Solar Light on Combustion. 185 



2. Exposure of the flame to the influence of intense solar 

 light, without heating the swrounding air. 



The first condition was secured by performing all the ex- 

 periments in a large lecture-room, with all the doors and win- 

 dows closed. To secure the second condition, I employed a 

 portion of the apparatus belonging to a large solar microscope, 

 consisting of the reflecting mirror, the condensing lens and tube, 

 together with the mechanical arrangements for adjusting the 

 direction of the light. As the condensing lens was upwards of 

 4 inches in diameter, I hoped to exaggerate enormously what- 

 ever effect the light might exert, by concentrating it on a com- 

 paratively small area. Inasmuch as the aperture in the window 

 through which the light was admitted was completely closed by 

 this arrangement, the exterior agitations of the atmosphere were 

 not felt in the room, while the pencil of light thus thrown on 

 the flame traversed it, as well as the surrounding air, without 

 imparting a sensible amount of heat to the latter. 



I used the best wax-candles (as they are called in the shops), 

 four to the pound, costing about 15 cents apiece*. By allowing 

 them to burn a sufficient length of time to form a well-defined 

 cup for the melted wax, and carefully turning the wicks so as 

 to render them self-snuffing, the combustion was found to go on 

 with remarkable uniformity in a calm atmosphere. The rate of 

 burning was determined in the following manner : — A portion of 

 candle, three or four inches in length, was secured to the bottom 

 of one of the scale-pans of a tall balance and ignited ; after 

 allowing it to burn for ten or fifteen minutes, so as to secure a 

 steady flame of constant size, it was nearly balanced by adding 

 weights to the opposite scale-pan, allowing a slight preponder- 

 ance to the candle-pan. In a short time the equilibrium was 

 established by the burning of the candle : the precise time at 

 which the balance indicated a condition of equilibrium was accu- 

 rately noted. Next, a given weight (say 60 or 100 grains) was 

 withdrawn from the weight-pan, and the time of restoring the 

 equilibrium by the loss of weight in the burning candle was in 

 like manner recorded. In this manner the rate of combustion 

 was determined by observing the time occupied in consuming a 

 given weight of the burning matter. The arrangements described 

 above enabled me to perform such experiments alternately in 



* From the close approximation to identity in the rate of consumption, 

 it is probable that these are the same as Dr. Ure's " genuine wax-candles." 

 He found the consumption to l)e, "upon an average of many experiments, 

 125 grains jjerhour." (Diet, of Arts, Manufactures, and Mines, 4th edition, 

 article " Illumination, Cost of.") My experiments gave resi)ectively 136"7, 

 lli.54, and \'2A7 grains per hour. Other kinds of candles burn at a much 

 more rapid rate. 



