CHEMISTRY. 1 7i) 



his own determinations, were represented in the form of curves, 

 and by means of the magnesium lantern projected on to a screen 

 for inspection. From the results of observations thus far made, 

 he concluded that Bunsen's and Roscoe's formula (a straight line 

 cutting the axis <?) would represent the greatest number of series, 

 and particularly the series belonging to the argand burners. 



That for those series belonging to the jet, the fish-tail and bat's- 

 wing burners, the common formula (a straight line passing 

 through the origin), which is a modification of Bunsen's, in which 

 B zzz O, very closely represents the relation found by experiment. 

 The number of cases in which the series or any considerable part 

 of the series could be represented by formula 3 (a parabola) were 

 very few. Mr. Stirnpson found, however, that when a gas flame 

 was on the verge of the smoky condition, a tangent to the curve 

 would almost always pass through the origin, showing that 

 for a limited range of consumption at that point the light is 

 proportional to the consumption. One other point was also very 

 apparent; such is the influence of the burner upon the flame, that, 

 in order to get the best result for any given consumption, the 

 burner must be of the kind best fitted for that particular con- 

 sumption. Amer. Gas-Light Jour, from Proc. Amer. Association. 



Farmer's Theorem. Many of Professor Silliman's experiments, 

 on which he relied to establish "Farmer's Theorem, 1 ' were per- 

 formed by mixing a very rich gas, which could not be burned in 

 a 15-hole argand-burner at the rate of 5 feet an hour, with a poor 

 gas of known power. Mr. Stimpson objects to this " method of 

 mixtures." Processor Silliman subsequently performed a limited 

 number of experiments by mixing a gas of known power with 

 hydrogen, and determining the candle-power of the mixture. He 

 finds the results sufficiently in accordance with theory to justify him 

 in asserting the applicability of the following rule : To find the 

 candle-power of a gas having, for example, an intensity greater 

 than 20 candles, mix the rich gas of unknown power with a 

 poorer gas of known power in such proportions that the intensity 

 of the mixture shall not be greater than 20 candle-power when 

 consumed at the agreed rate of not over 5 cubic feet per hour. 

 Then, to compute the candle-power of the rich gas, subtract the 

 intensity (6) of the poor gas from the intensity (cZ) of the mix- 

 ture ; multiply the remainder by the volume (a) of the poor gas ; 

 divide the product by the volume (c) of the rich gas ; add to the 

 quotient the intensity (d) of the mixed gas, and the sum is the in- 

 tensity (x) of the rich gas sought. That is, since 



ab -4~ ex (d 6) a 



= a, then x \- d. 



a-\- c c 



Amer. Jour. Science. 



Influence exerted on the Illuminating Power of Gas by the Pres- 

 ence of Carbonic Acid. That the presence of carbonic acid in 

 illuminating gas should exert some influence on its illuminating 

 power, it would be very natural to suppose. Owing to the (until 

 recently) general use of lime for the purification of gas, no careful 

 and direct experiments on the subject have been published. The use 

 of oxide of iron in the place of lime, and the consequent retaining by 



