276 



Scientific Life and Worh 



of a litre of air and on the density of mercury. The 

 values which he deduced are accepted on all hands as the 

 most reliable of any that we possess, and are quoted in 

 all the text-books as the standard determinations for these 

 important constants. He adds a formula by which the 

 weight of a litre of air may be found for any latitude 

 and any height above the sea-level.^ 



The fourth memoir treats of the measure of tempera- 

 tures. In this he shows the great superiority of air-ther- 

 mometers for scientific purposes. The experiments on 

 mercurial thermometers are very complete, and in all 

 cases include a chemical analysis of the glass envelope. 

 He thus compared three mercurial thermometers, having 

 differently constituted envelopes, with an air-thermome- 

 ter, in a series of sixty-three experiments, and afterwards 

 extended the comparison to other mercurial thermome- 

 ters. These investigations prove that beyond the boiling 

 point of water, the indications of the ordinary instru- 

 ments exhibit errors ranging from 1° to 10° C. The 

 following abstract of a part of his results makes this 

 statement clear: 



Air Thermometer. Thermometer with Thermometer with 



crystal envelope. ordinary envelope. 



100° 100°.00 100^.00 



150° 150°.40 149°.80 



200° 201°.25 199°.70 



250° 253°.00 250°.05 



300° 305°.72 301°. 08 



350° 360°.60 354°.00 



The same memoir also contains an extended discussion 



* Denoting the height above the sea by Ji, the latitude by I, and the mean 



1_ 



radius of the earth by R, the weight in grammes = 1.292673 X t , '^^ 

 [1 — . 002837 COS. 2ri. 



