new Standard of Light. 395 



therefore the light produced by a standard cannot alter per- 

 ceptibly in time. To make the light constant from the moment 

 the current passes, L e. to establish dynamic equilibrium be- 

 tween the heat produced and the heat lost per unit of time, it 

 is necessary to make the arrangement in such a manner that 

 the electric resistance offered by the standard is only in the 

 piece of platinum intended to be made hot by the current, and 

 not in the other parts of the circuit. 



For this reason I find it best to cut the piece of platinum 

 out of a platinum sheet. 



Figure 7, Plate XI. gives the form in actual size. The two 

 ears, left white in the drawing, may then conveniently form 

 the electrodes between the leading wires and the piece of 

 U-shaped platinum which has to produce the light. As the 

 U-shaped portion is left in its natural connexion with the ears, 

 the contact takes place over a large surface ; and therefore the 

 contact resistance must be small. This special form, if the 

 dimensions are defined as well as the weight of the platinum 

 sheet out of which it is cut, can be easily reproduced any- 

 where. Further, it is required to exclude the draught from 

 the heated platinum. This is best done by putting on a cover 

 of thin white glass. One half of it is left white ; the other half 

 is blackened on the inside. This precaution is required in 

 order to ensure that light emanating from one side only of the 

 platinum is used in the photometer ; otherwise light from the 

 back part of the heated platinum would be reflected into the 

 photometer. This part is unknown, and therefore could not 

 be taken into account when measuring the light emanating 

 from one side of another light. In fact, to be able to form 

 right conclusions from photometric measurements, it is neces- 

 sary to arrange the experiment in such a manner that either 

 the two lights under comparison throw the same fraction of 

 the total light into the photometer, or, if this is impossible, to 

 ascertain this proportion accurately. 



The platinum light-standard (P. L. S.), described before, we 

 will call in future A. Sending a current of 6*15 webers 

 through it (15° deflection on my large tangent galvanometer, 

 for which the constant =2-296 C. G. S.), the P. L. S. (A) 

 produces a light equal to 0*69 Sugg's candle, or, 



1 Sugg's candle=l-44 P. L. S. (A) with 

 6*15 webers. 



Hence, if this particular light were adopted as the unit, we 

 might define it as follows : — 



6*15 webers passing through a piece of platinum 2 millims. 

 broad, 36*28 millims. long and 0*017 millim. thick, weighing 



