On a new Standard of Light. 393 



Old Standards for Light-measurements. — Up to the present 

 in England the Standard Candle * has been adopted as the 

 standard of light, the nnit of light being defined as that light 

 which the said candle emits when burning steadily at a certain 

 definite rate. In France the Carcel Burner (bee Carcel) has 

 been introduced as the standard of light, the unit of light in 

 this case being defined as that light which emanates from a 

 good moderator lamp burning pure colza oil at a given definite 

 rate. The ratio of these two arbitrary units is given by 

 several authorities very differently, the mean value being 

 about 



10 standard candles = 1 Carcel burner. 



These two standards of light, although answering perhaps 

 certain practical requirements, are by their nature ill-adapted 

 to form the units of light- intensities. A good and trustworthy 

 standard should possess absolute constancy, or, if not, should 

 afford the possibility of application of a correction for the 

 variation, and moreover should be capable of accurate repro- 

 duction. These qualifications are certainly not possessed by 

 the standards at present in use. 



A candle of whatever compound and size will partake of 

 something of the nature of a complex body, an accurate re- 

 production of which must always be a matter of great diffi- 

 culty. Exactly the same holds good for the Carcel burner. 



Further, the amount of light these standards produce de- 

 pends to a very considerable extent on external influences, 

 which do not allow of easy control or measurement, and which 

 therefore cause variations in the standard light for which it 

 becomes impossible to introduce a correction. For instance, 

 the rate and regularity with which a candle burns and the 

 amount of light it gives depend, in addition to the material of 

 which the candle consists, on the ready and regular access of 

 oxygen. In a closed-up place, like the box of a photometer, 

 if the draught is not well regulated or the supply of fresh air 

 not quite constant, it can be easily observed that the very 

 same candle may emit light at different times varying as much 

 as 50 per cent. Another difficulty is introduced by the varia- 

 tion of the length of the wick, and of the candle itself, by 

 which the standard light necessarily alters its position in the 



* The Metropolitan Gas Act, 1860 (23 and 24 Vict, cap. 125, sect, xxv.) 

 defines the standard candle as : — " Sperm candles of 6 to the pound, each 

 burning 120 grains an hour." I have tried the standard candles as made 

 by two different manufacturers, Messrs. Field and Co. and Mr. Sugg. 

 These candles are sold as six to the pound, and consume, according to my 

 own experiments, about 8"26 grams per hour when placed in a large room 

 and direct draughts excluded. 



