Velocity of Electric Waves in Air. 127 



The means of each of these sets of five readings are taken 

 and plotted as a curve, the co-ordinates of whose points are 

 the distances of the coherer from the reflector and these mean 

 values. The resulting curve is given in fig. 1 . 



Fig-, l. 



H<jllct°, 



The velocity of the electric waves was determined from the 

 formula X = VT=2wV \/L(J, where \ is the wave-length. L 

 the self-induction, C the capacity, V the velocity, and T the 

 time of oscillation of the secondary spark. T was found by 

 the photographic process. 



The sparks directly given by the oscillator used in the 

 original experiment could not be photographed. They were 

 too small, and, moreover, their light was not actinic enough. 

 Sparks given by similar capacity and self-induction, but of 

 larger dimensions, and cadmium points in place of the platinum 

 globules were photographed. The self-induction here used 

 consisted of two parallel copper wires of the same diameter as 

 were employed in the original experiment, 5 cm. apart and 

 1051*1 cm. in length. The capacity was the same glass plate 

 with the tinfoil sheets four times as large. The time of 

 oscillation of the sparks thus produced was found to be 

 4-12382 x 10~ 7 = T' seconds. It was then assumed that the 

 formula must still hold good if we cut down the self-induction 

 and capacity, still keeping the copper wires 5 cm. apart. 

 That is to say, if we make our C one fourth and our L 1/100 

 of the L used in the photographic process, the time of 

 oscillation of our spark will be 1/20 of that of the spark 

 photographed. 



