NA TURE 



\_Nov. 5, li 



ordinary transparent media U is so small compared with u that 

 it may be neglected, and puts it zero everywhere. 



The results of the various theorie? differ in the form they give 

 for the dispersion formula. Lommel's theory has been shown 

 by Voigt to be untenable. The theories of Helmholtz, Thom- 

 son, and Sellmeier lead, when /' is small, to the same result, and 

 give 



'n- 



which Ketteler's gives 



m"= I + 



+ 2 



T is the period of the ether vibration, k of the matter vibration, 

 and q, D, &c., are functions of the constants. 



Voigt's formula, since he does not consider the matter motion, 

 is different and not so general. 



With regard to these formula;, I am not aware that Helmholtz's 

 has been tested by comparison with experiment. Ketteler's has, 

 and agrees excellently over a long range of values of t. 



Double refraction is generally explained by sujiposing /3- to 

 be a function of the direction ; but, as Sir W. Thomson has 

 pointed out, this involves for Helmholtz's theory — he did not, 

 however, apply his formulae to crystals — dispersion with double 

 refraction. For Ketteler's theory this is not the case. y. can 

 be a function of the direction independently of t. 



The mechanism which would make the action between the 

 matter and ether in each element of volume a function of the 

 acceleration is perhaps not so easy to conceive a; that supposed 

 by Helmholtz and Thomson ; but still Ketteler's theory seems 

 to overcome some of the difficulties inherent in the latter. 



Either of these theories can be shown to lead to Fresnel's 

 wave-surface, provided we do not consider it necessary that the 

 vibrations should lie in the wave-front. The vibration, as 

 indeed Ketteler and Boussinesq have pointed out, will be nor- 

 mal to the ray. In all other respects Fresnel's construction will 

 hold. I 



Ketteler and Voigt have tried, without much success, to apply I 

 their theories to reflection and refraction. I 



Thomson, in that most valuable appendix to his Baltimore 

 lectures, has given a complete theory. This can be readily 

 adapted to Ketteler's theory, and the results in many points 

 agree in a striking manner with experiments both for transparent 

 and op.aque bodies. The occurrence of a real negative value 

 for /u- is explained by the supposition that the period of the 

 incident light is higher than the highest possible mode of vibra- 

 tion for the matter-molecules in the medium. 



The l.ast section deals with Maxwell's eleclro-ma^netic theo: .■ 

 of light. 



Electro-magnetic disturbance travels in air with a velocity 

 equal to that of light ; and in a double refracting medium obeys 

 Fresnel's laws. The difficulty lies in giving a physical explanation 

 of light motions, an:! of accounting for the mechanical structure 

 of the ether required by the theory. No complete theoiy of dis- 

 persion has yet been given. The work of Willard Gibbs does 

 not explain why there is no dispersion in a vacuum. The 

 objection made to Cauchy's theory holds good. It is probable 

 that some theory such as is developed in the third section may 

 be successfully applied to the electro-magnetic disturbance. 



The theory has the great advantage of connecting naturally 

 with the theory of light the important electro and magneto- 

 optical discoveries of Faraday, Kerr, Kundt, and Quincke, and 

 to the development of this much is due to Prof Fitzgerald. The 

 theory of reflection and refraction as at present developed is only 

 approximate. 



ELECTROL YSIS 

 pROF. LODGE opened the discussion at the Aberdeen 

 ■*■ meeting of the British Association on Electrolysis by 

 reading a paper, the notes of which have already appeared in 

 Nature. 



Sir W. Thomson referred, in his remarks on Prof. Lodge's 

 paper, to a matter of importance in electro-plating — viz. the 

 selection which takes place in the electrolysis of solutions con- 

 taining several salts, as, for instance, in the electrolysis of copper 

 sulphate containing ferrous sulphate, which, when decomposed 

 by a strong cun-ent gives a deposit containing impurities, where- 

 as a slower decomposition yields a very pure deposit. Sir W. 

 Thomson spoke also of the necessity for the careful investigation 



of those cases in which the formation of deposits between the 

 electrodes had been observed, and it would be important to 

 know whether deposits could be formed in the line of conduction 

 without a nucleus at all. Such matters are of importance to physio- 

 logy, indicating a possible danger in the passing of long con- 

 tinued currents through the human body. 



Prof. Schuster explained the views propounded by Von 

 Helmholtz in his recent papers on this subject. Helmholtz 

 explains the phenomena of electrolysis by assuming a dif- 

 ferent .attraction of different chemical elements for electricity. 

 Tf this be admitted, most of the difficulties connected with 

 the phenomena of contact electricity disappear. In electrolysis 

 the element (say hydrogen) charged with positive electricity 

 travels to the negative electrode and forms a coating over it. 

 Any electromotive force, however small, is sufficient to produce 

 this effect, as no work is done. The hydrogen does not appear 

 as free hydrogen, however. It is only liberated when the 

 electromotive force is sufficient to produce a transfer of the 

 positive electricity from the hydrogen molecule to the electrode. 

 When the dissociated elements appear in a neutral state an 

 interchange of the electricities of the elements must have 

 occurred before dissociation. In this way we may explain the 

 conversion of stannic in stannous chloride, v/hich was mentioned 

 by Prof. Armstrong in his address. Prof. Schuster did not 

 think that Prof. Lodge had laid suflficient stress on the fact that 

 in very dilute solutions an ion has the same rate of transference, 

 no matter with what element it was combined. This fact affords 

 strong evidence in favour of the above views, from which it 

 follows as a necessary result. Prof. Schuster also explained his 

 own views of the discharge of electricity in gases. He believes 

 that the phenomena present some analogy to those exhibited 

 in electrolysis of liquids. The phenomena exhibited at the 

 negative pole are, he thinks, due to dissociation of the compound 

 molecule. They do not appear in the case of monatomic 

 mercury vapour. Experiments which he hopes to conclude in 

 the next few months will decide whether or not the law of the 

 constancy of molecular charge holds. 



The next contribution to the discussion was a paper by Dr. C. 

 R. Alder Wright, containing an account of the nature of his 

 investigations, conducted witli the view of measuring Chemical 

 Affinity in terms of E.M. F. 



On the Sensitiveiitss to Light of Selenium and Sulphur Cells, 

 by Shelford Bidwell, M.A., LL.B.— The author suggests that 

 the operation of annealing in the making of selenium cells in- 

 creases the sensitiveness to light by promoting the combination 

 of the selenium with the metal of the electrodes, forming a 

 selenide which completely surrounds the electrodes, and is, per- 

 haps, diftused throughout the selenium when in a liquid con- 

 dition ; further, that the apparently improved conductivity of 

 the selenium, together with the electrolytic phenomena which it 

 exhibits, are to be accounted for by the existence ot this selen- 

 ide. This view finds considerable support in the fact that cells, 

 constructed with sulphur, replacing the selenium and containing 

 a proportion of silver sulphide, are all more or less sensitive to 

 light, and exhibit properties of annealed selenium. The author 

 also read a paper On the Generation of a Voltaic Current by a 

 Sulphur Cell with a Solid Electrolyte, a short account of which 

 has already appeared in Nature (vol. xxxii. p. 345). 



MOLECULAR HEIGHTS 

 T""HE discussion on the Molecular Weights of Liquids and solids 

 was opened in Section B of the British Association by the 

 reading of a paper by Prof A. W. Reinold, F. R. S., the subject 

 of which was the Size of Molecules. In this paper an account was 

 given of the different lines of argument by which Sir W. Thomson 

 has been led to form an estimate of the size of molecules. The 

 estimate is based upon four lines of argument — the first, from 

 the refractive dispersion of light ; the second, from the phe- 

 nomena of contact electricity ; the third, from liquid films ; and 

 the fourth, from the kinetic theory of gases. All four agree in 

 showing that in liquids and transparent solids the mean distance 

 between the centre of contiguous molecules is something between 

 i/ioth and i/200th of a millionth of a millimetre. Recently 

 Exner (Monatschrift fiir Chemie, vi. 244-27S) has proposed 

 another method for estimating the diameter of gaseous molecules, 

 the results obtained by this method being slightly smaller than 

 those deduced from the above. The author gave an account of 

 his experiments on soap-films, conducted conjointly with Prof 

 Rucker (Nature, vol. xxxii. p. 210), the results of which are 



