458 Analyses of Books, [Dec. 



5*203 inches of mercury (see Table I, p. 13, of Mr. Daniell's 

 work). The barometer, therefore, which, at the surface of the 

 supposed sphere, stands at 30 inches, will, at this elevation, 

 indicate 24'797. But if the temperature of this aerial column 



fradually decrease from 32°, till, at the height of 5000 feet, it 

 ecomes 14*8°, it is required to determine the change which 

 this variation will produce in the height of the mercurial column 

 at the above elevation. The question seems to us to reduce 

 itself to a simple comparison between the weight of a column of 

 air 5000 feet high, of the temperature 32° Fahr. and that of an 

 equal column of the temperature 23*4°, which is the mean of the 

 temperature at the base and that at the summit. Now air by 



being reduced 1° Fahr. contracts in bulk— of the volume which 



it would occupy at 32° ; consequently a reduction of tempera- 

 ture equal to 8*6 (32'' — 23''*4) will be accompanied by a 



decrease of volume equivalent to — of its former bulk. The 



vacuous space which would be left by such a contraction must 

 be immediately filled up by air from above. Hence the mercu- 

 rial column at 5000 feet must, by falling, indicate this transfer- 

 ence of air from the superior to the lower strata, and this fall 



will be equal to ^^^ of 5-203 = -093. At the elevation of 5000 

 feet t hen, the height of the barometrical column will be equal to 

 30 - 5-2U3 + -093 = 24-704, instead of 23-949, the number 

 given by Mr. Daniell. The same result will be obtained by 

 means of a formula derived algebraically from one originally 

 given by Sir G. Shuckburgh.''*^ Let H denote the height of the 

 mercurial column at the surface of the earth, y that at a given 

 elevation p (in the present instance 5000 feet), and b the number 

 of feet of air of the given temperature (23'''4), equal to 1-lOth 

 inch mercury. 



Then 3/ = ^tt- x H. Substituting in this formula the 



values of h and j?, the former of which is obtained from a table 



I. o- i-< c?i. 1 1, u t, 600 X 85-044 - 5000 .. _ 



given by SirG.Shuckburgh, we have y = ,,, ^ ,,.^,^ ^ ^^ x 30 



= 24*64. The small difference between this result and the 

 former one may be attributed to Sir G. Shuckburgh's having 



estimated the expansion of air for each degree Fahr. at t^-. 



instead of j^ of its original bulk. 



Our limits will not permit us to enter at any length into the 

 account of Mr. DanielPs hygrometer, which is fully described 

 in his work, and also in the Quarterly Journal, Nos. 1 1 and 25 

 We consider it as an elegant instrument, and are satisfied by 



• Dalton's Meteorological Essays, p. 82. 



