LETTER OF J. HENRY TO REV. S. B. DOD. 157 



to shade the pile by a screen with a small opening introduced 

 between it and the flame. 



XII. Another series of experiments connected with the preced- 

 ing may be mentioned here. It is well known that the light from 

 a flame of hydrogen is of very feeble intensity ; the same is the case 

 with that of the compound blowpipe, while the temperature of the 

 latter is exceedingly high, sufficiently so to melt fine platinum wire. 

 It is also well known that by introducing lime or other solid sub- 

 stance into this flame its radiant light is very much increased. I 

 found that the radiant heat was increased in a similar ratio, or in 

 other words, that in such cases the radiant heat was commensurate 

 with the radiant light, and that the flame of the compound blow- 

 pipe, though of exceedingly high temperature, is a comparatively 

 cool substance in regard to radiant heat. To study the relation of 

 the temperature of a flame to the amount of heat given oif, four 

 ounces of water were placed in a platinum crucible and supported 

 on a ring stand over a flame of hydrogen ; the minutes and seconds 

 of time were then accurately noted which were required for the rais- 

 ing of the water from the temperature of 60 to the boiling point. 

 The same experiment was repeated with an equal quantity of water, 

 with the same flame, into which a piece of mica was inserted by a 

 handle made of a narrow slip of the same substance. With this 

 arrangement the light of the flame was much increased, while the 

 time of bringing the water to the boiling point was also commensu- 

 rately increased, thus conclusively showing that the increase of light 

 was at the expense of the diminution of the temperature. These 

 experiments were instituted in order to examine the nature of the 

 fact mentioned by Count RUMFORD, that balls of clay introduced 

 into a fire under some conditions increase the heat given oif into 

 an apartment. From the results just mentioned it follows that the 

 increase in the radiant heat, which would facilitate the roasting of 

 an article before the fire, would be at the expense of the boiling of 

 a liquid in a vessel suspended directly over the point of combustion. 



XIII. Another investigation had its origin in the accidental 

 observation of the following fact : A quantity of mercury had been 

 left undisturbed in a shallow saucer, with one end of a piece of 

 lead wire, about the diameter of a goose-quill, and six inches long, 



