Rfwaiis on the 7'"e:,i;j of Cold ^ubiei raimm IViiiJt. 235 



by heat, and will dry the fides of thefe caverns. Whence he thinks that the air may be 

 confidered as fufficiently diy to produce an evaporation which may cool it three degrees. 



After having given fo Interefting and accurate a detail of fadls on this curious fubjei^, 

 from the viork of a philofopher to whom the world is fo greatly indebted, I have thought 

 it a point of juftice to relate his theory in a manner fcarcely lefs copious, and that the more 

 particularly, as the notions which occur to me on the fubjed have led me to differ from him 

 in opinion. 



In the firfl place I mull remark, that we have no adual information of the exigence of 

 fuch vaft caverns, efpecially fo near the furface of the earth. 2. By the author's own ex- 

 periments it appears, that at very inconfiderable depths, compared with the thicknefs re- 

 quired for the roofs of fuch caverns, the influence of the fcafons is fcarcely perceptible. 

 He found the greateft difference at the depth of 29 1 feet, in compa£l ground, during three 

 years, was only from 8,95 degrees to 7,75 degrees, or 1,2 degrees in the whole. 3. If 

 the roof were ever fo thin, it appears from Count Rumford's experiments *, that heat, 

 which in air, and very probably in all fluids whatfoever, is fcarcely tranfmitted but by the 

 afcent of the heated parts or defcent of the parts cooled, would not pafs downwards, be- 

 caufe the rarefied parts could not defcend; confequentiy the expanfion, even in this mod 

 improbable fituation, could not take place. 4. From the expetiments of Duvernois re- 

 lated in the Encyclopedic Mcthodique, article Air, page 6S6, vi-hich are alfo to be found in 

 the firft volume of the Annales de Chitnie, it appears that common air dilates rather more 

 than ,\\\\ part of its bulk by the firft twenty degrees of Reaumur, and therefore fcarcely more 

 than -j-'-^th part by the difference of temperature fuppofed by M. De Sauffure. This cave 

 muft therefore contain near the furface of the earth, to be within the reach of the feafons, 

 eighty times as much air as flows out of all the apertures during the whole fummer. For 

 want of data, that is to fay, the apertures and velocities of emiffion, it is impoffible to infti- 

 tute a calculation ; but it feems utterly improbable that fuch a volume of air (hould have 

 been referved in an appropriate veffcl. At all events, it is not in the fmalleft degree likely, 

 that Mont Teftaceo contains, or communicates with, fuch a refervoir. 5. Laftly, it feems 

 a gratuitous fuppofition, that the refervoir fliould be always dry, and the paffage tKrough 

 which the air iflites always wet. 



After this undifguifcd examination of the confiderations of M. De SaulTurc, I fhall myfelf 

 prefent a few notions refpedling the caufc of this phenomenon ; with the perfe£t wifli to fee 

 them refuted, if cxifting fafts, future difcoveiics, or undetedcd errors (hould render it 

 neceflary. 



The whole, then, appears to me to depend upon the fimple cireumftance of the cavity be- 

 tween a confiderable mafs of flones or other fubft.inces not being capable of fpeedily changing 

 its temperature, by any other means than the conta£l of the air, or other fluids, which may 

 pafs through it. Let us fuppofc, for example, the cathedral of St. Paul's in London, the dome 

 of which is near 400 feet in height, to be filled with fragments of broken pottery. This large 

 mafs might be fuppofed at firit to poflcfs the temperature of the air at the time wlien it 

 was accumulated. From the imperfeft conducing power of pottery and moft earthy fub- 

 (laaces, the effctl of the fun's rays, or of the adual lieat of the furrounding air, would penc- 



' Exi>cu(iicnial Eir::y5, VII. 



H h J trite 



