EXPERIMENTS ON SPECIFIC HEAT. 819 



venting transmission of heat in all directions during the 

 experiment. By surrounding a body with bad con- 

 ductors it is possible to retard the mutual transmission 

 of heat between the body under experiment and sur- 

 rounding bodies; but it is absolutely impossible to pre- 

 vent gain or loss of some heat during the experiment, 

 and the difficulty of ascertaining and allowing for the 

 quantity thus lost or gained, makes the accurate 

 measurement of specific heats, and many similar opera- 

 tions, very complicated and .troublesome. The follow- 

 ing few determinations of specific heats are there- 

 fore only intended as general illustrations, not as 

 examples of experiments by which these specific heats 

 could actually be determined in an exact manner, since 

 the necessary corrections are completely neglected. 



500 grammes of water is poured into a large flask 

 of thin glass, and a thermometer is suspended in the 

 water. In another flask the same weight of water, viz. 

 500 grammes, is heated, and its temperature is also 

 ascertained by a thermometer. When the water in 

 the second flask nearly boils, it is quickly poured 

 into the flask containing the cold water, both ther- 

 mometers having been read off just previously. The 

 mixture is briskly stirred with a splinter of wood, 

 and the temperature observed. The temperature of the 

 mixture will be the arithmetical mean of the two 

 temperatures previously observed, viz. that of the cold 

 and of the hot water. If, for example, the temperature 

 of the cold water was 15, that of the hot water 95, 

 the temperature of the mixture will be found to be 



= 55, or somewhat less, because a portion of 

 3 G 2 



