VNITED STATES. 



The countries to the north are not less exposed' 

 to these vicissitudes, but there is this difference, in 

 the southern states the sudden variations sire chiefly 

 from hot to cold, while in the northern they are more 

 frequently from cold to hot ; so that, in the latter, 

 the effects produced on bodies generally arise from 

 dilatation, in the former, they are commonly owing 

 to con. triction. 



On comparing the therm ometrical tables of the dif- 

 ferent places of which I have just been speaking, 

 and in making daily observations myself on the 

 variations of the atmosphere, I could not avoid 

 perceiving a constant harmony between these va^ 

 riatinns and certain winds, wluch are uniformly 

 connected with them. The transitions from cold 

 to heat, I always found, took place with changes of 

 the wind from north-east or north-west to south- 

 east or south; and, on the contrary, the transitions 

 from heat to cold always occurred with changes of 

 the wind from south or south-east to north-east or 

 north-west ; and this was the case from Florida to 

 Canada and Hudson's bay. 



Hence we have one element of a theory applica- 

 ble to all the problems of this climate ; but as good- 

 theories are nothing but a systematic arrangement 

 and combination of all the facts of one kind, I shall 

 Dot be in haste to solve these problems by isolated 

 facts, but shall proceed to bring forward many sin- 

 gularities, which, at first sight, would appear to be 

 exceptions. 



The climate of the basin of the Ohio, and of the 

 Mississippi, is less cold, by three degrees^ of Utitudg^ 

 tlian that of the Atlantic coast, . 



