32 



NATURE 



[May 



[o, 1900 



(February 1900), I showed that these corpuscles existed 

 in the neighbourhood of a hot wire and of a metal plate 

 illuminated by ultra-violet light, and recently the discovery 

 by Giesel, Curie and Becquerel of the magnetic deflection 

 and electric charge carried by part of the radium radiation 

 may be interpreted as indicating the existence of cor- 

 puscles in this substance. 



I suppose, then, that there is a certain amount of what 

 may be called corpuscular dissociation taking place in 

 bodies ; that some of the molecules of the substance are 

 continually breaking up by the detachment of a corpuscle, 

 and are being reformed by the arrival of another corpuscle ; 

 the result of this is that at each instant there are a certain 

 number of free corpuscles with negative charges distributed 

 throughout the body, while the corresponding positive 

 charges are on the molecules of the body, the corpuscles 

 are much more mobile than the molecules ; indeed, in 

 solids and liquids, the latter may be regarded as almost 

 fixed in comparison with the former. We thus get the 

 conception of a body permeated with corpuscles which 

 are able under forces to move from one part of the body 

 to another. We must remember that, as the particles are 

 charged, any movement will be accompanied by electrical 

 efifects and, in general, a volume density of electrification. 



The actual number of corpuscles free at any instant is 

 the result of an equilibrium between the number of cor- 

 puscles produced by dissociation and the number which 

 recombine. Thus if 5^ is the number of corpuscles pro- 

 duced by dissociation in unit volume in one second, t the 

 time during which a corpuscle is free {i.e. the time which 

 elapses between its departure from one molecule and its 

 entry into another), // the number of free corpuscles in 

 unit volumes, then when there is equilibrium q = }iJT or 

 n=Tq = \qlu, if X is the mean free path of the corpuscle 

 and u its velocity of translation. In non-conductors we 

 suppose that there are very few corpuscles, but that they 

 are abundant in metallic conductors. Let us now trace 

 some of the consequences of the existence of these cor- 

 puscles in a solid, and suppose for the moment that the 

 positively charged molecules are fixed ; if the corpuscles 

 are acted upon by gravity (of which point we have 

 no evidence), then in a vertical bar of metal the 

 number of corpuscles in unit volume will be greater 

 at the bottom of the bar than at the top, for just 

 the same reason as the density of the air gets less 

 as we go higher ; thus in this case gravity would pro- 

 duce a displacement of electricity, the bottom of the 

 bar being negatively and the top positively electrified. 

 Again, in a rotatmg mass of metal the centrifugal force 

 would tend to drive the corpuscles towards the surface ; 

 there would thus from this effect be an excess of the 

 corpuscles near the surface and a deficit near the axis. 

 Thus the outer parts of the metal would be negatively and 

 the inner parts positively electrified, the rotation of the 

 negatively electrified corpuscles being no longer com- 

 pletely balanced by that of the positively electrified 

 molecules would give rise to a magnetic field ; thus a 

 large mass of rotating metal would act as a magnet. 

 Again, suppose we place a piece of metal in a magnetic 

 field, the action of the magnet on the moving corpuscles 

 will make them describe curved paths, and we can easily 

 see that the magnetic effect due to the particles moving in 

 this way is in the opposite direction to that of the external 

 magnetic field. Thus a metal containing these corpuscles 

 would tend to act like a diamagnetic substance. Again, 

 suppose the metal is exposed to an electric force X, the 

 corpuscles will acquire an average velocity along x equal to 

 XTe/2m, where m is the mass of a corpuscle and e its 

 charge. Let us call this velocity 7^X, then the electric 

 current across unit area is nevX ; thus nev or qe^X'/2inu^ is 

 the specific conductivity of the substance. If we suppose 

 that u, the mean velocity of translation of the corpuscles, 

 varies with the temperature in the same way as the velocity 

 of translation of the molecules of a gas, //tu- would be pro- 



NO. [593. VOL. 62] 



pottional to the absolute temperature, and the specific 

 resistance would, considered as a function of the absolute 

 temperature 0, vary as dlq ; if q, the amount of ionisation 

 increases as the temperature increases, the resistance will 

 vary more slowly than the absolute temperature ; if q 

 diminishes as the temperature increases, the resistance 

 would vary more rapidly than the temperature. These 

 corpuscles moving from place to place would carry not 

 merely electric charges, but energy from one part to 

 another ; and since the coefficient of diffusion of these cor- 

 puscles is proportional to 7/, the thermal and electric con- 

 ductivities would be proportional to each other. Again, 

 when we have conduction of heat we have unequal streams 

 of these corpuscles in opposite directions ; thus the un- 

 equal deflection of their paths produced by a magnet 

 would give rise to an electric displacement, and we should 

 have an electromotive force at right angles to the mag- 

 netic force and to the temperature gradient, an effect 

 discovered by v. Ettinghausen and Nernst. From the 

 conductivity of the gas we can deduce the value of nev. 

 We know the value of e, and hence another equation 

 would enable us to determine n and v ; for this purpose 

 we turn to the Hall effect, but here the results are dis- 

 appointing, for we can easily prove that when E* and E 

 are the tranversal and longitudinal electric forces and H 



the magnetic force, EVEH - '^4^ -J'-A^ where t/j and 



v.> are respectively the velocities of the negative corpuscles 

 and positive molecules under unit electric force, and ky 

 and >^2 the values of k for these ions where k == pres- 

 sure -^ number of systems in unit volume. If both the 

 negative corpuscles and the positive molecules behave 

 like perfect gases, k^ = k.^ and EVEH = \v^, since v.^ is 

 very small ; thus, on this supposition, the Hall effect would 

 give us the value of v ; but there seems no reason to 

 suppose that the positively electrified molecules in the 

 solid would produce the same pressure as an equal 

 number of molecules in the gaseous state, and thus 

 though v.^ is small compared with t/j, k.^ may be so small 

 compared with k^ that k^v.^ cannot be neglected in com- 

 parison with k.2Vy, and in this case the Hall effect would 

 not be sufficient to determine v. The fact that the Hall 

 effect is of different signs for different substances shows 

 that we have to take into account both terms in the 

 expression for EVEH. 



Again, if different parts of a metal bar were at different 

 temperatures, the "pressure" as it were of these corpuscles 

 would be different at different parts of the bar, so that 

 the corpuscles would tend to flow from one part of the 

 bar to the other, and cause an electric displacement ; 

 thus difference of temperature would cause an electric 

 displacement. This is the Thomson effect, measured by 

 the "specific heat of electricity." The value of the 

 "specific heat of electricity" will on this theory depend 

 not only on the variation with temperature of the kinetic 

 energy of a single corpuscle, but also on the way the 

 dissociation constant q varies with the temperature. 

 There are many other phenomena which can be inter- 

 preted in terms of these corpuscles, but these I must 

 leave for another occasion. J. J. Thomson. 



Cavendish Laboratory, Cambridge, April 30. 



SCIENCE IN RE LA TION TO ART AND 

 INDUSTRY. 



AT the annual banquet of the Royal Academy on 

 Saturday evenings Sir Norman Lockyer, in replying 

 on behalf of science, made the following remarks upon the 

 intimate relation between intellectual progress and the 

 study of nature, and also upon the necessity for a more 

 liberal provision for scientific work if England wishes to 

 compete successfully with other nations struggling for 

 industrial supremacy. Though the public mind may be 



