' S Or THh. .\ I K.R. 



360 i HOIOMKIMK TEST* . the results are stated in candle-powtr 



should be made at such a distance from the source of light that the latter 



may be retarded at practically a point. Where tests art made at shorter 



inccs, as for example in the measurement of lamps with reflectors, the 



results should always be given as " apparent candle-power " at the distance 



should always be specifically staled. 



861 HAIS FOR < a flu* ofhght in lumen*, or the 



mean spherical candle-power, should always be used as the basis for com* 

 paring various luminous sources with each other, unless there it a dear 

 understanding or itatemen- .try. 



362 l.\ AMifcM KM I. AW i customary to rate 



]>* on the basis of their mean horizontal can hut tn c 



lamps in which the . lumii 



inter, lie comp.i: tie based on .'. flui 



ligl lumens, or on their mean spherical candle-power. 



362a LIFE TESTS. Similar filaments may be assumed to operate at the 

 when their lumens per watt consumed are the i 

 tests are compar.t similar conditions 



as to filam :.i lures. 



363 The SPHERICAL REDLCTION-PACTOR of a lamp 



mean spherical candle- pox- 

 me 



364 x of light in lumens < y a lamp 4 V X mean 

 M! candle-power X spherical redu 



366 The- Si K should only be used when properly 



determined : 1 characteristics of cacn lamp. 



The spherical reduction-factor permits of substantially accurate com- 

 parisons being made between the total lumens, or mean spherical candle- 

 powers of different types of incandescent lamps, and may be used in 

 the absence of proper facilities for direct measurement of the total lu- 

 is, or mean spherical candle-po. 



366 " RBAI K." Where standard photometric measurements 

 are impracticable, approximate measurements of illmmnants such as 

 street lamps may be made by comparing their " reading distances. 



by determining alternately the distances at which an ordinary size of 

 reading print can just be read, by the same person or persons, when all 

 other light is screened. The angle below the horizontal at which the meas- 

 u-nt is made should be specified when it exceeds 15 degrees. Reading- 

 distance methods usually involve the comparison of very faint illuminations 

 and hence the results may be seriously affected by th< effect. 



367 In COMPARING DIFFERES SOURCES not only should their 

 candle-power be compared, but al lalive form, brightness, dis- 

 tribution of illumination and character of light. 



367s The following symbols are recommended in connection with photometry: 



Pholonwlnc magnitude Symbol 



Intensity of light. / International candle. 



Luminous flux. Lumen 



Illumina* Lumen/cm. 1 , foot-candle. 



/? Foot-candle. 



Brightness. * :n. 



Qi: Candle. 



Lighting. L Lumen-second, lumen-boor. 



AI < 1). SPARKING DISTANCES. 



368 Table of Sparking Distances in Air between Opposed Sharp Needle* 

 Points, for Various Root-Mean-Sauare Sinusoidal Voltages, in indie* 

 nnd in centimeters. The table applies to the conditions specified in Sees. 



