February 1, 1895.] 



SCIENCE. 



133 



the reflecting surface when the electric force 

 of the incident wave is normal to the plane 

 of incidence ; no retardation takes place if 

 this electrical force is in the plane of inci- 

 dence. The extinction of the wave in its 

 passage through the metal develops heat 

 and Poincare calculates the rate at which 

 the heat is developed by a given current, 

 obtaining the interesting result that it is 

 proportional to the square root of the 

 product of frequency, specific resistance and 

 permeabilit}' . The results of these consider- 

 ations are now compared to expeiiment. 

 The most important experiments bearing 

 ufwn this part of the theory are those of 

 Bjerkness (1. c). A circular resonator 

 having a small plate condenser interposed 

 in place of the spark gap was employed. 

 Between these plates a small aluminum 

 sheet was suspended and measured bj- its 

 deflection the mean square of the potential 

 difference between the plates. The oscil- 

 lator was gradually tuned and the resonance 

 etfect in the resonator measured by the 

 deflection of the aluminum sheet. Six re- 

 sonators of the same dimensions but of 

 different material were investigated. The 

 resonance curve of copper was highest, then 

 followed brass, silver, platinum, nickel and 

 iron, in the same order as required by theory. 

 The resonator decrement of iron, for in- 

 stance, was nine times and that of platinum 

 twice as large as that of copper. To meas- 

 ure the depth of penetration these materials 

 were deposited electrolj^tically, say u"on on 

 a copper resonator, or vice verxa, and the 

 resonator effect measured for the various 

 thicknesses of the deposit. Results agreeing 

 very fairly with the theory were obtained. 

 Propagation of Electrical Wave.-* through Die- 

 lectrics other than Air. — Another crucial test of 

 the correctness of the Faraday-Maxwell the- 

 ory is furnished by the well known relation 

 that the sjiecilic inductive capacity of a di- 

 electric is equal to the square of its index 

 of refraction. This relation is an immedi- 



ate inference fi-om the new electromagnetic 

 theory. Since the index of refraction of a 

 substance is equal to the ratio of the ve- 

 locity of propagation in vacuum to that in 

 the substance it follows that the velocity of 

 propagation of a Hertzian wave in dielec- 

 trics having a specific inductive capacity 

 larger than unity should be smaller than in 

 air. This relation was tested by Blondlot 

 in the experiments cited above by immers- 

 ing Ijoth the conducting wire and the reso- 

 nator in a liquid dielecti'ic and measuring 

 the wave length. Another method liased 

 upon the same principle was that emploj-ed 

 by Rubens & Arons (Wied. Ann. 40 p. 585). 

 The neutral point of a rectjingular resonator 

 was connected directly to one side of the 

 spark-gap of the oscillator. No spark was 

 then observed in the spark-gap of the resona- 

 tor. If, however, the balance of the resona- 

 tor was now disturbed by inserting on one 

 side of it a certain length of wire immersed 

 in a dielectric the spark appeared. The 

 balance was again restored by inserting a 

 sufficient length of wire in the other side of 

 the resonator. The ratio of these two 

 lengths of wire measured the ratio of the 

 velocities of piopagation in air and in the 

 dielectric. 



Another method, first employed by J. J. 

 Thomson (Phil. Mag. :iO, p. 129), was based 

 on the relation which exists between the 

 capacity of a plate condenser and the di- 

 electi-ic constant of the insulator separating 

 its plates. The period of an oscillator or 

 resonator will vary with the dielectric be- 

 tween the condenser plates. Thomson mea- 

 sured the period of an cscillator for vari- 

 ous dielectrics placed between its condenser 

 plates and calculated from it the specific 

 inductive capacity. Several otlier electro- 

 magnetic nu'thods are described briefly by 

 Poincare, and then the stiitical methods, be- 

 longing most of them to tlie pre-llertzian 

 epoch, are passed in quick review. Finally 

 the experimental results are coordinated 



