Measurement of Electromagnetic Radiation. 63 



The length of the wire (/) is 103 centimetres, and its 

 resistance 19 ohms = 19x 10 9 C.G-.S. units. 



In the case of platinum fju may be taken as unity. 

 From this it follows that 



R' = 29 ohms, 



and since H (half the heat developed in the two wires) 

 = •000342 calorie per second 



r2 _JH 



= •000049. 



Owing to the harmonic distribution of the current along the 

 wire, in order to produce a given amount of heat in the whole 

 wire, the strength of the current at the centre must be \^2 

 times as great as that which would be necessary if the current 

 were of uniform strength in all parts of the wire. 



.*. the square root of the mean square of the current at 

 the centre = '01 ampere. 



While it is impossible from the heat found to form any 

 conclusion as to how fast heat was being generated during 

 each spark, or what fraction of the whole time the oscillations 

 lasted, yet here, as in the former case, the final result depends 

 on the mean square of the current. We found by the first 

 method that the square root of the mean square was certainly 

 not *003 ampere, if the theory was correct ; while by the heat we 

 find the square root of the mean square to be *01 ampere. The 

 arrangements in the two cases were not strictly comparable, 

 for in the first the oscillator was only 30 centimetres long, while 

 in the secondit was 109 centimetres, and because the distance of 

 the primary from the secondaries was relatively greater in the 

 first than in the second case. We intended to have made a 

 second set of measures in which the suspended cranked oscil- 

 lator should itself be in the tube of the convection-thermometer 

 together with the fixed cranked oscillator. In this way the 

 force, if any, between them could be observed and measured 

 at the same time that the deflexion of the mirror due to the 

 heat was read ; then, by making contacts with the ends of the 

 fixed cranked oscillator alone, and passing a current throuoh 

 it so as to produce the same deflexion, all the data would be 

 obtained to tell for certain whether the absence of force, or, if 

 force was observed, if the actual force was that which the 

 provisional theory set out at the beginning of this paper in- 

 dicated. The discrepancy of 3 : 10 in the present instance 



