274 Mr. Joule on the Calorific Effects 



On comparing the corrected results in column 5 with the 

 squares of magneto-electricity given in column 6, it will be 

 abundantly manifest that the heat evolved by the coil of the 

 magneto-electrical machine is 'proportional [ceteris paribus) 

 to the square of the current. 



Column 7, containing the heat due to voltaic currents of 

 the quantities stated in column 2, is constructed on the basis 

 of three very careful experiments on the heat evolved by 

 passing currents through the coil of the small compound 

 electro-magnet. I observed an increase in the temperature 

 of the water equal to 5 0, 3, 5°*46, and 5°*9 respectively, when 

 2*028, 2*078, and 2' 145 of current voltaic electricity were 

 passed, each during a quarter of an hour, through the coil. 

 Reducing the first and second experiments to the electricity 

 of the third according to the squares of the current, we have 

 5°-93, 5°*82 and 5°-9 for 2-145 of current. The mean of these 

 is 5 0, 88, a datum from which the theoretical results of the 

 preceding and subsequent tables are calculated. 



But in comparing the heat evolved by magneto-, with that 

 evolved by voltaic electricity, we must remember that the 

 former is propagated by pulsations, the latter uniformly, 

 Now since the square of the mean of unequal numbers is 

 always less than the mean of their squares, it is obvious that 

 the magnetic effect at the galvanometer will bear a greater 

 proportion to the heat evolved by the voltaic, than the mag- 

 neto-electricity; so that it is impossible to institute a strict 

 comparison without ascertaining previously the intensity of 

 the magneto-electricity at every instant of the revolution of 

 the revolving electro-magnet. I have not been able to devise 

 any very accurate means for attaining this object: but judging 

 from the comparative brilliancy of the sparks when the com- 

 mutator was arranged so as to break contact with the mer- 

 cury at different positions of the revolving electro-magnet 

 with respect to the poles of the stationary electro-magnet, 

 there appeared to be but little variation in the intensity of the 

 magneto-electricity during f of each revolution. The re- 

 maining ^ (during which the revolving electro-magnet passes 

 the poles of the stationary electro-magnet) is occupied in the 

 conversion of the direction of the electricity. In the experi- 

 ments all flow of electricity during this 4 is cut off by the di- 

 visions of the commutator. In illustration of this I have 

 drawn fig. 5, in which the direction and intensity of the mag- 

 neto-electricity are represented by ordinates A#, &c, perpen- 

 dicular to the straight line ABCDE; the intermediate 

 spaces B C, CD, &c, represent the time during which the 

 electricity is wholly cut off by the divisions of the commutator. 



