292 Obituary Notices of Fellows deceased. 



liim " to throw a new light on thermodynamics " ; and, in the early 

 eighties, copies of the memoir were there highly valued, but procured 

 with difficulty. In the abstract of Maxwell's discourse, published in 

 "Proc. Camb. Phil. Soc.," iv, p. 427, the theory is illustrated from 

 F. Guthrie's published experiments on selective precipitation by the 

 introduction of solid substances into a mixed solution, and by observa- 

 tions recently made, perhaps for this purpose, by P. T. Main, on the 

 phases of the triple system composed of chloroform, alcohol, and water ; 

 while the simple and natural exposition of the nature of catalytic 

 .action was singled out' for remark. 



It is characteristic of Prof. Gibbs' extreme care for completeness 

 .and perfect elaboration, that probably the most interesting and 

 instructive account in existence of this great memoir of over 300 

 pages is the abstract of 18 pages which he contributed himself to the 

 "American Journal of Science" for December, 1878. 



The nineteenth century will be remembered as much for the 

 establishment of the dynamical theory of heat at the very foundation 

 of general physics, as for the unravelment of the nature of radiation 

 and of electricity, or the advance of molecular science. In the first 

 of these subjects the name of Carnot has a place by itself; in the 

 completion of its earlier physical stage the names of Joule and 

 Clausius and Kelvin stand out by common consent; it is, perhaps, not 

 too much to say that, by the final adaptation of its ideas to all rever- 

 sible natural operations, the name of Gibbs takes a place alongside 

 theirs. 



Afterwards Gibbs turned his attention to the electrical theory of 

 light, then in the tentative stage of development, and published in 1882 

 three papers in which the electrical relations forming the foundation of 

 Maxwell's theory were expounded on the most general formal basis. 

 The medium is taken to be heterogeneous (molecular) in its smallest 

 parts, but of an averaged homogeneous structure as regards elementary 

 regions of dimensions comparable with a wave-length. General linear 

 relations of a formal type between the Maxwellian vectors are assigned, 

 involving the case of rotational media when they are not self -con jugate. 

 The precise part of the electrical basis utilised was the universally 

 admitted general type of formula (Neumann-Maxwell) for the 

 kinetic energy of the (circuital) displacement-currents in the field. The 

 object of the papers was to point out how naturally the laws of optical 

 reflexion, and of double refraction including its dispersion, flow from the 

 electrical ideas as contrasted with the mechanical theory of Cauchy 

 and Green. Thus the analysis is in some respects open to the remark 

 .that the electrical foundation is refined and generalized until there is 



