Illumination and the Measurement of Light, 271 



light ilhimination we must not depend upon any direct comparison 

 between the light emitted by the source to be measured and any 

 given recognised standard of light ; but that we should rather make 

 our standard of comparison an area illuminated to a given intensity 

 whatever be the source of light. We do not want to know so 

 much the intensity of the light emitted by a lamp, as the intensity 

 of illumination of the surface of the book we are reading, or of 

 the paper on which we are writing, or of the walls upon which 

 we hang our pictures, or of the surface of the streets and of the 

 pavements upon which the busy traffic of cities circulates. Illumi- 

 nation of this character depends not on one source only, but on 

 many sources of light distributed in innumerable ways. Hence, I 

 propose to measure the illumination of surfaces quite independent of 

 the sources of light by which they are illuminated. The difficulties 

 in comparing the illumination of spaces have hitherto been insur- 

 mountable, but now, thanks to the beautiful little glow lamps intro- 

 duced by Mr. Swan, these difficulties have disappeared. 



The standard I propose is the space illuminated by 1 British 

 standard candle at 12'7 inches distance, which is the same as the 

 illumination give n by 1 French standard "bee " on the same space at 

 1 metre distance. 



2. The principle of measurement which guides us in nearly all the 

 methods at present in practical use is to compare two surfaces equally 

 illuminated to the same intensity by lights placed at different dis- 

 tances, and to compare them by the law of inverse squares determined 

 by Kepler. This law, however, assumes a clear space through which 

 the light is transmitted, and the emission of this light from a point. 

 Distance therefore becomes essential, and the error introduced by 

 large flames or by the diffusion of light becomes difficult to eliminate ; 

 indeed, by the present system it is well-nigh impossible to measure 

 satisfactorily the light-giving power of several lamps distributed over 

 a room or large area. Moreover, the principle ignores the presence 

 of different colours, and it involves the necessity, with electric arc 

 lamps, of taking measurements at different angles. 



Up to the present time, the simplest and perhaps the most favourite 

 mode of measurement is Rumford's plan of contrasting shadows, but 

 Bouguer's method of comparing equally illuminated surfaces, as 

 carried out by Ritchie and Bunsen, is also in great favour. 



3. All these plans, however, are dependent upon the establishment 

 of a uniform standard of light. Although in England, Parliament has 

 legalised the standard candle, it has been found so inconvenient in 

 practice, owing to its want of uniformity, that Mr. Yernon Harcourt 

 has felt it advisable to replace the standard spermaceti candle with a 

 flame resulting from the combustion of pentane. In France, again, 

 the standard is the Carcel bee, burning at the rate of 0*0042 kilo, of 



