CHEMICAL SCIENCE. 237 



then, to get the same light from the same burner, you must supply 600 cubic 

 inches per hour. He adds, that with coal-gas the deterioration appears to be 

 more rapid. For if such gas, when first made, will give the light of 1 candle 

 by the consumption of 400 cubic inches per hour, when kept four days it will 

 require the consumption of 460 cubic inches per hour to give the same light. 

 On my first attempt to obtain some definite results, I filled a large receiver 

 from the street main, and placed it on the shelf of the pneumatic trough; 

 the next evening I filled a second one, and put it alongside of the first; the 

 following evening I filled a third receiver, and still the following evening, 

 the llth inst., I filled a fourth receiver. On the evening of the 12th I was 

 thus provided with four jars of gas, one of which had been standing twenty- 

 four hours, or one day, over the pneumatic trough; this I will call No. 1. 

 Another, No. 2, had been standing two days; No. 3 had been standing three 

 days, and No. 4 had been four days in contact with the water. The diminu- 

 tion in volume by such exposure was indicated by a receiver graduated to 

 cubic inches, into which I introduced 130 cubic inches of gas on the evening 

 of the 8th; on the evening of the 12th this had lost about ten and one-half 

 cubic inches, indicating a loss of about eight per cent, of the original bulk. 



The effect produced on the illuminating power of the gas by the loss of 

 volume became at once apparent as I proceeded to contrast the value of the 

 flames furnished by the contents of the several receivers, 1, 2, 3, and 4. 

 I used for this purpose the ordinary photometer arrangement, taking the 

 relative intensity of the shadows produced, as a measure of the relative 

 intensity of the light. The candle employed for the comparison was the 

 patent candle already referred to, and the burner was the kind known as 

 fish-tail burner, which had been previously gauged, and known to consume 

 a trifle more than five cubic feet per hour, with the average maximum pres- 

 sure of the gas-works. I need hardly add, that the burner was the same in 

 all the trials, and occupied exactly the same position. The burner and the 

 screen on which the shadows fell were not moved at all during the experi- 

 ments. The only adjustment wanted was to bring the candle nearer to or 

 farther from the screen ; and by beginning with the most luminous gas, the 

 adjustment became simply a gradual withdrawal of the candle. 



The capped receiver from which the gas was passed, floated freely in a 

 large glass jar, supported in an erect position by the perpendicular sides of 

 the jar, its own weight, with all attachments, making a difference of level 

 between the water around it and that within, equal to three and one-half 

 inches, a little exceeding the ordinary evening pressure in the gas-pipes. 

 This ditference of level, and consequently the pressure on the escaping gas, 

 was kept uniform by the spontaneous sinking of the receiver as the gas was 

 consumed, a flexible tube communicating between the stop of the receiver 

 and the gas-burner. This arrangement gave me a steady, equable flame, 

 which continued perfectly uniform long enough to enable me, after a few 

 trials, to note very exactly its true value. The results as first obtained were 

 too startling to be at once believed, but subsequent repeated trials satisfied 

 me that they were very close approximations to the truth. The first trial 

 was with the gas from the street main, which I found equal to 10'71 candles. 

 The same gas, transferred from the pipe to the capped receiver, and burned 

 immediately, gave exactly the same power, 10 - 7l candles. Gas No. 1 was 

 next used, and found equal to only 3 '50 candles; gas No. 2, after standing 

 two days, gave the light of 3'20 candles; gas No. 3, three days old, was 



