Mr. (X J. Lodge. Experiment* on 



[June 



distance , and diameter O8 ; or whether I reckon it a* a circle of 

 perimeter 2 (50 -I- u) and average thickness 0*3. For cases where the 

 rectangle is elongated, the former approximation is best, BO I use it 

 in preference to the other always ; reckoning the self -induction, there- 

 fore, as 



L/ji = 26-r200 log -^ cm- (24), 



lie real value will be somewhat greater than this. 



36. In the following tables the experimental data were obtained 

 and recorded carefully. The notes appended to each, concerning the 

 amount of agreement between calculation and experiment, are 

 capable of further refinement : but they suffice to show that the dis- 

 crepancies between calculation and observation are as small as could 

 be expected, and that, to a first approximation, experiment and 

 theory agree ; in other words, that if the velocity of a pulse, along 

 thin isolated copper wires, differs from the velocity of light, it does 

 not differ to any considerable extent. 



RESULTS (12th May, 1888). 

 37. Two Toss Jars end to end, as shown, with " Shorter Leads." 



Here the maximum occurs somewhere about the 35 cm. distance, 

 probably on the hither side of it ; so we may judge that the value 

 u = 34 is about the place where the waves emitted agree with twice 

 the length of the " shorter leads," i.e., are 31 J metres long. 



To see how this agrees with calculation, we must decide how much 

 of the capacity of those wires ought to be added to the capacity of 

 the jars in calculating the period of the discharge oscillation. If we 



