PHYSICS. 287 



sealed in a globe, which is then exhausted to oiie-niillioiith of an atinos- 

 l)here. Subsequently, in exi)eri)nenting with various libers, iMlison (;ame 

 to the oonclusion that a natural liber was more durable tlian an artihcial 

 one, and that Japanese bamboo furnished the best. Tlic lamps which 

 are now made on the large scale are constructed with these libers. 



liowland and Barker have made a series of measurements to test the 

 efliciency of Edison's lamp. They compared the loss of energy in the 

 lamp, as measured in the calorimeter, with the amount of light emitted, 

 measured on the photometer. The lamps used were live in number, 

 four of them having carbons made from paper, and one a natural liber 

 carbon. The comi)arisons were made by having two lamps in circuit, 

 one being in the calorimeter and the other on the photometer in the lirst 

 experiment, and their positions being reversed in the second. No. 1^01, 

 which gave a mean of 10.1 candles on the pliotometer, gave 318G foot- 

 pounds of energy in the calorimeter ; No. 580, which gave 13.1 mean 

 candle lights, evolved 3540 foot-pounds. But when No. liOl was made 

 to give 28.9 candles, its energy was only increased to 4898 foot-pounds ; 

 and No. 580, which gave 33.5 candles, evolved only 5181 foot-pounds; 

 the cost of doubling the light being- one-half more energy only. ]No. 

 809 j,ave 14.3 candles, and used 3330 foot-pounds ; No. 850 gave O.li 

 candles, and consumed 2483 foot-pounds. No. 817, the liber lamp, gave 

 17.2 candles for a consumption of 2708 foot-pounds of energy. The 

 mean of the lirst two is 109 candles i)er horse power of current ; of the 

 second two, 2*;4.3 candles; of the third two, 133.4; and of the fourth, 

 209.G. Or, in the lirst, G.8 gas jets of IG candles, in the second J2.8, in 

 the third 8.3, and in the fourth 13.1 gas jets would be yielded by each 

 horse-power of current. From the efficiency of Edison's machine, the 

 authors thnik that the yield per indicated horse-power of the engine 

 would not be lar from 70 per cent, of these values. — {Am. J. Sci., Ill, 

 xix, 337, April, 1880.) 



Brackett and Young have also made measurements on Edison's lamp, 

 in addition to others upon his Faradic machine. Comparison of the 

 Edison dynamometer with the Prony brake showed that the latter reg- 

 istered 93.2 per cent, of the power transmitted by the former. In their 

 lirst experiment, the power exi)ended was measured by the dynamometer 

 and the current by the copper voltameter. The total efficiency was 

 found to be 82.3 per cent., and the available efficiency 78.7 per cent. In 

 the second trial the energy produced was measured by the calorimeter, 

 and the total efficiency was found to be 84. G per cent., the available, 78.2 

 per cent. A third test gave 84.5 total, and 78.2 available efficiency. But 

 one lamp was tested. No. 853, the carbon of which was from paper. 

 The mean illumination was 10.1 candles ; the resistance, hot, 99.G ohms; 

 the difference of potential, 74.33 volts; hence the current strength was 

 0.75 weber, and the lamp was consuming 0.075 horse-power. This gives 

 ]37 candles per horse-power of current, or 107 candles per horse-power 

 at the dynamometer; a near correspondence with the preceding re- 

 sults.— (A?«. J. >S'd., Ill, xix, 475, June, 1880.) 



