Intelligence and Miscellaneous Articles. 



231 



formula wliich gives the period of the oscillations, a formula the 

 tenability of which is doubtful in the case of very rapid oscillations. 

 More recently, E. Lecher has measured the dielectric constants of 

 several substances by the aid of a method founded also on the 

 determination of the wave-length of very rapid oscillations, but 

 without employing any formula. 



The conclusions of M. Lecher are diametrically opposed to those 

 of Prof. Thomson: "Not only," says he, "does the dielectric con- 

 stant, calculated by the aid of the capacity, not become smaller for 

 very rapid oscillations, but it even considerably increases." 



These contradictory conclusions have led me to engage in new 

 researches ; the method which I have adopted is based on the 

 employment of very rapid oscillations, but I do not make use of 

 any formula. 



A large rectangular plate of copper A A' is fixed vertically ; a 

 second, smaller plate BE' forms a coudenser with the first. This 

 condenser can discharge itself by the intermediation of the knobs 

 a and b ; a is in communication with the gas-pipes, b with 

 one of the poles of an induction-coil the other pole of which com- 

 municates with the gas-pipes. When the coil is working the 

 condenser A B is the seat of oscillatory charges and discharges the 



period of which is of the order of the 



of a second. 





25,000,000 



We have thus in the space situated on the side of A A' opposed 

 to B B' a periodic electromagnetic field which — and this is the chief 

 point — has x X as the plane of symmetry. 



Let us fix in this field two square 

 plates C D, C D' parallel to A A' and 

 symmetrical with respect to X X , 

 and then solder to the middle points 

 D, D' of their inner edges two wires 

 terminating at E and E' in two carbon 

 points kept facing each other at a 

 very small distance apart. 



Although the coil is working, we 

 do not observe any light between E 

 and E' ; this results from the sym- 

 metry of the apparatus. If we , jC 



interpose a glass plate between 



A A' and C D, sparks immediately A 

 pass between E and E' ; this is 

 caused by the induction received by 

 C D becoming less strong than that 

 received by C D'. 



Let us now interpose between 

 A A' and CD' a plate of sulphur; 

 if we give to it such a thickness 

 that the inductive action on C D' 

 is equal to that on C D, the sparks disappear between E and E.' 

 Conversely, the disappearance of sparks between E, E' indicates 



B' 



A 



B' 



