92 CAUSES which affetl the ACCURACY 



cepting in one circumftance ; for the formula makes the expan- 

 fion increafe with the heat continually, though not uniform- 

 ly ; wherea? the experiments give the greateft expanfion be- 

 tween the temperatures of 60 and 70. But this feems to 

 be fo anomalous a fact, that it looks more like .fome accidental 

 effect, produced from the particular manner of making the 

 experiments, than a part of that law of nature, which connects 

 the variations of bulk in bodies with their variations of tem- 

 perature. 



7. EOT this is not the only irregularity to which the ex- 

 panfion of air by heat, and its contraction by cold, appear to 

 be fubject. We learn from the manometrical experiments of 

 the fame excellent obferver, that a given variation of tempera- 

 ture is accompanied with more or lefs variation of bulk, accord- 

 ing as the* air is comprefled by a greater or a lefs force. Air, 

 for inftance, comprefled by the weight of an entire atmofphere, 

 was expanded by the 180 degrees from freezing to boiling, no 

 lefs than 484 of thofe parts, whereof, at the temperature 32, 

 it occupied 1000. But the fame air, when comprefled only by 

 i of an atmofphere, was, by the fame difference of heat, ex- 

 panded no more than 141 parts; and that though the heat of 

 boiling water was applied to it for an hour together. It is not 

 eafy either to aflign the caufe, or to determine the law of this 

 inequality. GeneraLRov has, indeed, conftructed a table of the 

 correction to be made on account of it ; which proceeds on 

 the fuppolition, that the expanfion, for one degree of heat, de- 

 creafes in the fame proportion that the column of mercury in 

 the barometer exceeds a given length. This given length is 

 nearly 4.5 inches ; fo that if b be the length of the column of 

 mercury in the barometer, and .00252 the expanfion for one de- 

 gree of heat, when the barometer is at 30 inches, and the tem- 

 perature of the air 3 2 ? , then - : x -00252, will be the expanfion 



of air of the fame temperature, for the fame change of heat, 



when 



