March i6, 1899] 



XATURE 



477 



In No. 3, potassium and sodium were fused together to form 

 the alloy liquid at ordinary temperatures. The ditference of 

 potential was 075 volt at starting, and fell in the course of seven 

 and a half years to o'49 volt. On opening the tube there was 

 little appreciable change in this value. The fall in this case also 

 was therefore no doubt due to tarnishing of the zinc surface. 



Experiment i is the only one of the three which lends some 

 degree of support to the hypothesis which, however, from 

 evidence in other directions, seems nevertheless to be the true 

 theory. If so, the negative results here obtained are no doubt 

 due to the difficulty of removing the last traces of active matter 

 from the gas employed. 



Experiments by C. Christian.sen ( IVied. Ann. , vol. Ivi. 

 p. 644), confirm this view. He shows that, if the metals be 

 exposed for only a minute fraction of a second in hydrogen the 

 difference of potential is very much lower than when the 

 exposure is continuous in air. Here the active matters have not 

 time to diffuse through the hydrogen to the metal in sufficient 

 quantity to produce the full effect. 



Physical Society, March to.— Prof. Oliver Lodge, F.R..S., 

 President, in the chair. — Mr. A. A. Campbell Swinton described 

 and exhibited the Wehnelt current-interrupter. A glass cell 

 contains a large cylindrical negative electrode of lead, and a 

 small positive electrode consisting of a platinum wire about 

 1/16 inch or J inch in length, in a solution of one part sulphuric 

 acid to about five parts water. The platinum wire may pro- 

 'ect from the top of the shorter arm of a J-shaped ebonite tube, 

 so that it can point upwards immersed in the solution. Or it 

 may be fused into a similar glass tube ; but glass is apt to 

 crack in the subsequent heating. Wehnelt's interrupter replaces 

 the make-and-break apparatus of an induction coil ; it also 

 replaces the ordinary condenser of that apparatus. In its pre- 

 sent form it requires rather a strong current. The resulting 

 spark at the secondary terminals differs in character from the 

 ordinary spark of an induction coil : it is almost unidirectional, 

 and in air takes a V-form, bright, continuous, and inverted- 

 somewhat like a pair of flaming swords rapidly crossing and 

 recrossing one another at their points. By blowing upon the 

 V it breaks up, and then more nearly resembles the customary 

 discharge of a coil. The sound emitted by the spark has a 

 pitch that varies with the conditions of the circuit. As the 

 self-induction of the circuit is diminished, the spark-pitch rises ; 

 it becomes infinite when the self-induction vanishes, i.e. the 

 Wehnelt interrupter will not work in a circuit devoid of self- 

 induction. As the applied potential-difference diminishes, the 

 spark-pitch diminishes. In Mr. Campbell Swinton's experi- 

 ments, twenty-five volts was the minimum primary voltage at 

 which his apparatus would work. The spark-pitch also varies 

 with the length of the platinum wire electrode in the solution. 

 If the circuit is closed by dipping this electrode into the solu- 

 tion, the apparatus will not work ; the wire must be dipped in 

 before closing the circuit. After working for about a quarter 

 of an hour the action often ceases ; this fatigue-effect is not 

 due to heating of the solution, for it is not obviated by keeping 

 the temperature constant by a water-bath. It is supposed 

 that the o.xygen generated at the platinum electrode forms 

 a more or less insulating film which interrupts the current until 

 absorbed by the surrounding water. The fact that oxygen is 

 more easily absorbed than hydrogen may explain why it is 

 necessary to connect the platinum electrode to the positive pole 

 of the battery or dynamo. When the platinum electrode is 

 dipped gradually into the solution, the wire gets red-hot, and 

 the interruptions do not take place. Again, when the apparatus 

 stops, from fatigue, the platinum gets red-hot. The action is 

 further complicated by a series of small explosions, and by the 

 formation of a kind oi electric arc at the platinum electrode. 

 The coil exhibited was connected to the loo-volt electric-light 

 mains at Burlington House ; in this case the potential difference 

 at the terminals of the primary was 30 volts, and that across the 

 interrupter 150 volts — a total of 180 volts, showing the effect of 

 impedance. For Rontgen-ray work the apparatus would be 

 very effective, but unfortunately the sparks produce great heat- 

 ing, so that the kathodes of tubes are melted. Mr. Campbell 

 Swinton suggested that as the sparks were more nearly con- 

 tinuous than ordinary discharges, they might produce Hertz 

 waves less rapidly attenuated than those now applied to wireless 

 telegraphy ; the trains of waves would also follow one another 

 at shorter intervals than those from the sparks at present 

 employed. The President said he was rather surprised that the 



NO. 1533, VOL. 59] 



self-induction of the primary coil was not sufficient of itself to 

 form the induction factor in the impedance necessary for perfect 

 working. He would like to know ho'w the apparatus behaved 

 when an alternating current was used. Did the secondary coil 

 become damaged by over-heating ? Did reversal of the current 

 assist the recovery from the fatigued condition of the 

 apparatus? The natural period of the circuit depended upon 

 its capacity and its self-induction. There was undoubtedly 

 capacity at the surface of the platinum electrode in the 

 liquid ; this capacity acted together with the auxiliary self- 

 induction, and the self-induction of the rest of the circuit, 

 in the orthodox way, and there was automatic adjustment 

 of resonance to the frequency of the interruptions, probably by 

 variations of the capacity at the electrode. The heating effect, 

 when a wire was made to close a circuit with a liquid, was dis- 

 covered many years ago. Prof G. M. Minchin thought that 

 the usefulness of the apparatus would be greatly increased if it 

 could be made to work with less current. He had himself 

 succeeded with 12 applied volts, but not with 10 volts. As a 

 tentative experiment he had used a horizontal lead plate, with 

 disastrous effect, for the apparatus went suddenly to pieces. Ex- 

 plosions were frequently obtained, but they were not attended 

 with much real danger. In a later and safer apparatus he used 

 a platinum wire about \ inch long, projecting from a glass tube 

 around which the lead plate was bent. There appeared to be a 

 definite depth of immersion of this wire, at which the apparatus 

 worked with minimum current. In his apparatus this critical 

 position was when half the wire was below the surface of the 

 liquid, the other half projecting into the air. He attributed the 

 fatigue to the presence of gas about the electrodes, for he ob- 

 served that a mechanical tap to the base of the apparatus 

 restored the working condition. Mr. Rollo Appleyard pointed 

 out that the improved result at half immersion, observed by 

 Prof. Minchin, taken together with the phenomena described 

 by Mr. Campbell Swinton as to the effect of dipping the elec- 

 trodes into the solution, suggested that the liquid immediately 

 around the submerged part of the wire was at some instants in 

 the spheroidal state. The breaking-down of the spheroidal 

 state would be facilitated by heat lost by the immersed part to 

 the non-immersed part of the wire. The capacity for heat of 

 the non-immersed part, and the degree of roughness or smooth- 

 ness of the immersed part, would thus appear as factors in the 

 explanation. No doubt the evolved gases were the primary 

 cause of the interruption of current, but the wire having once 

 become red-hot the spheroidal condition would introduce 

 a further cause of electrical separation between the wire 

 and the liquid. Prof. Vernon IJoys asked whether it was 

 the liquid or the electrodes that became fatigued. Ex- 

 periments should be made to determine the effect of 

 variations in the hydrostatic pressure around the platinum 

 electrode. Mr. T. H. Blakesley said that the rise of 

 potential at the terminals of the interrupter proved that the 

 arrangement possessed capacity. Such a rise of potential could 

 not occur without there bemg capacity, any more than it could 

 without self-induction. Mr. D. K. Morris described experi- 

 ments he had made with a Wehnelt interrupter, using a i 

 kilowatt -transformer with a transformation of 4 to 5, intended 

 for 10 amperes at 100 volts. The anode of the interrupter was 

 designed to have an adjustable surface to correspond with the 

 load on the secondary — a platinum wire at the end of a copper 

 wire could be projected more or less through the drawn-out 

 lower end of a glass tube containing oil. The best results with 

 the interrupter were obtained with about 45 volts on the 

 primary circuit. At this pressure, an average current of i 

 ampere sufficed to give 125 (alternating) volts very steadily on 

 the secondary. As measured by an electrostatic instrument, 

 the "no-load" loss was only 45 watts. The secondary could 

 then be loaded up with lamps, provided that the exposed surface 

 of platinum wire was proportic^ately increased. The energy 

 delivered to the lamps, however, was not at any load much 

 greater than 45 per cent, of that taken from the mains. 

 By connecting the interrupter with a condenser of i microfarad, 

 the efficiency at small loads was increased to nearly 60 per cent. 

 He had observed that the fatigue of the interrupter could be 

 temporarily remedied by reversing the current. Mr. C. E. S. 

 Phillips asked whether Mr. Campbell Swinton had tried other 

 liquids than dilute sulphuric acid. So far as his own experi- 

 ments went, he had only obtained good results with that electro- 

 lyte. Mr. Campbell Swinton, in reply, said that with the appa- 

 ratus arranged in a simple circuit, an alternating current applied 



