539 



must suppose, that these hailstones at the mo- 

 ment of their formation are larger in the tempe- 

 rate than in the torrid zone. We yet know so 

 little of the conditions, under which water con- 

 geals in a stormy cloud in our climates, that we 

 cannot judge whether the same conditions be 

 fulfilled under the equator above the plains. I ? 

 doubt hail's being always formed in a region of 

 the air, of which the mean temperature is nought, 

 and which with us is found in summer only at 

 one thousand five hundred, or one thousand six 

 hundred toises of height. The clouds, in which 

 we hear the rattling of the hailstones against one 

 another before they fall, and which move hori- 

 zontally, have always appeared to me much less 

 elevated : and at these smaller heights we may 

 conceive, that extraordinary refrigerations are 

 caused by the dilatation of the ascending air, of 

 which the capacity for caloric augments ; by 

 currents of cold air coming from a higher lati- 

 tude ; and above all, according to Mr. Gay- 

 Lussac, by the radiation from the upper surface 

 of the clouds. I shall have occasion to return 

 to this subject, when speaking of the different 

 forms, under which hail and hoar-frost appear 

 on the Andes, at two thousand and two thousand 

 six hundred toises of height ; and when examin- 

 ing the question, whether we may consider the 

 stratum of clouds, that envelopes the mountains, 

 as a horizontal continuation of the stratum, 



