VOL. XC.] PHILOSOPHICAL TRANSACTIONS. 779 



given instances where the assumed hypothesis of visible and invisible heat, in 

 certain proportions, would require a greater stoppage than our experiments will 

 admit ; and now, on the contrary, it appears, that interceptions calculated accord- 

 ing to the same hypothesis, should be less than the results in the 3d table give 

 them. From which we conclude, that every other proportion fixed on, would 

 always be erroneous, either in excess or in defect. 



Equal contradictions maybe shown to attend all endeavours to account fortheresults 

 contained in our4th table, by admitting any visible heat at all, let the quantity be what 

 it will. To make the proof of this general, let 1000 be the total heat, and assume x 

 for that part of it which we would suppose to be occasioned by visible rays ; then will 

 1000 — oc be the remainder, which must be ascribed to rays that cannot be seen. 

 Now by our table we find that crown glass, of which one side has been rubbed on 

 emery, stops 854 rays of light. These alone, if not a particle of invisible heat were 

 stopped, would be equal to .85437 visible rays of heat, that must be intercepted by the 

 glass. When the other side of this glass has also been rubbed on emery, it will 

 stop 932 rays of light, which will give .932#, for the quantity of heat to be 

 intercepted, on the same supposition, that all invisible rays of heat are transmitted. 

 But, by our 4th table, we have the actual stoppage of heat of these glasses ; 

 which will therefore give us the two following equations ; .854 x = 464, and .932 



x = 667. Then, taking the first from the last, and reducing the remaining equa- 



203 20s 

 tion, we obtain x = — — -, for the visible part of the total heat. But , or 2602, 



•U/o •0/8 



being only a part, comes out greater than 1000, which is the whole ; and this be- 

 ing absurd, it follows that visible rays of heat cannot be admitted, in any 

 proportion whatever. This will equally hold good with any additional stoppage 

 of invisible heat, provided it be equal in both glasses ; and of this equality, the 

 l65th and 167 th experiments can leave us no room to doubt. 



But it is high time that we should now take into consideration a more direct proof, 

 which may be drawn from our prismatic experiments. The results of them are 

 here brought into a table, as follows. 



Table 5. — Stoppage of Prismatic Heat of the Refrangibility of the Red Rays, and of the Invisible Rays. 

 Red rays. Invisible rays. Red rays. Invisible rays. 



Bluish-white glass stops 375 71 Pale-blue 700 750 



Flint glass 143 000 Dark-blue 71 167 



Crown glass 294 182 Indigo 367 222 



Coach glass 200 143 Pale-indigo 313 250 



Iceland crystal 200 Purple 444 273 



Calculable talc 133 250 Violet 400 250 



Dark-red glass 692 000 Crown glass, one side rough. . 389 600 



Orange 500 273 Coach glass, ditto 500 500 



Yellow 417 200 Crown glass, both sides rough 471 6*00 



Pale-green 588 375 Coach glass, ditto 833 714 



Dark-green 786 500 Calcined talc 737 889 



Bluish-green 462 800 



As a necessary introduction to the decisive experiment I am going to analyse, I 

 must remark, that it has been shown in a former Paper, that the prism separates 



5 G2 



