April 17, 1902 J 



NATURE 



555 



tables ; but the method now adopted is attended with 

 the great disadvantage of adding enormously to the 

 labour of examining the work critically. 



In 1 858 the most serious desideratum in Russian 

 meteorology was the large number of undetermined 

 heights among the stations. But shortly thereafter 

 Wild commenced vigorously to make good this great 

 defect, and his successor, Rykatchew, is successfully 

 carrying out this good work. It is much to be desired 

 that in a few years it will be completed. The more im- 

 portant desiderata still outstanding are large portions of 

 northern and south-eastern Siberia. 



The new hypsometrical and meteorological data put 

 before us in this atlas are exceedingly valuable accessions 

 to meteorology. For while the broad features of the geo- 

 graphical distribution of pressure and temperature, as pre- 

 viously disclosed, remain substantially the same, yet the 

 fresh data now submitted result in many cases in more 

 accurately defined positions of the isobars, isotherms 

 and wind direction in their occurring changes from 

 month to month. 



It is truly a genuine pleasure to the climatologist to 

 use these rectified monthly isobars and isotherms in 

 explanation of the monthly changes in the geographical 

 distribution of snow and rainfall, number of days of pre- 

 cipitation, humidity, cloudiness and other weather 

 phenomena in their all-important bearings on the agri- 

 cultural and other economic interests of the Russian 

 Empire. Prof. Rykatchew and his singularly able staff 

 are to be congratulated on the successful termination of 

 this great work. Alex.^nuer Buchan. 



A FRENCH CRITIC OF MAXWELL. 

 Les Theories dlectriques de J. Clerk Ma,x'well. Etude 

 historique et critique. Par M. P. Duhem. Pp. 228. 

 (Paris : A. Hermann, 1902.) Price fr 8. 



IN this work the earlier writings of Maxwell on 

 electrical subjects, as well as his " Treatise on Elec- 

 tricity and Magnetism," are discussed. The general 

 attitude is somewhat severe, as may be inferred from the 

 following extract translated from the preface : — 



" The different theories of the .Scoitish physicist are 

 irreconcilable with the traditional theory ; they are 

 irreconcilable with each other. ... At each instant it 

 seems that the result even of his own reasoning and of 

 his calculations is going to drive Maxwell to impossible 

 and contradictory results ; but . . . Maxwell makes an 

 embarrassing term disappear, changes an inacceptable 

 sign, transforms the meaning of a symbol ; then, having 

 passed the dangerous spot, the new electric theory, en- 

 riched by a paralogism, pursues its deductions.'' 



Of the electrostatics in the paper " On Faraday's lines 

 of force," which Prof. Uuhem regards as the first of three 

 different theories of statical electricity propounded by 

 Maxwell, his concluding remark is that it is only the 

 semblance of a theory. Maxwell states sufficiently 

 clearly, we should have thought, that his object was 

 not to establish any physical theory, but to point out 

 certain analogies between lines of force and lines of flow. 



Of the theory developed in the paper " On physical lines 

 of force," Prof. Duhem writes that it does not even 

 lead to the expression in equations of the problem of 

 NO. 1694, VOL. 65] 



the polarisation of a given dielectric medium ; this seems 

 to be because Maxwell assumes without formal proof 

 that the function whose spacial differential coefficients 

 express on his theory the electromotive force is identical 

 with the potential of the classic theory ; it seems very 

 easy to rectify the omission. 



Altogether too much capital is made of Maxwell's 

 unfortunate confusions of sign, and it seems puerile to 

 complain of the use of the popular term " electric tension " 

 where " electric pressure " is required by strict analogy. 



Prof. Uuhem's objection to Maxwell's interpretation of 

 the various terms in the expression obtained for the 

 magnetic force' appears well grounded. 



In discussing the third electrostatic theory of Maxwell, 

 as contained in the paper " On a dynamical theory of the 

 electromagnetic field" and in tlie "Treatise," comments are 

 made on the obscurity of Maxwell's idea of electric dis- 

 placement and on the confusion caused in the form of 

 the equation of continuity by thinking of a charge of 

 electricity sometimes as a real thing and sometimes as a 

 mere fiction representing the effect of nonuniform polaris- 

 ation. All readers of Maxwell know these difficulties ; 

 most will be disposed to agree with Hertz that if we 

 interpret the word " electricity " in a suitable way, nearly 

 all the apparent contradictions can be made to disappear, 

 rather than follow Prof. Duhem in his reductio ad, 

 absiirdum. 



Prof. Duhem contrasts Maxwell's theory of displace- 

 ment through a dielectric considered as a continuous 

 medium having an elastic constant different from that of 

 pure itther with the theory which regards all the pheno- 

 mena as due to action at a distance on the analogy of 

 Poisson's theory of induced magnetism. His expression 

 for the electrostatic energy on the latter theory contains 

 a term which is furnished by the surface separating two 

 different dielectrics and which corresponds to the ficti- 

 tious charge of electricity due to change of polarisation 

 on crossing that surface ; consequently he infers that 

 this theory clashes with that of Maxwell. Considera- 

 tion of the work actually required to charge the 

 conductors in such a case renders it difificult, how- 

 ever, to see how on either theory the expression for 

 the total organised energy can differ from that given by 

 Maxwell. Prof. Duhem then refers to Gouy - as having 

 shown that the classic doctrine completely explains the 

 actions observed between conductors and dielectrics by 

 Pellat, 2 among others, and concludes that such actions 

 could not be deduced from Maxwell's theory. In the 

 absence of precise numerical calculation and comparison 

 with experimental results, this inference appears some- 

 what rash, even if the premisses were correct. Gouy, 

 however, in the paper referred to does not consider 

 the case of two different dielectrics at all. Pellat, 

 moreover, considers the results of his experiments to 

 be in perfect accordance with Maxwell's theory ; he 

 points out, however, that as his calculations of the effec- 

 tive forces on the surfaces of conductors and dielectrics 

 are deduced from the variation of the electrostatic 

 energy, his experimental results, as well as those of 

 Quincke, may be explained without accepting Maxwell's 



1 "Scienlific Papers,' vol. i. p. 463. 



- Journal dc Physique, 3" serle, t. v. p. 154, 1896. 



3 Annates de CltiuiU et de Physique, j" strie, i. v. 1895. 



