^6 



NATURE 



{Nov. .15, 



Whether a substance will be paramagnetic or diamagnetic will 

 depend, in the first place, on the distribution of lines in its 

 spectrum, and also upon the relative values of Tj and Tg, 

 calculated from equations (39) to (42), and therefore upon the 

 other molecular constants which determine the relation between 

 the wave-length and the period of vibration. The fact of the 

 magnetic behaviour of a substance being partly determined by 

 the values of these molecular constants would appear to make it 

 impossible to predict its magnetic properties from the nature of 

 its spectrum only. It is clear that, according to the author's 

 theory, similar effects might be produced by mechanical vibra- 

 tions, as by heat, for any excitation of the molecules to a 

 sufficient extent must give rise to phenomena of the kind 

 described. 



The results obtained may be formulated in the following 

 statement : — 



If a body is traversed by an electric current in the positive 

 direction, it will give rise to a series of pairs of oppositely- 

 directed currents in the neighbourhood of any molecule, and 

 each of these currents will be equivalent to a luminous vibration 

 of definite wave-length. The body will be paramagnetic, when, 

 in consequence of internal absorption, the excess of the right- 

 handed current over the left-handed one is positive ; and it will 

 be diamagnetic if this excess is negative. 



When a body is magnetized the internal energy of its mole- 

 cules is increased, and therefore it will become heated — a result 

 which is in agreement with observation. The magnetic satura- 

 tion will increase as long as such an increase of internal energy 

 can continue ; but the limit of saturation will be attained when 

 the molecular impacts have excited currents of the same kind 

 as those which were absorbed, and of equal intensity. 



The author explains permanent magnetism by assuming that 

 the molecules of steel can have internal vibrations more easily 

 excited by molecular impacts in some directions than in others. 

 Suppose the sensitiveness to be very small in the direction of the 

 axes of the molecules and very great in perpendicular directions, 

 then vibrations perpendicidar to the axis will be the most easily 

 excited by magnetization, and will be transformed into circular 

 vibrations. The planes of these circles will, in general, be inclined 

 obliquely to the axis, and therefore every molecule will give rise 

 to several circular vibrations in its neighbourhood, and the 

 centres of these circles v ill lie upon a straight line nearly coin- 

 ciding with the axis of the molecule. Every such circular 

 vibration will be equivalent to a small magnet with its axis 

 perpendicular to the plane of the circle, and the poles of these 

 magnets lie in a straight line. There is therefore a molecular 

 rotation in the direction indicated by Weber's theory. Again, 

 when the axes of all the molecules have become parallel, there 

 will be more frequent collisions between neighbouring molecules 

 in a direction perpendicular to the axes than in the direction 

 parallel to them. The critical relations will therefore be excited and 

 communicated to the ether, giving rise to circular currents, which 

 will again be partially absorbed. The mngnetization will be 

 permanent when the mutual action of the molecules and the 

 ether is a steady one. This stationary motion requires a certain 

 supply of external energy, which is continuously transformed into 

 small vibrations of the molecules about their positions of equi- 

 librium. It is an experimental fact that when this sujtply of 

 energy is considerably diminished by cooling the magnet to a 

 sufficient extent, the magnetism is greatly weakened. 



According to this theory, the amount of light absorbed will 

 be equal to the amount emitted, but the latter will have a diff- 

 ferent vibration-period from the former, since T may be different 

 from Tj and Tj, and the critical vibrations most easily txcited 

 by the molecular collisions will in general be of different period 

 from those which are most easily absoibed. Therefore rapidly 

 succeeding molecular impacts may give rise to luminous vibra- 

 tions, the periods of which may be different from those proper 

 to the molecules. 



§ 19. Lorenz's and MaxivcU's Electro-dynamic Theories of 



Light. 

 The author observes that he has preferred to base the explana- 

 tion of electrical phenomena upon those of optics rather than the 

 reverse, because optical phenomena are much more completely 

 understood than electrical phenomena. Lorenz and Maxwell 

 both endeavoured to explain optical by means of electrical 

 phenomena. Lorenz ' bases his speculations upon the resem- 

 blance between the differential equations of the motion of 

 ' Poggendorff's Anna ten. vol. cii. 1856. 



electricity and those which represent vibrations of the ether> 

 which can be made identical by the introduction of certain very 

 small terms. His theory has not been sufficiently developed to 

 admit of its application to the discussion of any definite problem. 

 He comes to the general conclusion that the motions of light 

 consist in electric currents, and that the latter consist essentially 

 in rotatory vibrations of the ether about certain axes. In this 

 point I.orenz's theory presents a certain similarity with that of 

 the author, but in the former no distinction is assumed between 

 the magnitude of electrical and luminous' vibrations respectively. 



Maxwell has developed his theory to a much greater extent. 

 He, too, starts from the similarity in the differential equations, 

 which are different from those of Lorenz, and also from the 

 author's. Magnetic and electrical actions at a distance are 

 attributed to the energy of an intervening medium, and explained 

 by the assumption of the existence of a strain in this medium. 

 The assumption of the identity of the electrically excitable inter- 

 vening medium with the luminiferous ether receives strong 

 confirmation from the fact that the ratio of the electro-magnetic 

 to the elecl:ro-static unit of quantity is the same as the velocity of 

 light. Maxwell arrives at the result that electrical as well as 

 luminous vibrations are entirely transverse to the direction of 

 propagation, but he does not obtain any further analogy between 

 electrical and optical phenomena, and his explanation of the 

 electro-magnetic rofation of the plane of polarization involves a 

 series of complicated hypotheses respecting the action between 

 matter and ether. Maxwell expressly excludes the consideration 

 of molecular structure, and supplies its place by the hypothesis of 

 molecular vortices.^ A further important difference between tlie 

 two theories is that while the author assumes that the material 

 molecules suck up energy from the ether, Maxwell deduces the 

 repulsive actions between two similarly charged conductors from 

 an accumulation of electricity in the intervening medium, 

 especially in the case of an optical excitation of the medium. He 

 does not appear to have arrived at any definite distinction between 

 electrical and luminous energy. 



In the preceding investigations the molecules have always 

 been assumed to be of the same size. If there should be any 

 great -difference in the sizes of molecules in the case of different 

 substances, then the difference between optical and electricaJ 

 phenomena would be entirely relative to the size of the mole- 

 cules of the body considered, so that an ethereal vibration which 

 would give rise to electrical excitation in one body might pro- 

 duce only optical effects in another. The different behaviour of 

 different substances with regard to light and electricity may per- 

 haps depend partly upon this condition as well as on the values 

 of the critical periods and other molecular constants. An 

 interesting question which arises is, What would be the effect of 

 ether vibrations neither very large nor very small in compari- 

 son with the size of a molecule ? The author has not suceedel 

 in obtaining any definite answer to this question. 



§ 20. Conchiding Observations. 



In concluding the paper the author observes that the only 

 hypothesis which he has made use of is that space is filled with 

 a continuous elastic medium — namely, luminiferous ether, the 

 density of which is so small that it may be neglected in com- 

 parison with that of matter. The existence of this is sufficiently 

 established from the known phenomena of light. 



It is not found necessary to assmne a difference in the elasticity 

 of the ether in crystals in different directions, the existence 

 of a special force of chemical affinity, of electric or magnetic 

 fluids, or of molecular vortices. 



Thomson's assumption with respect to the constitution of 

 molecules and their relation to the ether explains the most 

 diverse phenomena of physics and chemistry from a single stand- 

 point — namely, the transference of energy between the mole- 

 cules and the ether, in obedience at every stage to the law of 

 conservation of energy. 



The author then suggests that the theory may possibly provide 

 a means of escape from the conclusion, known as the " Dissipation 



' Maxwell's " Electricity and Magnetism," Arts, iii, 645,794, 830. 852. 

 Ill Art. Ill he says : '• I have not been able to make the next step— namely, to 

 accovmt by mechanical considerations for these stresses in the dielectr.c. ' In 

 Art. 806 the analigy which the author deduces between a solenoid and a 

 circularlx-polarized ray is characterized as faulty on the ground that two 

 opposite circularly polarize'! rays do not neutralize each other, biit produce 

 a plane-polarized ray. The author points out, however, that it is only 

 nece-^sary that the electrical actions should neutralize one another. The 

 simple rela.ion deduced by Maxwell between the specific inductive capacity 

 of a medium and its index of refraction does not follow from the author's 

 hypothesis, and this relation has been shown to be only very roughly true. 



