Intelligence and Miscellaneous Articles. 375 



of a galvanometer-wire is put to earth while the other is connected 

 with the second flame, an electrical current is produced as soon as 

 the machine is worked. Lighting the flame did not affect the 

 needle of the galvanometer ; electrostatic induction was completely 

 avoided, as the needle was completely surrounded by a copper 

 damper, and the latter was put to earth. 



]t was found that when the machine was rotated with a constant 

 velocity the galvanometer showed a constant current, the direction 

 of which changed with the direction of the poles of the machine ; 

 it was observed that these opposite currents were never equal to 

 each other. Wiedemann's galvanometer (Edelmann's construction) 

 was used (one division corresponds to 5*99 . 10 -8 ampere) ; or a 

 specially constructed pure metallic galvanometer of W. Lerman- 

 tow, with a bell-magnet (one division corresponding to 2*21 . 10 -8 

 ampere without astatic magnet, and 2-05 . 10 - 9 ampere with astatic 

 magnet). 



When the flames were at a distance of E metres, the following 

 deflections A : and A 2 were observed, corresponding to the case in 

 which one or the other conductor of the machine was connected 

 with the flame : — 



E 1-09 1-68 2-54 4-19 



93 n 9 -, 9 f Lermantow's 



^ Z6 iX Z lA 1 galvanometer. 



A 2 30 7 2 



and with E = 0*305 A x was found equal to about 300 divisions. 



The same phenomenon is also observed by taking two pairs of 

 insulated flames, so that between the machine and the galvanometer 

 there are air-spaces ; if a galvanometer is brought between the 

 middle flames a constant current is also observed here. 



The deflections of the two galvanometers are seen to be always 

 proportional, which may serve as a proof that the circuit in ques- 

 tion is a closed one. A metal or a non-metallic screen insulated 

 from the earth and placed between the flames gives directly a de- 

 crease of the strength of 1-45 times in the case of glass, and 1*25 

 for metal ; with an uninsulated screen the deflections of the needle 

 disappear altogether. 



Instead of the machine a voltaic battery may be used ; the phe- 

 nomenon is of course much feebler. 



The author connects further one pole of an inductorium to earth 

 and the other with the flame, while two telephones connected in 

 series are used instead of the galvanometer. With an air-space 

 between the flames of 6 metres (or as much as 11 metres when the 

 telephone is connected with a large metal shield), the telephones 

 sound quite distinctly ; hence intermittent currents are transmitted 

 directly through the air. 



If, now, the air does actually conduct the electrical current, the 



