Candle-Poicer of Incandescent and Arc Lamps. 137 



consisting in stating the whole flux of light in lumens per 

 Avr.tr. which gives a figure of about 3*5 for glow-lamps. 



The expression of the efficiency in this manner involves the 

 determination of the mean spherical candle-power ; and the 

 method described in this paper affords accurate results with 

 no greater expenditure of time than that required for one, or 

 at most two, measurements by existing methods. 



Apparatus and Method. 



The experiments were carried out with the following objects 

 in view: — 



(i.) To obtain a series of curves showing the variations of 

 candle-power in incandescent lamps, in a horizontal plane. 



(ii.) To obtain reduction factors by which the mean 

 horizontal candle-power may be calculated from the maximum 

 horizontal candle-power, for different types of filament. 



(iii.) To obtain reduction factors for obtaining the mean 

 spherical candle-power from the maximum or mean hori- 

 zontal candle-power. 



Objects (i.) and (ii.) were met by designing a carriage, 

 to run on the existing photometric bench in the Pender 

 Electrical Laboratory, provided with a scale of degrees and 

 means for rotating the lamp. 



The carriage maybe briefly described as follows : — 



A strong cast-iron frame A (PI. III. fig. 1) carries four flanged 

 wheels B and a cord grip C. In the centre of the frame a 

 bearing for a vertical shaft D is arranged, supported at E by 

 a collar-bearing. This shaft is hollow for more than half its 

 length, and carries a second sliding tube, which can be fixed 

 in any position by the nut F. To the upper end of this is 

 screwed the lamp-holder G. 



On the lower portion of the shaft is fixed a bevel-wheel H 

 and a pair of slip-rings J arranged in a horizontal plane on 

 an ebonite base. On these rings press a pair of gauze brushes 

 K K fixed in spring-holders and mounted on insulating 

 supports. 



A scale of degrees is arranged at L (shown dotted in plan), 

 the pointer being fixed to the collar-bearing E. 



The vertical shaft is driven by a motor M carried on the 

 same carriage. The motor is mounted in a gun-metal ring N 

 by throe screws, which, together with the slots at 0, allow 

 of a small movement in any direction. 



A pinion on the motor-shaft en^aiit^ with the bevel-wheel 

 H, the reduction ratio being 4 to 1. The motor voltage is 

 adjusted until the lamp rotates about four times per second. 



Two pairs of flexible leads are brought from suitable 



