42 KEPORT— 1 888. 



Mr. Livingston kindly undertook to chronicle and keep the records of 

 the experiments made, and also assist in the testings, and to him the Com- 

 mittee is beholden for the very considerable trouble he took on its behalf. 

 The experiments were conducted as follows: — The central flame 

 being adjusted, one observer made a determination of its illuminating 

 power by candles in the first arm and by the Pentane lamp in the second 

 arm, making the I'eadings alternately. At the same time the second 

 observer made a determination of the illuminating power of the central 

 burner by the Amyl-acetate lamp in the thii'd arm and by the Pentane 

 standard in the fourth arm, also making the readings alternately. The- 

 observers changed positions after each set of observations. The candle 

 readings were corrected in the usual way for sperm consumed. On all 

 the four bars the standards were kept fixed in position and the readings 

 were made by moving the discs. The candles employed were those now 

 manufactured by Mr. Miller and by Messrs. Brecknell and Turner for gas- 

 testing, but tests were made also of other candles, of the same composition 

 as the gas-testing candles, but larger in diameter, which were made by 

 Mr. Miller. The Pentane lamp was of the new form manufactured by 

 Messrs. Woodhonse and Rawson. 



It was found that the central burner was nearly but not exactly con- 

 stant in illuminating power from day to day, but during the hour and 

 a half or two hours the tests lasted each day the light did not vary 

 appreciably, according to the most uniform standards. In order to show 

 the uniformity of the standards under trial, the average of the day's tests 

 with each standard is taken and the individual tests compared with this- 

 mean. The divergencies of each standard from the mean of several tests 

 by the same standard are evidence of the want of uniformity of that 

 standard. 



In the following table the corrected result of each test is given; 118 

 complete tests by each of the four standards were taken. By the side of 

 each result is a number (obtained by dividing that result by the average 

 of the day's tests and multiplying by 100) which expi'esses the ratio of 

 that result to the average of the day by the same standard, the day's 

 average being called 100. A glance at the table shows the uniformity or 

 want of uniformity of the standards. Out of the 118 complete candle 

 tests 86 differed by 1 per cent, from the day's average, 57 difi"ered by 2 

 per cent., and 19 by 5 per cent. Variations of 9 and 10 per cent, 

 occurred occasionally. Of the other standai'ds the Amyl-acetate lamp 

 showed a variation of 2 per cent, from the daj^'s average on four occasions 

 out of the 118 tests, and a vai'iation of 1 per cent, on 11 occasions. The 

 Pentane lamp twice only showed a variation of 1 per cent., the Pentane 

 standard once only. 



In the second table the averages of each day are tabulated and the 

 general average of each standard is given. The Pentane standard and 

 the Pentane lamp gave practically the same light ; and the light was not 

 altered when pentane of specific gravity '632 and of specific gravity •628 

 was employed, instead of pentane of specific gravity -630. The Amyl- 

 acetate lamp was set too high, but although this was discovered soon 

 after the experiments were begun it was thought better to maintain the 

 flame at the same height than to alter it during the tests. The three 

 lamps tested gave uniform results. The bi'oader candles made by Mr. 

 Miller gave less light than those ordinarily used for gas-testing, and they 

 did not appear to burn more uniformly in this photometer. 



