1 20 CA USES which affett the ACCURACY 



> 



any depreffion below the furface, as, z plog.^- r + (log.\ 



b is * b I 



the temperature being fuppofed r, and the term 



P 2 I 3 

 l--r being left out, as relating only to the diminution of 



S O 



the weight of the quickfilver in the lower barometer. If, then, 

 b, or the column of mercury in the barometer at the furface, 



be 30 inches, or of a fathom, and (3 =. 4343, we find/>%--r- = 



joooo x tabular log. 10423 401 80 fathoms 45. 6 miles near- 



P 2 I & \ 2 



ly. The fecond term, \ &'~r) ( or tne fq ua re of the for- 

 mer divided by the diameter of the earth), = + .25 of a mile, 

 fo that z 45-85 miles nearly. The approximation might be 

 carried to much greater exac~lnefs if it were necefTary ; but 

 this is fufficient to {hew, that, at a lefs depth under the furface 

 than 46 miles, the denfity of air would become equal to that of 

 quickfilver ; and if this conclufion appear, in any degree, pa- 

 radoxical, it need only be confidered, that, abftracYmg from 

 any diminution of the power of gravitation, the denfity of air 

 would be nearly doubled by every 37 miles of defcent below 

 the furface of the earth. 



35. IF, again, we would form any conelufion concerning the 

 limit to which our atmofphere may extend upwards, we muft 



J (J+JC) 



refume the formula, y ~ - j+^ 



9 (' 



and, if we would abftracl from the effect of the cold in the 

 higher regions to reduce the atmofphere within narrower 

 limits than thofe to which it would otherwife extend, 



we 



