RICHARDS AND LAMB. — SPECIFIC HEAT OF LIQUIDS. 075 



the thermostat-pipettes at a temperature just low enough to counterbalance 

 the heat of the reaction, so that the mixture also vvould be at the temper- 

 ature of the room. This application of the device will be tried in the 

 near future. 



The Heat of Dilution. 



For the determination of the heat of dilution the same apparatus could 

 have been used, adjusting all the thermostats to the same temperature. 

 It was, however, much more convenient to use for this purpose a simpler 

 and prol)ably more accurate device, described below. 



This device consisted merely of a beaker divided into equal right and 

 left-hand compartments, by means of a detachable partition of silver foil. 

 The partition was held iu place by a narrow U-shaped ledge of glass 

 cemented to the walls and floor of the beaker. To this ledge the silver 

 foil was fastened, water-tight, by means of hard parafline. A glass rod 

 was securely attached by wires and cement to the partition, and by its 

 means the partition could be easily detached from its supports. Wa'er 

 was placed in one compartment and the solution in the other. Each was 

 provided with a stirrer and a sensitive Beekmann thermometer. The 

 whole was then placed in a calorimeter at the desired temperature. As 

 a rule, it was more expeditious to adjust the temperature of the liquids on 

 either side of the partition to within a few one-hundredths of a degree of 

 each other, before they were placed in the calorimeter. When the liquids 

 in each compartment had reached practically the same temperature and 

 were either stationary or very slowly rising in terajjerature, readings 

 were made of the thermometers at the minute intervals. At a given 

 moment, having made a final reading of the thermometers, the partition 

 was removed, active stirring continued, and the thermometers read at in- 

 tervals of a minute, until the same regular rate of warming or cooling had 

 established itself. This seldom required more than two or three minutes. 

 It was not difficult to make the necessary corrections for radiation over 

 so short an interval. Tlie average change recorded in both thermometers 

 evidently represents the actual temperature change — since both liquids 

 were not usually at exactly the same temperature. 



Since the heat of dilution of these concentrations only amounts to a few 

 one-thousandths of a degree, there is evidently no need of great accuracy 

 in determining the quantity of liquids used. We therefore measured out 

 equal volumes, using a pipette of 0.173 litre capacity. This reproduced 

 very nearly the actual weights of water and solution used in the above 

 determinations of specific heats. The heat cajjacity of the beaker 



