ON A NEW METHOD OT" CALIBllATING, &C. 347 



error which it lias beeu fouiul most difficult to giiard against iii tlie use 

 of the furnaee is altocrether avoided. 



SUMMARY. 



Te metliod briefly outlined in this paper possesses the foUowing ad- 

 vantages over inethods hitherto practiced for the calibration of thermo- 

 elemeiitsand for the détermination of the melting points of certain inetals. 



(1) The use of the air thermonieter is avoided. 



(2) The cumbersoine and laborious ])rocedure of fusing the metals 

 in considérable niasses within a ftirnacc is reiidered unnecessary and the 

 difficulty of bringing the melted nietal and the juiiction of the therrao- 

 element to precisely the same température is eliminated. 



(3) The elaborate précautions necessary to prevent the vitiation of 

 the junction by the action of fumes, which constitutes one of the chief 

 difficulties of sncli work wlicii a furnaee is used is avoided. 



(4) The amonnt of nietal which is necesary to use in determining 

 melting points or in calibrating a thermo-element is much sinaller in 

 this than in the inethods u.sually employed. 



The accuracy of the inethods described is believed to be fuUy equal 

 to that of other known inethods of calibration. 



Physical Laboratory of Cornell University. 

 Oct. 1900. 



