528 PROCEEDINGS OF THE AMERICAN ACADEMY. 



tated by means of a perforated ring of heavy sheet copper, lifted by an 

 electric motor and allowed to fall by its own weight. The stirring in 

 the cover need not be so energetic as that below, since the temperature 

 in this part of the bath need not be so accurately adjusted as in the 

 lower compartment. 



As the temperature in the calorimeter rose, because of the condensa- 

 tion of the vaporized liquid, concentrated sulphuric acid was run from 

 burettes into both jacketing compartments, so that the environment 

 around the calorimeter was changed in temperature as fast as the 

 calorimeter itself. The details are so similar to those of the other 

 applications of this adiabatic method that they need not be reviewed. 

 The stirring was so effective that the thermometers in different parts 

 of the bath all rose at practically the same rate. 



It was found possible to follow the rising temperature of the calo- 

 rimeter within 0.05° in the lower compartment and within 0.1° in the 

 upper compartment throughout the entire course of the experimeut. 

 Suitable tests proved that with so small a difference in temperature 

 there was no danger of appreciable heat-exchange with the surround- 

 ings during the brief progress of a determination. 



The rise in temperature of the calorimeter was measured by a care- 

 fully standardized Beckmann thermometer, having a slender bulb and 

 rather a long scale. Readings to 0.001° were made with certainty by 

 means of a lens. The standardization was effected with very great 

 care by comparison with two Baudin thermometers standardized by the 

 Bureau des Poids et Mesures of Paris. The comparison was con- 

 ducted by immersing all the thermometers in the water of the adia- 

 batic calorimeter surrounded by its jacket. The calorimeter was 

 covered tightly with a non-conducting cover through which the ther- 

 mometers and stirrer passed, and the thermometer to be studied was 

 immersed to the depth at which it was subsequently to be used. All 

 the thermometers were placed as closely together as possible in order 

 to insure their having the same temperature. The temperature of the 

 calorimeter water could be held perfectly constant for any desired 

 length of time, so that there was no error due to the lag in any of the 

 thermometers. Headings were made (after gently tapping the thermom- 

 eters) at every -i\j° mark, and the corrections were calculated with due 

 account of the corrections for the various standardized thermometers 

 used. The Baudin thermometers had previously been found to agree 

 very closely with an accurate Fuess thermometer, which in its turn 

 had been carefully studied by the Reichsanstalt of Berlin. 



The various parts of the apparatus having been described in detail, 

 the actual execution of the experimental work may now be considered. 



