I 



431 



at tho close of spring the teniporaturo of the external aii' and ol 

 that in the rents is the same, no pai-ticular occurrence takes place ; 

 but as soon as the wave of cold begins to make impression on 

 tho rocky parietes of tho fissures, then tho air in them will be 

 somewhat cooled, contracted, and rendered specifically heavier. 

 This being so, tho weight of the column of air in these rents will be 

 greater than that of a column of equal altitude of the external at- 

 mospheric air, and the consequence will necessarily be, that the colder 

 air will descend, tho warmer atmospheric air from above will supply 

 its place, which, in its turn, will bo cooled and descend, and thus a 

 current of cold air through the crevices into and through the cavern 

 will be established. As the temperature of the rocky parietes 

 gradually falls with each successive wave of cold, the air in the 

 fissures will become colder and colder, and in the same proportion 

 will descend more rapidly. 



But the rapidity of descent does not only depend upon the increas- 

 ing coldness of tho air in the fissures, but is further augmented by 

 the warmth of the summer expanding the external air, so that the 

 difference of weiaht between the external and internal columns be- 

 comes greater. In the manner now explained, a current of cold 

 air is constantly descending and flowing through the cavern, produc- 

 ing all the surprising frigorific effects displayed within it. 



That such a current does exist, Mr Murchison gives a satisfactory 

 proof; he says, " That upon unlocking the frail door of the cavern, a 

 volume of air, so surpassingly keen, struck the legs and the feet, that ho 

 was glad to rush into a cold bath in front of him to equalize the effect." 

 This downward current will continue the same till the close of au- 

 tumn, when its course comes to be changed ; by that time the first ap- 

 proaches of May's surface warmth will begin to be experienced, the cold 

 of the sides of the rents begins to diminish, and the temperature of the 

 external air must have fallen to nearly that of the internal current. 

 As soon as an equality between the temperatures and densities of tho 

 extei'nal and internal columns shall have been established, all current 

 must cease. At this period, namely, tho commencement of winter, 

 tho wave of the summer's heat begins to reach both the walls of the 

 air-channels and of the cavern, and gradually communicates a warmth 

 which progressively elevates the temperature, and dissipates every 

 mark of tho preceding summer's congelation. It might at first 

 be reasonably expected, that at this time the preceding order of things 

 would be reversed, and that a current iu the opposite direction would 



