on the terrestrial Rays that occasion Heat. 439 



refrangibility of the rays of heat follows of course. So far then, 

 a great resemblance again takes place. 



I must now point out a very material difference, which is, that 

 the rays of heat are of a much more extensive refrangibility 

 than those of light. In order to make this appear, I shall de- 

 lineate a spectrum of light, by assuming a line of a certain 

 length ; and, dividing it into seven parts, according to the di- 

 mensions assigned to the seven colours by Sir Isaac Newton, 

 in the fourth figure of the second part of his Qptics, I shall 

 represent the illuminating power of which each colour is pos- 

 sessed, by an ordinate drawn to that line. And here, as the 

 absolute length of the ordi nates is arbitrary, provided they 

 be proportional to each other, I shall assume the length of that 

 which is to express the maximum, equal to |~| of the whole line. 



Thus, let GQ * represent the line that contains the arrange- 

 ment of the colours, from the red to the violet. Then, erecting 

 on the confines of the yellow and green the line LR = y-J- of 

 GQ, it will represent the power of illumination of the rays in 

 that place. For, by experiments already delivered, we have 

 shewn that the maximum of illumination is in the brightest 

 yellow or palest green rays.-f From the same experiments we 

 collect, that the illuminations of yellow and green are equal to 

 each other, and not much inferior to the maximum ; this gives 

 us the ordinates K and M. Then, by the rest of the same ex- 

 periments, we obtain also the ordinates H, I, N, O, P, with 

 sufficient accuracy for the purpose here intended. All these 

 being applied to the middle of the spaces which belong to their 

 respective colours, we have the figure GRQG, representing 

 what may be called the spectrum of illumination. 



* See Plate XX. f See page 262. 



3L2 



