Comparison of Mercury and Platinum Thermometers. 165 

 Again 



.*. sds= — fid [a, 



/j? = l — s 2 = l ultimately, 



s= - ; .*. ds= — ; 



a a 



,.d„=-s^=-^; 



a a' 



p,_^Pn a 2 dJ (kp) 



71 dfi p dp 



= — Ji(*p). 



r 



XXIV. -A Comparison of Rowland's Mercury Thermometers 

 with a Griffiths Platinum Thermometer. By C. W. 

 Waidnee and F. Malloey*. 



THE determinations of the mechanical equivalent of heat 

 by electrical and mechanical methods show a close 

 agreement between the values obtained by different experi- 

 menters using the same method of experiment. There 

 appears, however, a very appreciable difference between the 

 values as determined by the two methods, a difference greater 

 than can be accounted for by errors of experiment. This 

 difference can only be explained on the assumption that the 

 results of the different experimenters are based on different 

 thermometric standards or of a still undiscovered error in the 

 system of electric units employed. The recent comparison 

 by Professor Schuster of Joule's thermometers has indirectly 

 furnished a connexion between Professor Rowland's air 

 thermometer and the nitrogen standard of the Bureau 

 International. 



This comparison pointed to differences in these two 

 standards as great as '05° C, but as the details of Joule's 

 comparison of Rowland's thermometer (No. 6166) with his 

 own are not known, this correction is uncertain. It was there- 

 fore thought advisable to make another comparison of Professor 

 Rowland's mercurial thermometers under conditions as nearly 



* From the Johns Hopkins University Circulars, June 1897. 



