460 SCIENTIFIC RECORD FOR 1884. 



meltiug" point as the photometric staudard. This standard was sug- 

 gested by Violle, who has made many experiments, first with silver and 

 subsequently with platinum. Having assured himself that the radia- 

 tion from silver was constant during solidification, he i^roposed as the 

 absolute unit of light the radiation emitted by a square centimeter of 

 surface of melted platinum at the temperature of solidification. The 

 platinum must be perfectly' pure and must be melted in a lime crucible. 

 By means of a diaphragm blackened on its surface, having a square 

 opening, it is easy to obtain a beam of definite cross- section. When 

 the measurement is to be made, the gas is shut oif and the liquid metal 

 is, allowed to cool. This, which is at first rapid, becomes slower and 

 slower, becoming finally stationary. . Then a "lightening" passes over 

 the surface, and the cooling continues. The moment of measurement 

 is at the instant when the temperature is stationary. By continuing 

 the flow of gas, instead of cutting it entirely off, this point may be pre 

 served for a longer time. The author has compared the light thus ra- 

 diated with that of a staudard carcel lamp, first, when the light was 

 emitted by the platinum at an angle of 45^, and, second, when it was 

 emitted normally and reflected from a mirror at 45°. The photometer 

 used in the first comparisons was the E-umford instrument, as used in 

 the light-house service ; in the second, besides this instrument, a Fou- 

 cault photometer was used, of the form used in testing officially the gas 

 supplied to Paris. By the first method the light emitted by the plati- 

 num was 2.118 carcels; by the second, 2.079 for series 1, group A, and 

 2.077 for group B, series 2 giving 2.077. The mean of the whole is 2.08 

 carcels. Assuming the flame of the carcel lamp to have 5.25 sq. 

 cm. of surface, an equal surface of platinum would emit 2.08x5.25, 

 or 10.92 carcels. The intrinsic intensity of the platinum staudard is 

 then 11 times that of the carcel lamp. Comparisons were also made 

 with a Swan incandescent lamp, the light emitted by it under known 

 current conditions having been carefully determined. The mean of 

 these comparisons gave 2.069 carcels. ( C. J?., April, 1884, xoviii, 1032; 

 J. FMjs., June, 1884, II, in, 241; Phil. Mag., June, 1884, V, xvii, 503; 

 Am. J. Sci., July, 1884, III, xxviii, 72.) 



Werner Siemens has contrived an apparatus for putting the above 

 unit into practice, although it determines the light emitted by plat- 

 inum at its melting point and not at its solidifying point. A very thin 

 j)latinum plate is inclosed in a metallic case provided with a hole 0.1 

 sq. cm. in section, which is immediately over the metal. The sides 

 about the hole are inclined toward it, the platinum plate being consider- 

 ably larger. At the instant of melting, the light radiated through 

 the opening is 0.1 of the standard. By suitable arrangements the 

 current may be modified- and controlled so as to produce the melting 

 at will when the comparison is to be made. Preliminary experiments 

 show that the light emitted is about 1.5 standard candles. ( Wied. Ann.^ 

 XXI, 304 ; Am. J. Sci.j August, 1884, III, xxviii, 150.) 



