Mr. H. F. Talbot's Experiments on Light. 333 



brightness, I think I have here suggested a method that is 

 founded upon sound principles. 



But here let me be permitted to leave for a moment the 

 immediate subject of this paper for the purpose of offering a 

 suggestion relating to the measurement of quantities in va- 

 rious other branches of natural philosophy. I think the same 

 method which I have here advanced as leading to a correct 

 numerical estimation of intense light, is also applicable to the 

 measurement of very intense heat. It is well known what dis- 

 crepancy of opinion exists as to the proportion of ordinary 

 temperatures to those which are very exalted (such, for in- 

 stance, as that of melted platina). Might we not measure this 

 heat with an ordinary thermometer, provided we so contrive 

 the experiment that its exposure to the heat shall be inter- 

 mittent, and shall only last for an extremely small portion of 

 the whole time ? To illustrate this idea, let us suppose a red- 

 hot cannon ball suspended near the edge of a rapidly re- 

 volving wheel. If a thermometer were fixed on the circum- 

 ference of the wheel, it would be only exposed to the influence 

 of the high temperature during a small part (let us suppose, 

 for example's sake, a hundredth part) of its revolution. If, 

 therefore, it was observed to mount five degrees, we might 

 fairly argue that it would have sustained an increase of tem- 

 perature of 500°, if it had remained stationary at the point 

 nearest to the ball. 



It is true, that in this form the experiment would be very 

 inaccurate, owing to the cooling power of the atmosphere on 

 the revolving thermometer and other causes, therefore I only 

 mention it for the sake of illustration ; or both the healed 

 body and the thermometer might be at rest, and a rotatory 

 screen might intervene between them, having an aperture of 

 given dimension, during the passage of which the thermo- 

 meter would be affected by the heat, but not at other times. 



If there is any truth in the preceding argument, as I trust 

 there is, it offers a method (and perhaps the only possible 

 one of subjecting to numerical comparison some qualities of 



into a surface uniform in light, which will have , — L X , that is. 



(7 20)* 1000 * 



about of the brightness of the sun ; that is, speaking accurately, 



not of the sun himself, but of his image reflected successively from both 

 mirrors when at rest. It is, however, to be observed, that the velocities 

 of the two mirrors must not be equal, nor in any simple ratio to one an- 

 other, otherwise their effect would be different from what is here contem- 

 plated. 



