314 . MR. W. DUDDELL ON THE RESISTANCE AND 



considerable sudden change in the P.D. between the armatures of G sent sufficient 

 current through M to burn it up. Even this precaution did not prevent M from 

 being burnt up several times, owing to the battery connections being broken or the 

 arc going out while the key k was depressed. 



The switch S t was to enable M to be completely disconnected when taking its zero, 

 as it was thought that at these high frequencies, so long as one pole was connected, 

 there might be a small current flowing into the instrument, due to the capacity of the 

 instrument with surrounding objects, but this effect was not observed. 



Owhig to the very high frequencies used, very great care had to be taken in 

 arranging the circuit so as to avoid self-induction and capacity errors in the leads and 

 connections. All the leads through which the alternating current flows were carefully 

 twisted and bound together. 



To reduce any possible error caused by the lead tj) between the arc and R, the 

 drop along which had to be included either with the arc or with R, this lead was 

 made by twisting together 12 No. 23 double cotton-covered wires, to avoid possible 

 skin effects, and its length was reduced to GO centims. As first constructed, each 

 wire of the lead t.J> was twisted with the corresponding wire of lead t^, which was of 

 the same length and made in the same way, and then the 12 pairs of wires were 

 twisted together. It Avas found on testing these leads that at a frequency of 

 18,000 -- per second, and with a P.D. of 33 volts between the two leads, an 

 alternating current of about 1'9 X 10~ :! ampere floAved between them due to their 

 capacity. To reduce this capacity current, tAvo IICAV leads were constructed, exactly 

 the same as before, only that instead of twisting the individual wires belonging to 

 each lead together in pairs, all the Avires belonging to each lead were stranded 

 together so as to form tAvo separate leads. The lead tj) Avas then bound over with a 

 layer of silk tape and the t\vo leads were twisted together. On re-testing in 

 the same way as before, the capacity current was found to be reduced to about 

 0'53 X 10~ 3 ampere with 33 volts between the leads. As the alternating P.D. 

 between the leads Avas under 0'5 A^olt in most of the experiments, this capacity 

 current Avas negligible compared Avith the working current of O'l ampere. 



As the self-induction of II had to be determined and allowed for, the' lead tjb was. 

 included with II, so that its small self-induction could be corrected for at the same 

 time. 



The lead c, d Avas brought back along the connections b, c between the 12 coils 

 of R, so as to neutralise as well as possible the magnetic field of these connections. 



In arranging the circuit the capacity of those parts of the main circuit between the 

 measuring points t { , t z , t 3 , as Avell as of the whole of the measuring and alternator 

 circuits, to surrounding objects and to earth, was kept as small as possible, so as to 

 avoid what might be called capacity leaks. The alternator itself was practically 

 insulated from earth by being fixed down to a wooden frame, and the field circuit of 

 the alternator was well insulated and removed from earth. 



