782 PHILOSOPHICAL TRANSACTIONS. [ANNO 1800. 



Let us now examine what information we may draw from the facts which are 

 recorded in this table. The first that must occur is, that a candle which emits 

 light, is also a copious source of invisible heat. If this should seem to require a 

 proof I give it as follows. That the candle emits heat along with light, the ther- 

 mometer has ascertained ; and, that a considerable share of this at least must be 

 invisible, follows from comparing together the quantity of light and heat which 

 , are stopped by different glasses. The bluish-white one, for instance, stops 86 

 rays of light, and 625 of heat. Hence, if only visible rays of heat came from 

 the candle, a glass stopping more light, as for instance the dark-red glass, which 

 stops 999.8, ought to stop all heat whatever ; but the fact is, that it even stops 

 100 rays less than the former. This instance alone shows plainly, that the exist- 

 ence of invisible terrestrial heat in the flame of a candle, is proved; while, on the 

 contrary, heat derived from rays that are visible, remains yet to be established, by 

 those who would maintain that there are any such. But, for the sake of argu- 

 ment, let us endeavour to explain how visible rays of heat may be reconciled with 

 the contents of our 6th table. " Now though we must allow," it may be said, 

 " that there is a certain quantity of candle-heat which cannot be seen, we are 

 however at liberty to assign any ratio that this may bear to its visible heat-rays. 

 Let us therefore begin with the bluish-white glass, and make the most favourable 

 supposition we can, in order to explain its phenomena. Visible or invisible, it 

 stops 625 rays of heat, and also 86 of light. Now, as in the last column of the 

 table we have likewise the proportional quantity of invisible heat it intercepts, 

 which is 700 out of 1000, we may surmise that the 914 rays of light, together 

 with the 300 of the invisible rays which are transmitted, make up the 375 rays 

 of heat which pass through the glass. . Hence, by algebra, we have the number 

 of invisible heat-rays 878, and the number of the visible ones 122. Then, to 

 try how this will answer, if 1000 rays of light give 122 of heat, 80 will give 10; 

 and if out of 1000 invisible rays 700 be stopped, 878 will give 6l5 to be inter- 

 cepted. The sum of these will be 625, which is exactly the number pointed out 

 by our table." Now this being a fair solution of one instance, let us see how it 

 will agree with some others. 



Before proceeding however, I cannot help remarking, that the supporters of 

 visible heat-rays must feel themselves already considerably confined, as our present 

 argument will not allow them more than 122 of such rays out of 1000. Now, 

 if the assumption that terrestrial heat is owing to a mixture of visible and invisible 

 rays, in the proportion of 122 of the former to 878 of the latter, be well-founded, 

 it ought to explain every other phenomenon collected in our table. The purple- 

 coloured glass stops 993 rays of light, which, according to our present hypothesis, 

 should stop 121 rays of heat: it also stops 730 invisible rays, which will give 

 641 rays of intercepted heat; therefore this glass should stop 762 rays of heat, 

 out of every 1 000 that come from a candle ; but from our table we find that it 

 stops no more than 520, so that 242 rays cannot be accounted for. The glass 



