134 



SCIENCE. 



[N. S. Vol. I. No. 5. 



and briefly discussed. In a large number 

 of cases Maxwell's relation is confirmed ; 

 but, again, the cases are numerous in which 

 the agreement between theory and experi- 

 ment is far from satisfactory ; this is es- 

 pecially true of dielectrics showing traces 

 of conductivitj' and large electric absorp- 

 tion, and even more true of electrolytes. 

 This part of Poincare's work is rather in- 

 complete, probably because it offers fewer 

 opportunities to a mathematical physicist 

 than any other part of Maxwell's electro- 

 magnetic theory. The most serious critic- 

 ism, perhaps, that may be brought against 

 it is its omission of some of the most im- 

 portant investigations on dielectric con- 

 stants, as, for instance, the investigations of 

 Boltzmann. Again, not a single word is 

 said concerning the influence which the 

 study of the dielectric properties of sub- 

 stances had upon Faraday and Maxwell and 

 how much it had contributed to the forma- 

 tion of their electromagnetic theory. 



The reflection of electrical waves from the 

 surface of a dielectric is taken up and it is 

 shown by a reference to analogous phe- 

 nomena in optics why reflection cannot occur 

 when the thickness of a dielecti'ic plate is 

 small in comparison to the wave leng-th of 

 an electi'ical wave. Trouton's experiments 

 (Nature, Vol. 39, p. 391) form the basis of 

 this discussion. 



The experimental evidence furnished by 

 the study of the reflection of electrical waves 

 is cited which supports the view that the 

 plane of polarization as defined in optics is 

 perpendicular to the direction of the elec- 

 trical force in the wave-fi-ont. 



A very interesting experimental investi- 

 gation published by Klemencic (Wiener 

 Sitzungsber, 19. Feb., 1891) is next de- 

 scribed. It treats of wave reflection by di- 

 electrics. The dielectric experimented with 

 was a slab of sulphur 120 cm. long, 80 cm. 

 wide and 7 cm. thick. The wave length 

 employed was 60 cm. A rectilinear oscil- 



lator placed in the axis of a cylindrical para- 

 bolic mirror furnished the plane waves. The 

 reflected and refi-acted waves were studied 

 by means of thermoelectric couples attached 

 to rectilinear oscillators placed in the axis 

 of parabolic mirrors similar to the one used 

 in connection with oscillator. There was a 

 reflection at every angle of incidence when 

 the du-ection of oscillation of the electrical 

 force was perpendicular to the plane of in- 

 cidence. But when it was parallel to it then 

 there was an angle of incidence at which no 

 reflection occurred. Fresnel's fundamental 

 formulse, however, were not quite satisfac- 

 torilj^ verified. Poincare ascribes it to the 

 insufficient thickness of the slab. Klemen- 

 cic found also that the energy of the inci- 

 dent wave was smaller than the sum of the 

 energies of the reflected and refracted 

 wave, a result which he believed to be due 

 to the presence of diffraction. 



Conductors in Motion in an Electromac/netic 

 Field. — The last chapter gives the essential 

 features of Hertz's essay : On the funda- 

 mental equations of the electromagnetic 

 field for conductors in motion. 



Poincare considers first the electromotive 

 force induced in a circtut which is moving 

 through a variable electa'omagnetic field. 

 He proceeds as follows : Consider a surface 

 formed by the circuit under consideration. 

 Let it move with the circuit. Consider two 

 consecutive positions of this surface, the 

 time of passage from the first to the second 

 position being infinitely short, the velocity 

 of motion being finite. Consider now the 

 space bounded by the initial and the final 

 position of the surface and by the ring- 

 shaped surface whose boundary is the 

 initial and the final position of the circuit. 

 The total magnetic flux through this siu'face 

 is according to well known relations pro- 

 portional to the total amount of what Hertz 

 and Poincare call true magnetism included in 

 the bounded surface. The total induced 

 electromotive force being equal to the total 



