somewhat analogous to Newton s Rings. 85 



Thus, by Equation (1) we obtain b = ^. From this value of b and 

 those last used for a. and /B we derive the curve shown by T, fig. 5. 



This, as in the case of Experiment V, lies wholly below the experi- 

 mental curve. But here, again, the correction to the experimental 

 <jurve applied in accordance with Equation (4), and the higher value 

 of <T from Experiment IV yields the curve C, fig. 5, lying still lower 

 than the theoretical one. 



In Experiment VII. It may be seen from the Equations (3) or from 

 general considerations that the total intensity of the reflected waves 

 must be complementary to that of the transmitted ones, that is p + T 

 = 1. Consequently the theoretical curve showing p as ordinates 

 instead of T would be obtained by inverting the theoretical one for 7. 

 Hence, we see that the general characteristics of the curve for p may 

 be thus outlined : 



(1) A damped wavy-formed curve with 



(2) narrow troughs at Z = 0, \/2, X, &c., and 



(3) broad crests at I = V/4, 3A/4, &c., 



where Z is the length of the abnormal part. These general charac- 

 istics are possessed by the experimental curve obtained. It must, of 

 course, be borne in mind that the ordinates of the curve thus experi- 

 mentally determined do not represent p, but represent approximately 

 ratios proportional to 1 4-/>, since we have both the reflected and the 

 incident waves passing the electrometer. On this account the wavy 

 form of this curve is less strongly marked than in the case of trans- 

 mission. Probably also other, and undesired, interferences arose 

 between the on-coming and reflected waves in the neighbourhood of 

 the electrometer, thus causing the two anomalous humps at 1*5 m. 

 and 5'5 m. 



In Experiment VIII. For the cases to which Equation (1) applies 

 we see that there is no reflexion when r = 1, that is, when no change 

 in the capacity of the secondary occurs. Now the approximate 

 expression for the capacity of two equal parallel cylinders depends 

 not upon their absolute but upon their relative dimensions only. 

 Hence, if Equation (1) is correct, it must be possible to arrange a 

 part of the wires in a form which appears very abnormal, but yet so 

 ;as to produce no reflexion. We have simply to introduce, at any 

 part, thinner wires placed proportionately nearer together or thicker 

 ones in like manner further apart. Thus, since in either case, the 

 capacity is unaltered by the change in question, we have r = 1, and 

 the fraction of wave-energy reflected disappears. 



In the experiment under consideration the readings taken pointed 

 to the distance 1*37 cm., from centre to centre of wires, as being that 

 which, with the w^ires in use, would give no reflexion. Theory gives 

 -as the correct distance T32 cm. Thus the discrepancy is not great ; 



