402 On a neiv Standard of Light. 



are proportional. These suppositions are almost correct for 

 small variations of the current. 



In conclusion it may be stated that it was ascertained that 

 the platinum light-standard (B) produced the unit intensity 

 of light (the unit of light equal to the light emitted by the 

 standard candle) at a total expenditure of energy equal to 427 

 £2 ergs per second. Of these, 300 H ergs were actually trans- 

 formed into light by heating the platinum up to a high tem- 

 perature ; while the remaining 127 H ergs were lost for 

 illuminating-purposes, being used for raising the temperature 

 of the circuit exclusive of the platinum standard. 



The platinum light-standard (A) being made of much 

 thicker platinum sheet, showed a much less favourable result. 

 The unit of light by (A) was produced at a total expenditure 

 of energy equal to 1226 XI ergs per second, of which 725 12 

 ergs were actually transformed into light ; the remaining 501 

 XI ergs were wasted in heating the circuit to low temperature 

 (no light). Considering that the unit of light can be pro- 

 duced in an electric arc at a total expenditure of energy of 10 

 12 ergs per second only (see my ' Precis of Report on Electric- 

 Light Experiments,' London, 1st Nov. 1878, p. 11), when 

 produced by Siemens's intermediate dynamo-electric machine, 

 it follows that, from an engineering point of view, light by 

 incandescence can scarcely be expected to compete with light 

 by disintegration (electric arc). 



In fact, it appears that light by incandescence is scarcely any 

 cheaper than light by combustion. The reason for this is that 

 the temperature of an incandescent platinum wire is not very 

 much higher than the temperature of a flame, and that for unit 

 volume the mass which has to be kept heated in a piece of 

 platinum is much larger than the mass in a flame. Unless we 

 should be fortunate enough to discover a conductor of electri- 

 city with a much higher melting-point than platinum, and 

 that the specific weight and specific heat of that conductor is 

 also much lower than for platinum, and that at the same time 

 the new conductor does not combine at high temperatures 

 with oxygen, we can scarcely expect that the principle of in- 

 candescence will be made use of for practical illumination. 



Further, it was ascertained that the resistances of the platinum 

 light-standards (not including contact resistance) were as 

 follows : — 



P. L. S. (B) = 0*136 ohm at 22°'2 C. 



=. 0*876 ; , at the temperature of the standard 



where the light was measured, or increase Tfrtn = 6-44. 



