54" THE NATURAL HISTORY OF CLOUDS. 



^w\s^ be certain pf, until the current, in its course of propa- 

 gation downward, had betjun to afiect the densei- clouds, 

 thrown up by the superficial evaporation. It is not very 

 uncoinrnon to see the cirro-stratus evidently brought'by a 

 wind, moving in a different direction from that wherein the 

 cumuli are immersed on which it settles. In this case the 

 latter are speedily arrested by it, ar,d assume the new course, 

 or descend in rain, by a change of their electricity. 



Of the Nature of the Cirro-cumutus. 



Nature of the i<et us now reverse the former case, and consider the 

 c»rro-cumulus, upper current as both vaporised, arid warmer than the air 

 below. 



it is probable, that the upper is then cooled by that par^ 

 of the lower which is next to it, though very slowly, from 

 the difficult transmission of caloric downward. The decom- 

 position of the vapour in the upper current by this means 

 may give origin to the cirro-cumulus ; and the peculiar ag- 

 gregation of this cloud, »s distinguishable from that of the 

 tirro-slratijs, rnay be the result of j*s acquiring electricity 

 in its {jescent in a much greater de||^ee. Such, at least, ^ 

 the inference we may deduce froai its abundance before 

 thunderstorms; when it is occasionally seen to arrive with 

 the wind in extensive fiocks or strata, moving with unequal 

 velocity, and by consequence overtaking each other, until 

 they form a den&e stationary mass. 

 Judications. This explanation of the origin of the cirro-cumulus is 



principally deduced from an observation, which we have no^v 

 so often repeated, as to regard it as a meteorological axiom ; 

 that the temperature of the day following, exceeds that of the 

 ddy on which it appears. Hence, when it continues to recur 

 daily, the weather still grows warmer, until a thunder-storm, 

 in some quarter of the heated tract, puts a period to the in- 

 sulation of the clouds. 



Of the Nature of the Cumulo-stratus, 



In attempting to assign causerto phenomena so compH- 

 sated, as those which this modification presents, we may' be 



in 



