 

 of geology, or of general topography. 



These stations of the second order, in reference to St. Louis, have thus 

 served as a base for a greater number of stations of the third order. All 

 the heights observed in the upper regions, to the south and west of St. 

 Peter's, have been referred to the barometer of this station ; and all the 

 heights taken in the region of the northwest, in the great valley of the Mis- 

 souri, and on the western slope of the Coteaii des Prairies, have been re- 

 ferred to the station near Council Bluffs. 



In regard to the stations of the third order, I must remark, that there are 

 many whose altitudes are determined, not by a transient reading of the 

 barometer, but by sets of observations taken during my stay at such sta- 

 tions, both for this purpose and for the astronomical part of my work, or 

 for the reconnoissance of the adjacent country. Such are the stations at 

 the Traverse des Sioux, Lac qui-parle, lake Shetek, Spirit lake, the sources 

 of La Hontan river, Okamanpidan lake, the head of the Coteau des Prai- 

 ries, lake Mini-wakan, at Otuhu, or Talle de Chene.s, on the Tchan-san- 

 san river. Fort Pierre Chouteau, Huppan-Kutey prairie, Cedar island. Five 

 Barrel island, on the Missouri, and some others. And in 1836, during my 

 campaign to the sources of the Mississippi, I stopped, with the same object, 

 at Crow wing river, at Guyashk lake, at Kadikomeg lake, Sandy lake, 

 Leech lake, Itasca lake, and elsewhere. 



Further, as to the relative level of the ground, and its irregularities 

 which we passed over in our daily marches, the method pursued has 

 been this : " We observed the meteorological instruments in the morning, 

 before breaking up the camp ; we also observed them in the evening, 

 when arrived at a new spot, where we were to pass the night ; thus in- 

 forming ourselves of the barometric range during the night, and obtain- 

 ing some premonitions of the state of the atmosphere to be expected on 

 the morrow. During the day, wherever the soil presented to us character- 

 istic undulations, ridges, hills, hillocks, or btittes, or deep glens enclosing 

 some stream, we halted to observe for their relative differences of level ^ 

 i. e„ we measured the highest points in reference to the general surface 

 of the plain at their base ; and the surface of plateaux bordering rivers, in 

 reference to the bed of the stream; then, to reduce these levels, so ob- 

 tained, to that of the sea, I could refer them to those which had been 



