40 EEPOET— 1888. 



cup of the candle accordiDg to the movement of the surrounding air. 

 But these variations are to some extent neutralised by taking the average 

 of many testings. A source of error, which is at once less easily recog- 

 nised and more grave in its practical result, is the change in the average 

 value of the standard candle, which is due to the improvements made 

 from time to time in the manufacture of spermaceti. The manufacturer 

 aims at separating as perfectly as possible the solid parts of his crude 

 material, which constitute spermaceti, from the liquid parts which form 

 sperm oil. To obtain as muf^h oil and as ' dry ' a spermaceti as possible is 

 the object sought, and we are informed by one of the principal manufac- 

 turers of sperm candles that within the last ten years considerable im- 

 provements have been made in the process by which the sepai'ation is 

 effected. As a consequence of the use in the making of standard candles 

 of a drier sperm, it has been found necessary to provide them with 

 thicker wicks, in order that the required rate of consumption may be 

 maintained. Probably the drier sperm has a higher melting-point and 

 furnishes a less limpid liquid. A thicker wick means less light for a 

 given consumption ; and thus the result of the improvements in sperma- 

 ceti has been that standard candles give less light than they gave ten 

 years ago, and probably still less light than they gave at an earlier date, 

 when the average consumption of a sperm candle of six to the pound was 

 140 grs. per hour. 



(2) Professor Violle's standard of molten platinum is, in the opinion 

 of the Committee, not a practical standard of light. On this point your 

 Committee is anxious to have it understood that it is quite prepared to 

 agree to the adoption of the light emitted by a square centimetre of 

 molten platinum as a unit of light, but not as a standard of light. Hence, 

 in its recommendation, it is not challenging the conclusions arrived at by 

 the International Congress of Electricians. The experiments detailed in 

 the second appendix to this report confirm the Committee iu its belief 

 that there is no means at present known by which molten platinum can 

 be used practically as a standard. In fact, a comparison of Violle's unit 

 with any standard could be made only with great labour and but rarely. 

 Violle's unit not having been universally adopted, your Committee would 

 not propose to change the name of the unit hitherto in use, but would 

 call it a standard candle, giving it that value which appears to have been 

 intended by the Legislature, and was used in the adjustment of the 

 Pentane standard. 



(3) There seems unfortunately no prospect of any reliable electric 

 standard of light being constructed. In the report of 1885 your Com- 

 mittee delayed making a definite recommendation until further informa- 

 tion should be gained by experiment about the laws of radiation from 

 carbon heated by an electric current. The information gained since then 

 has led it to believe that but little help is to be gained from this quarter. 

 Some of the reasons are given in the report of 1886 ; others are the 

 variation in light by the blackening of the glass bulbs in which the 

 carbons are enclosed, and by the wasting of the carbon filaments. Also 

 the amount of radiation from carbon depends upon its surface and upon 

 the treatment to which it has been subjected. 



(4) The Amyl-acetate standard is Yery constant, but its red colour is 

 a serious objection to its use. This conclusion has been arrived at as the 

 result of a very large number of experiments in the hands of the Com- 

 mittee. The following experiments by the Committee show this well : — 



