DICKINSON AND OSBORNE: SPECIFIC HEAT OF ICE 339 



sensible increases in the apparent heat capacity with ice samples 

 of only ordinary purity are accounted for by the assumption of 

 incipient fusion caused by the lowering of the melting point by 

 the dissolved impurities. 



The present investigation has been undertaken with the object 

 of securing further evidence as to the thermal behavior of ice 

 at temperatures near the freezing point, and of obtaining reliable 

 data for the construction of tables of total heat of ice and water 

 in the range of temperature with which refrigerating engineers 

 are concerned. 



The measurements were made by means of a calorimeter of 

 aneroid type (i.e., without stii-red liquid as calorimetric medium) 

 which is described in detail in the foregoing paper. Briefly 

 described, the tin lined metal cell containing the specimen is 

 enclosed within a shell of copper, the copper acting as the calori- 

 metric medium for transmission and distribution of heat elec- 

 trically developed in a coil built into the shell. The caloruneter 

 is surrounded by air and enclosed in a metal jacket which is 

 surrounded by a stirred liquid bath, the temperature of which 

 can be controlled within a few thousandths of a degree at any 

 temperature between — 55° and +40°C. Measured amounts of 

 heat are supplied to the calorimeter electrically by means of the 

 built-in heating coil. The calorimeter temperature is measured 

 by means of a platinum resistance thermometer also built into 

 the shell. DifTerences in temperature between the calorimeter 

 and jacket surfaces are measured by means of multiple thermo- 

 couples distributed on the surfaces. During an experiment 

 thermal leakage is minimized by keeping the temperature of 

 the jacket nearly equal to the changing temperature of the 

 calorimeter. 



The samples used were from 400 to 470 grams each. Three 

 samples were of redistilled water of fairly high purity, while a 

 fourth, which was distilled directly into the container, appeared 

 from the experunental results to have a much higher degree of 

 purity. 



In the determinations of specific heat it is found that over 

 the range of temperature covered by the experiments, i.e., —40° 



