METEOROLOGY. 



23 



and three years' continuous observations at Fort Pierre, with one also at Fort Randall, furnish 

 much more reliable data. The mean results for winter are given in the table below.* 



A comparison of these winters is made with those at the nearest points where long series 

 have been obtained, from which it appears that the mean of the three winters at Fort Pierre 

 is probably near the true mean, since the same winters at Fort Kearney vary only 0.60 from 

 that of six years at that post, although each separately varies widely. 



But the observations at Forts Clark and Union in 1822-'23, are not so easily reconciled- 

 The same years observed at Fort Snelling, the nearest post, show that entirely diflerent 

 climatic conditions affect the three places. 



The only conclusion from the comparison is, that Fort Union partakes of the milder winter 

 climate of the western coast, as observed still more strikingly at Fort Benton. Forts Kearney 

 and Laramie occupy a similar relation to each other and to the Pacific climates, and the long- 

 series at those posts show that the former has a winter temperature 7°. 10 colder than the 

 latter, while the periodical curves follow nearly the same course, keeping from four to ten 

 degrees below. A similar relation of climate may be expected to prevail between Forts Union 

 and 'Benton. But Fort Clark shows two winter months much colder than for the same period 

 at Fort Snelling, which, together, were 3°. 85 below the average of thirty-five years. This, 

 together with the great difi"erences in the three winters observed at Fort Pierre, favor the 

 conclusion that this portion of the Missouri lies intermediate between the two sj'stems of 

 climate (for winter) east and west of it, and that the prevalence of winds from either quarter 

 determines the character of the weather. Both extremes seem to be represented by the Fort 



« Mr. Blodgett's book gives the isochimenal line for this part of the Missouri five degrees higher than he had fixed it in 

 the charts of the Surgeon General's Meteorological Register of 1855, but he does not give any data for the change, and the 

 former appears most correct. 



