September 26, 1895] 



NA 1 URE 



OJJ 



electricity of waterfalls were all confirmed. The daily curve 

 of atmospheric potential in the valley of I'ontresina shows a 

 maximum at ii a.m., dipping a little and rising again to an 

 afternoon maximum at 5 p.m., then rapidly descending a.s the 

 evening lirceze .sets in. Discussion on the paper related chiefly 

 to the behaviour and temperature errors of portable electro- 

 meters, the latter being somewhat large and quite unexplained. 

 Trof Ayrton .suggested a crucial experiment to determine whether 

 atmospheric electricity is due to an actual distribution in the air, 

 or to induction from the earth's surface. 



The rejiort of the Hen Nevis Observatory for 1894 was pre- 

 sented. The mean hourly velocity of the wind at the top of the 

 mountain, and the mean rainband, are included in the report 

 for the first time. Dr. Buchan and .Mr. Omond have made 

 progress in collating the simultaneous records made at Fort 

 William and the summit ; the differences between them are to 

 be examined especially with respect to their bearing on coming 

 storms. Even at this stage the results indicate that the present 

 theory of cyclones requires great mollification. 



The first part of Monday's sitting was devoted to a discussion 

 on the nature of combination tones. Prof. Riicker gave an ad- 

 mirable account of the history of the subject, pointing out that 

 Helmholtz originated both the theory that they are objective, 

 anil that which supposes them subjective. Me reviewed the 

 theories of Prior and others, according to which summation 

 and difference tones are explained as beat tones of various 

 kinds ; and he called attention to Helmholtz's proof that an 

 asymmetrical elastic body, such as the disc of a microphone or 

 the dr\miskin of the ear, would resound to the difference tone 

 between two notes. Prof. S. P. Thomp.son regretted that in his 

 historical survey Prof. Riicker did not refer to his own work. 

 He read communications from K'inig and Hermann, defining 

 their views. Kiinig distinguishes between beat tones, which can 

 be resonated, antl difference tones, to which the resonator does 

 not respond ; the latter are subjective. Hermann objects to 

 Helmholtz's theory that it is inadequate to account for the loud- 

 ness of the cojnbinational tones. Prof. Thompson mentioned 

 exiJcrimenls to show that difference tones may Ije obtained by 

 semljng one sound to each ear, and in other cases where the 

 drumskin does not receive the sounds. He described also the 

 effect of periodically intermitting a single tone, or of suddenly 

 and periodically changing its jihase, in both of which cases a 

 tone is heard the jiitch of which is the frequency of i)hase-change 

 or intermittence. 



Prof. Everett sought for the cause of combination tones in the 

 air itself, W'hich would be disturbed unsymmetrically by two 

 soimds of finite amplitude. He thought, however, that in the 

 combined effect of two tones, the vibration corresponding to the 

 fundamental Fourier term conunon to each would be louder than 

 the difference tone, a view in which Lord Kelvin concurred. 

 Dr. Burton pointed out that Prof. Everett's explanation of com- 

 bination tones would ai>])ly also to phase tones antl intermittence 

 lone.s. Dr. (I. J. Stoney thought resonance by the mouth -cavity 

 was an important factor in hearing, and in the selection of 

 separate sounds from among a number. There was a general 

 agreement that summation tones have never been heard, and 

 probably do not exist. 



Mr. E. H. Griffiths opened a discussion on the desirability 

 of a new Practical Heat Standard. He showed thattheu.se 

 of water as the standard substance in heat measurements had 

 led to great confusion, on account of (he various a.ssumptions 

 as to its variation of heat-capacily witli temperature. The 

 curves of heat-capacity of water and temperature, used by 

 different experimenters, were exhibited ; according t<j which the 

 author's results furnished a value about the mean of tliose of 

 recent observers. Mr. Oriffiths suggests as a heat unit, absolute, 

 independent of any one person's results, and convenient in 

 magnitude, the heat energy of 42 million ergs. To interpret it 

 as a water standard he proposes to take it as the thermal capacity 

 of a gramme of water at 10'^ C, as measured l)y the hydrogen 

 thermometer ; and he gives a formula to find the heat-capacity 

 at other temperatures than 10° C. Lord Kelvin said that Prof. 

 Rankine had |)reviously suggested the dynamical specific heat 

 of water as a standard. Mr. VV. N. Shaw thought it advisable 

 to make a distinction between the numbers for the absolute 

 thermal capacity and the specific heat of a substance. He 

 believed this would be done most simply by taking the thermal 

 unit as the heat energy of a million ergs ; the specific heat of 

 water at 10' C. would then be unity, and its thermal cajiacily 

 42 imits. The choice of a thermal unit has been referred to the 

 Electrical Standards Committee. 



NO. 1352, VOL. 52] 



Dr. C. H. Lees gave an account of the method and results of 

 experiments on the thermal conductivity of mixtures of liquids. 

 The method used was that of Christiansen, in which the heat i.s 

 conducted through the liquid enclosed between two copper discs, 

 and confined by an elxmite ring if necessary. The results show 

 that the conductivity of mixtures of two liquids is less than the 

 value calculated by the ordinary law of mixture, at any rate for 

 water, ethyl alcohol, methyl alcohol and glycerine. Dr. Lees 

 undertook the experiments to verify certain relations .suggested 

 by Prof. IL !•'. Weber between molecular weight, density, 

 specific heat and thermal conductivity. 



\ paper by Prof. Ramsay and Nliss Dorothy Marshall was 

 rea<l by the latter, the subject being a method of comparing 

 heats of evaporation of liquids at their boiling-points, -\fter 

 remarking that the data of heats of evaporation are very scanty 

 and discrepant. Miss Marshall described a method by which 

 two liquids, kept at their boiling points byjacketsof their vapour 

 surrounding them, are boiled by means of equal bare platinum 

 wires heated by an electric current. A comparison of the 

 amounts of the liquids evaporated in a given time gives the ratio 

 of heats of evaporation. For absolute values a special determina- 

 tion was made on benzene by Mr. Criffiths and -Miss Marshall. 

 Alcohol was carefully compared with benzene, and all other 

 liquids were then comjjared with alcohol. Water was very 

 erratic in its behaviour, probably because of its greater electric 

 conductivity. 



Mr. G. U. \'ule exhibited a harmonic analyser. 



At the meeting on Tuesday, Lord Kelvin described the results 

 of experiments for the electrification and diselectrification of air 

 and other gases, made by Messrs. Maclean and Gait, and him- 

 self. In the earlier experiments the air inside a metal can was 

 electrified by points, the can being put to earth : on insulating 

 the can and blowing out the air, the charge acquired by the can 

 was equal and opposite to that of the air. Electrification of air 

 and other gases in gas-holders over water, by points and flames, 

 was also tried, greater electric densities being thus obtained than 

 by the previous method. The maximum efl'ects were I'J x' 10"'' 

 electrostatic units per c.c. for air, and 2 '2 x io"'forCO.^. The 

 gases were diselectrified by "filtering" them through metal 

 tubes containing conducting wire gauze and cotton wool. \'ery 

 little electrifying effect was found when uncharged air p.assed 

 through a jilatinum tube ic» cm. long and i mm. diameter, 

 until the tube was made red-hot, in which case the air acquired 

 a strong positive charge. Prof. Oliver Lodge suggested the jse 

 of a filter consisting of a metal tube, highly ]ioIished inside and 

 illuminated by an electric beam shining into its interior. Lord 

 Kelvin said that in all Hertz's or Elsterand Geitel's experiments 

 on diselectrification by light, the charge of the air round the 

 illuminated body should be examined. 



Prof. Riicker made a communication on vertical (earth-air) 

 electric currents. At the meeting of the .-Vssociation last year. 

 Dr. Adolph Schmidt accounted for a portion of the earth's 

 magnetism by assuming electric currrents to ]>ass vertically 

 between earth and air. Such currents woidd be shown by 

 the non-vanishing of the line-integral of magneti* force when 

 taken round a closed circuit on the earth's surface. The matter 

 was tested in this way by Messrs. Kay and Whalley, using four 

 independent circuits, three in Great Britain and one in Ireland, 

 and olitaining the data of magnetic force from the surveys of 

 1886 and 1891. The results do not decide the general question, 

 but they show that in the United Kingdom the upward current 

 has certaijily not mcire than one-tenth of the value required in 

 Dr. Schmidt's theory. Lord Kelvin calculated that the current 

 assumed by Dr. Schmiilt (01 ampere per scpiare kilometre of 

 surface) amounts to a removal of the fine-weather charge of the 

 air near the earth 36 times per second. Dr. Rijchevorsel said he 

 understood that magnetic observations were about to lie made in 

 Switzerland, which" would furnish data for similar calculations 

 there. 



Mrs. Ayrton made a communication on the connection 

 between potential difl'erence, current and length of arc, in the 

 electric arc. The results of carefully-performed experiments, 

 verified also by recalculation from the data of other observers, 

 show that the foUosving relations hold :— ( I ) For constant length 

 of arc the power (number of watts used in the arc) is a linear 

 function of the current ; (2) for constant currents the power is 

 a linear function of the length of arc : {3) for constant length 

 of arc the curve of potential difference and current is a 

 rectangular hyperbola. .\ll these laws are included in the 

 President's statement that the surface with potential difl'erence, 

 current and arc length as coordinates, is a hyperbolic paraboloid. 



