CONTRIBUTIONS FROM THE CHEMICAL LABORATORY 

 OF HARVARD COLLEGE. 



THE ELIMINATION OF THERMOMETRY LAG AND 



ACCIDENTAL LOSS OF HEAT IN 



CALORIMETRY. 



By Theodore W. Richards, Lawrence J. Henderson, and 

 George S. Forbes. 



Presented March 8, 1905. Received March 8, 1905. 



1. TWO NEW METHODS OF AVOIDING ERROR IN CALORIMETRY. 



— By Theodore William Richards. 



2. THE CONSTANCY OF RESULTS OBTAINED BY ONE OF THE NEW 



METHODS. — By Theodore W. Richards and George S. Forbes. 



3. THE COMPARISON OF THE TWO METHODS, AND THE EXACT 



ESTIMATION OF ADIABATIC RISE OF TEMPERATURE. — By 

 Theodore W. Richards and Lawrence J. Henderson. 



1. TWO NEW METHODS OF AVOIDING ERROR IN CALORIMETRY. 

 By Theodore W. Richards. 



It is very well known that the error most difficult to overcome in 

 calorimetric work is the accidental exchange of heat with the environ- 

 ment by radiation, convection, and conduction. Very few calorimetric 

 results are wholly free from error due to this cause, even comparative 

 methods, like that of Pfaundler,* being dependent upon the assumption 

 that the loss of heat in two similar sets of apparatus are equal, one to the 

 other. Thermo-chemical results, especially those involving deliberate 

 reactions, are notoriously subject to uncertainty on account of the mag- 

 nitude of their corrections for cooling. The effort to annihilate such 

 error as this was the guiding thought in a preceding investigation of 



* Sitzungsber. d. Akad. Wiss. Wien, 62-2, 379 (1870). 



