386 M. MELLONI ON THE REFLECTION OF RADIANT HEAT. 



respect to glass, but also with respect to rock crystal, alum, fluate of 

 lime, topaz, sulphate of barytes, &c., so that a thin plate of any of these 

 substances, if very pure and well polished, will, when placed behind a 

 thick plate of the same substance, always transmit 0*923 and lose 0'077. 



The same numbers will be found also when the thin plate is placed 

 behind a thick plate of a different substance, provided the latter be less 

 permeable to the direct rays of the source. Thus a thin plate of rock 

 crystal transmits 0'923 of the radiation from thick glass ; and a thin plate 

 of glass transmits the same proportion of the heat emerging from water 

 or alum. The heat thus transmitted is so purified, that although it issues 

 from a very thin layer, it is still capable of traversing considerable depths 

 of glass or rock crystal without suffering any absorption. Hence it is 

 that plates of glass or rock crystal measuring 7 or 8 millimetres in thick- 

 ness will, when exposed to the rays emerging from a layer of water or 

 alum measuring 1 or 2 millimetres in thickness, transmit 0*923, as is done 

 by plates only half a millimetre thick. 



Concluding from all this that radiant heat undergoes a reflection of 

 about four hundredths of the incident heat which falls perpendicularly on 

 the surface of a diathermanous substance, we perceive at once the method 

 that is to be pursued in order to determine the quantity of calorific rays 

 reflected by athermanous bodies. We first observe the effect of the 

 calorific transmission through a plate of rock salt when the radiation, 

 emitted by a constant source, is perpendicular to its faces : the plate is 

 afterwards inclined towards the incident rays. In the quantity of heat 

 transmitted, there appears no sensible diminution so long as the inclina- 

 tion does not exceed 25° or 30° around the normal. The reflection of the 

 perpendicular rays is then sensibly equal to that of the rays forming an 

 angle of between 60° and 65° with the reflecting plane. Now, this be- 

 ing supposed, let us bring a bundle of radiant heat to fall on the well- 

 polished surface of glass or rock crystal at an incidence of between 60° 

 and 65°, and receive the reflected portion in the interior of the tube 

 wliich envelopes the thermomultiplier. After having marked the calo- 

 rific force indicated by the galvanometer, let us repeat the same expe- 

 riment on the polished surface of the athermanous body without making 

 any change in the respective positions of the several parts of the appa- 

 ratus. 'We shall thus obtain a second calorific force, differing from the 

 first. The reflection of the athermanous body will evidently be 0*0393, 

 which represents the value of the reflection at the surface of the rock 

 crystal multiplied by the ratio of the observed forces. 



The following are the mean results of several comparisons made be- 

 tween the quantities of heat reflected by rock crystal and yellow copper. 



Reflection of Reflection of Ratios of die Product of the two numbers 

 Rock Ci-ystal. Y fallow Copper, two reflections. (00393 and ] 1'3). 



3*15 35*63 11*3 0*4441, 



