$//? WILLIAM SIEMENS, F.R.S. 333 



98*5 Volts, and an internal resistance of 18'5 Ohms, giving a 

 current of 7'8 Amperes when the cells are short-circuited. The 

 resistance of a regulator such as Dr. Tyndall used in his experi- 

 ments may be taken at 10 Ohms, the current produced in the arc 



98'5 



would be = 4 Amperes (allowing one Ohm for the 



18-5 + 104-1 



leads), and the power consumed 10 x 4 s = 160 Watts ; the light 

 power of such an arc would be about 150 caudles, and comparing 

 this with an arc of 8,308 candles produced by 1,162 Watts, we 



find that ( -), i.e., 7'3 times the electric energy produce 



\ 160 / 



(Of>AQ\ 

 - ) , i. e., 22 times the amount of light measured horizontally. 

 150 / 



If, therefore, in Dr. Tyndall's arc T Vth of the radiant energy 

 emitted was visible as light, it follows that in a powerful arc of 



8,300 candles, x ', or fully i, are luminous rays. In the 

 10 7*3 



case of the incandescence light (say a Swan light of 20 candle- 

 power) we find in practice that 9 times as much power has to be 

 expended as in the case of the arc light ; hence i x = ^ part of 

 the power is given out as luminous rays, as against J^th in 

 Dr. Tyndall's incandescent platinum a result sufficiently ap- 

 proximate considering the wide difference of conditions under 

 which the two are compared. 



These results are not only of obvious practical value, but they 

 seem to establish a fixed relation between current, temperature, 

 and light produced, which may serve as a means to determine 

 temperatures exceeding the melting point of platinum with greater 

 accuracy than has hitherto been possible by actinimetric methods 

 in which the thickness of the luminous atmosphere must necessarily 

 exercise a disturbing influence. It is probably owing to this 

 circumstance that the temperature of the electric arc as well as 

 that of the solar photosphere has frequently been greatly over- 

 estimated. 



The principal argument in favour of the electric light is 

 furnished by its immunity from products of combustion which not 

 only heat the lighted apartments, but substitute carbonic acid and 

 deleterious sulphur compounds for the oxygen upon which respira- 

 tion depends ; the electric light is white instead of yellow, and 



