MUTUAL DEPENDENCE OF METEOROLOGICAL PHENOMENA. 337 



perature are seldom local in their occurrence; they 

 are for the most part distributed in a uniform manner 

 over extensive districts. The amount of deviation has its 

 maximum at some one determinate place, in receding from 

 which it decreases gradually until its limits are reached ; and 

 when these limits are passed, great deviations in the oppo- 

 site direction are met with. Similar relations of weather 

 extend more often from south to north than from west to 

 east. At the end of 1829 (when I had just completed my 

 Siberian journey) the maximum of cold was at Berlin, while 

 North America enjoyed unusual warmth. The assumption 

 that a severe winter will be followed by a hot summer, or 

 a mild winter by a cool summer, is a wholly arbitrary one, 

 resting on no foundation. Opposite conditions of weather 

 in adjacent countries, or in two corn-producing continents, 

 are the means of effecting a beneficial equalisation in the 

 prices of many products of cultivation. 



It has been justly remarked that it is the barometer alone 

 which indicates to us what is taking place in all the strata 

 of air above the place of observation, ( 419 ) while the thermo- 

 meter and hygrometer are purely local, and can only inform 

 us concerning the warmth and moisture of the lowest 

 stratum in close proximity to the ground. The simultaneous 

 thermic and hygrometric modifications of the upper regions 

 of the atmosphere (when direct observations on mountains 

 or in aerostatic ascents are wanting), can be sought only by 

 hypothetical combinations, whereby the barometer may indeed 

 serve also to determine temperature and moisture. Impor- 

 tant change^ of weather do not usually arise from a local 

 cause situated at the place of observation itself : their origin 

 is to be looked for in a disturbance of the equilibrium 



