10 



Prof. E. Edlund's Researches on the Passage 



the air, one is forced to rely exclusively on the observations 

 made with more considerable rarefactions of the air. 



In order to be able to determine how e and r^l vary, each 

 by itself, with the density of the air, it is necessary to proceed 

 with a series of observations in which e and i\l are combined 

 in a different manner from that in which they were in the 

 preceding series. For this purpose a Ruhmkorff induction - 

 apparatus was employed in the following manner : — Having 

 found the difficulty there was in obtaining constant deflections 

 when the Foucault interruptor belonging to the apparatus was 

 employed, I made use of only one of the induced currents 

 arising from the opening of the inducing current ; and that 

 current was measured by means of a sensitive magnetometer. 

 If the inducing current preserves its intensity without modifi- 

 cation, the induced currents do not undergo, at the opening, 

 any modification with respect to quantity, in whatever man- 

 ner the opening is effected. If the circuit of the induced 

 current is composed of solid and liquid conductors only, the 

 induced currents also produce deflections of equal magnitude 

 on a galvanometer inserted in the circuit. But when the cir- 

 cuit of the induced current is interrupted so that the current 

 is compelled to traverse an extent of air, for the deflections to 

 be equal it is necessary that the opening always take place in 

 the same manner. After some fruitless attempts, this could 

 be done with the aid of a spiral spring arranged in the manner 

 indicated by fig. 2. A B is 

 a wooden support, on which 

 is placed a glass jar, a, half 

 filled with mercury. On the 

 surface of the mercury is put 

 an ebonite disk perforated 

 at its centre. At g is fixed 

 one extremity of a spiral 

 spring, b, of steel, the other 

 end carrying a small iron 

 cylinder, c, to the opposite 

 extremity of which a thin 

 platinum wire, h 3 is soldered. 

 When the spring is kept 

 stretched by the lever-arm d, 

 the platinum wire descends 

 to the mercury through the 

 perforation of the ebonite disk. One of the electrodes, e, of 

 the battery which produces the inducing current dips into the 

 mercury ; the other, /, is fixed to the iron cylinder instead of 

 being at g, in order that the spring may not be heated and its 





