4-i PROCEEDINGS OF THE AMERICAN ACADEMY 



2. It remains to point out that, as the heating bulbs g g and h h were 

 of sufficient volume to hold as much gas as would be transpired by the 

 capillaries in a quarter to half an hour, so that the passage of the 

 gas through them was extremely slow, as the bath around them, 

 whether of ice or oil, was so arranged or stirred as to insure satis- 

 factory contact with the bulbs, and as proper precautions were taken 

 as to the position of the thermometer, the temperatures given by the 

 latter must have been very close to the actual temperature of the gas. 



By the foregoing discussion it has, I think, been shown, that the 

 sources of constant error are unlikely to produce accumulated errors in 

 y exceeding 0.5 joer cent at the highest temperatures used, and less at 

 lower temperatures. I am disposed to believe that in the case of air 

 these errors are smaller than for carbonic acid. 



Rkferences. 



Graham Phil. Trans. Roy. Soc, London (1846-49). 



Maxwell .... Phil. Mag., xix., xx. (1800) ; x.xxv. 211. 



Pliil. Trans. Roy. Soc, London (1806). 

 Clausius .... Phil. Mag., xix. 434 (1860). 



Stefan Wien. Ber., xlvi. (1802). 



Stokes Trans. Cambridge Phil. Soc, viii. 287 (1847). 



O. E. Meter . . . Pogg. Ann., cxxv. 177, 401, 564 (1865) ; cxxvii. 199, 353 

 (1806) ; cxliii. 24, 226 (1871) ; cxlviii. 1, 203, 5:^6. 



PcLUJ Wien. Ber., Lxix. 287 (1874); Ixx. 268 (1874) ; Ixxiii. 



(1876) ; July, 1878. 



Wied. Ann. i. 296 (1877) ; (1878). 

 A. vox Oeermater Wien. Ber., Ixxi. 281 (1875) ; Ixxiii. 433 (1876). 



Carl's Rep., xiii. 130 (1877). 

 E. Wiede-mann . . Arch. Sci. Phys. Geneva (2), xlvi. 278. 



Fortschr. d. Physik , xxxii. 206 (1870). 

 HoLMAN Proc. Amer. Acad. Arts and Sci., xii. 41 (1876). 



Phil. Mag., iii. 81 (1870). 



Wied. Beibl., i. 222. 



Warburg Pogg. Ann., clix. 899 (1870). 



Guthrie .... Phil. Mag., v. 433 (1878). 

 P.Hoffman . . . Wied. Ann., xxi. 470 (1884). 



