SPECIFIC HEAT OF AMMONIA. 347 



temperatures slightly above room temperature, two in the neighbor- 

 hood of 60° C, and two in the neighborhood of 115° C. It requires 

 comparison at three different temperatures to determine the constants 

 of the Callendar-Griffiths equation. Of the three room temperature 

 comparisons, two were at practically the same temperature, so these 

 were averaged to give a single point, making two sets of comparisons 

 of three each. From each of these sets the constants for the observ- 

 ing thermometer were computed, and then these two sets of constants 

 were averaged to obtain the final accepted values. 



8. Mass of Ammonia. 



The mass of the empty container was determined by the average 

 of three weighings, and the mass when loaded with ammonia, was 

 determined by the average of six weighings. "Weighings were made 

 on a 2 Kilo Bunge Balance, using the Laboratory Standard weights. 



9. Volume of Ammonia Container. 



The volume of the ammonia container was determined by weighing 

 the quantity of distilled water required to just fill it at room tempera- 

 ture. The specific volume of water at the observed temperature was 

 taken from Landolt and Bornstein's TaheUen. The average of three 

 determinations was taken as the value of the volume. Weighings were 

 made on the 2 Kilo Bunge Balance using the Laboratory Standard 

 ^Yeights. The volume was again determined at 78° C. by the same 

 method and no change was detected. 



V. Experimental Observations and Data. 



In this section are given all of the original observations and such 

 other data as were used in the computation of the results, or were 

 necessary to determine the precision of these measurements. 



1. Measurements of the Heat Capacity of the Apparatus: 



(a) "When filled with Ammonia. (Ammonia Measurements.) 



(b) AVhen Empty. (Calibration Measurements.) 

 Explanation of the tables. — The first two columns give the bridge 



readings of the resistance of the observing thermometer as a function 

 of the time. These readings were made at the end of each minute, 

 The notations "S^^itch on" and "S^vitch off," indicate that at those 

 instants the heater switch was thrown on and off, respectively. This 



