414 SHORT MEMOIRS ON METEOROLOGICAL SUBJECTS. 



Since moist air, so soon as precipitation occurs therein, must almost 

 ulvvays experience an impulse upward, because the upward diminution 

 of temperature is seldom so slow as to correspond to that of an ascend- 

 ing moist current of air, therefore it seems to nie that in all precipita- 

 tions the ascending movement of the air plays an important part and 

 increases its intensity. When a cooler air-current begins to flow over 

 warmer air, there first occurs at the surface of contact a slight precipi- 

 tation, above which then the air breaks with strong ascensive power 

 upward into the upper cold air, in consequence of the continually 

 increasing difference of temperature, and thereby loses almost the 

 whole of its aqueous contents. On the other hand, if a warmer current 

 flows above a cold one, this process must be feebler; and therefore the 

 precipitations are less intense when warmer air approaches. 



Since the effect of the precipitations of the rising moist air is to ele- 

 vate the temperature of the upper strata, but to lower that of the lower 

 strata, it follows that a baron)eter at the earth's surface can remain 

 quite unaffected by this re-establishment of the equilibrium in a vertical 

 direction of the temperature; but at a certain altitude, the barometer 

 must, before the rain, stand higher than afterward. The diminution 

 of pressure caused by the abstraction of the elastic force of the aqueous 

 vapor is compensated by the inflowing air of the neighborhood. In ex- 

 tended rains, a relatively warmer air collects above the rain-clouds, but 

 an active ascension of this air can only take place at the edges, where 

 the difference of temperature produces a strong upward impulse, and 

 the cooler air of the neighborhood can rapidly flow in after. Also, in 

 the case of a condensation of vapor over an extended area, the barome- . 

 ter will sink somewhat only in the central portions. We have no ob- 

 servations of this, because any such i^recipitation must immediately 

 produce currents of air from the sides inward ; but with the accession 

 of cooler air, the warming of the higher strata is again compensated, 

 and the barometer will rather rise than sink. The causal relation 

 between the precipitations and the origin and the progress of storms 

 can therefore only consist in this, that the condensation of aqueous 

 vapor gives the air a more or less decided upward impulse,'^ and 

 thereby produces and continually maintains a flow of air from all sides 

 toward the rain-cloud. But so long as no whirl is produced thereby, 

 this process causes no notable change in atmospheric pressure in the 

 horizontal strata of air. Therefore the heaviest rainfalls are, in the 

 equatorial zone, accompanied by no low barometer, nor in our latitudes 

 either, so long as the causes leading to the formation of a whirlwind are 

 wanting. The pressure sinks decidedly only when an extended whirl 

 has formed, in whose center probably the air is rising with increased 

 violence. I therefore believe the diminution of pressure in the center 

 of a storm-area to be a mechanical effect of the whirling movement of 

 the air. This whirl, however, continually receives new force, and is con- 

 trolled by the process of precipitation. I therefore agree with Reye 



