Determining the Specific Heat of a Liquid. 107 



every way. All that is required is to pump out the contents 

 of the spiral, wash it out with distilled water, pump tliis out 

 and rinse the spiral with a little spirits and then apply an 

 air-pump for half an hour to thoroughly dry the interior of 

 the spiral. During all this time the contents of the calori- 

 meter are in no way disturbed, the covers and shields are not 

 removed and no evaporation can go on. While the spiral is 

 being dried, the tube in the jacket is re-tilled and all will be 

 in a steady condition again by the time the spiral is dry. 

 Then the motor is re-connected, and a few seconds before the 

 experiment is to take place the calorimeter is pushed up to 

 the projecting part of the jacket and the loose end of the 

 rubber tube from the hot tube is slipped over the projecting 

 end of the glass spiral. Then the thermometers are read and 

 the pinch-cock released and in a minute to a minute and a 

 half the thermometer in the calorimeter has reached its 

 maximum. 



Some experience of working with the calorimeter as above 

 described showed the necessity of a slight alteration in detail. 

 It was found that in the course of the various manipulations 

 required, but especially in shifting the calorimeter to and from 

 the balance, there was a risk of some of the contents being 

 splashed up against the cover, and a little wet also would 

 occasionally get on the outside cover; as these were made of cork 

 they absorbed the moisture, and the quantity of liquid inside 

 the calorimeter might thus be altered during the course of an 

 experinient without auy indication being shown by the balance. 

 To avoid this, the cork discs were replaced by a piece of solid 

 wood turned to fit and thoroughly varnished so as to be 

 impervious to moisture. 



The investigations concerning two further corrections now . 

 remain to be described. In the first place the question arose 

 whether the heat generated inside the calorimeter by the action 

 of the stirrer would be suflScient to be perceptible or need the 

 introduction of a correction. To test this matter the calorimeter 

 was set up and the stirrer kept in motion for about one week 

 without interruption and the temperatures of the air in the room 

 and the water in the calorimeter observed at intervals. At 

 first there were considerable irregularites, but after a time the 



