I3 8 ELECTRIC LIGHTING. 



" These experiments," he says, " were made with Carre* car- 

 bons and a Serrin lamp. It was observed that with naked 

 carbons those of the smallest diameter had the longest points, 

 as might have been expected ; but with the metallized carbons 

 the reverse was the case, a circumstance difficult to explain. 



" We may, however, conclude from these experiments : 



" i. That independently of the improvement in the shape of 

 the positive carbon, covering it with nickel lengthens the duration 

 by 50 per cent, of the carbons of 9 millimetres, and by 62 per 

 cent, that of the 7 millimetre carbons. Covering with copper 

 also increases the duration by an amount intermediate between 

 that of the naked and of the nickelized carbons ; 



" 2. That with equal sections the metallization of the 

 carbons does not seem to modify the amount of light yielded 

 by them ; 



" 3. That the luminous power of the carbons of small diameter 

 is much superior for the same electric intensity to that cf car- 

 bons of large diameter, which is explained by the fact that con- 

 ducting bodies placed in a circuit composed of conductors of 

 large section or of high conductivity become the more heated as 

 their diameter is smaller ; and also from the fact of the polariza- 

 tion being the more energetic, the smaller are the carbons. This 

 the more concentrates the resulting calorific effects of which we 

 have spoken above. It depends also upon the circumstance, 

 that to obtain the maximum of light it is necessary that the re- 

 sistance of the circuit of the voltaic arc should be as nearly as 

 possible that of the generator ; 



"4. That the metallization by allowing carbons of small 

 instead of large section to be used for the same time of action, 

 gave advantageous results. This metallization is effected gal- 

 vanically." 



According to A. Ikelmer, this system of metallization can 

 have no advantage except in so far as it may lessen the re- 

 sistance of the carbons, and as the thin layer of metal is 

 oxidized by the influence of the high temperature and disin- 

 tegrated for a distance that may reach as much as 10 centi- 

 metres, an improvement of the conductivity cannot in this 

 way be obtained, any more than the prevention of lateral 

 combustion. Consequently, he thinks that the problem 



