266 Mr. Joule on the Calorific Effects 



stirring the water until the heat was equably diffused, its tem- 

 perature was ascertained by a very delicate thermometer, by 

 which I could estimate a change of temperature equal to about 

 j x ^th of Fahrenheit's degree. A cork covered with several folds 

 of greased paper was then forced into the mouth of the tube, 

 and kept in its place by a wire passing over the whole, and 

 tightened by means of one or two small wooden wedges. The 

 revolving piece was then restored to its place as quickly as 

 possible, and revolved between the poles of the large electro- 

 magnets for a quarter of an hour, during which time the de- 

 flections of the galvanometer and the temperature of the room 

 were carefully noted. Finally, another observation with the 

 thermometer detected any change that had taken place in the 

 temperature of the water. 



Notwithstanding the precautions taken against the injurious 

 effects of radiation and convection of heat, I was led into error 

 by my first trials : the water had lost heat, even when the tem- 

 perature of the room was such as led me to anticipate a con- 

 trary result. I did not stop to inquire into the cause of the 

 anomaly, but I provided effectually against its interference 

 with the subsequent results by interpolating the experiments 

 with others made under the same circumstances, except as 

 regards the communication of the battery with the stationary 

 electro-magnets, which was in these instances broken. And 

 to avoid any objection which might be made with regard to 

 the heat, however trifling, evolved by the wires of the large 

 electro-magnets, the thermometer employed in registering the 

 temperature of the air was situated so as to receive the influ- 

 ence arising from that source equally with the revolving 

 piece. 



I will now give a series of experiments in which six Daniell's 

 cells, each 25 inches high and 5| inches in diameter, were 

 alternately connected and disconnected with the large station- 

 ary electro-magnets. The galvanometer, connected through 

 the commutator with the revolving electro-magnet, had a coil 

 of a foot in diameter, consisting of five turns of copper wire, 

 and a needle six inches long. Its deflections could be turned 

 into quantities of current by means of a table constructed 

 from previous experiments. The galvanometer was situated 

 so as to be out of the reach of the attractions of the large 

 electro-magnets, and every other precaution was taken to 

 render the experiments worthy of reliance. The rotation 

 was in every instance carried on for exactly a quarter of an 

 hour. 



