On the Numher oftht Primitivt Colorific Rays in Solar Light. - 387 



If the folar light confided of feven primitive, homogen^al coloured rays, and that thefc 

 homogeneal rays were equally refrangible, the fpedrum would confift of feven circles of 

 different colours, fince the homogeneal rays of each colour would paint a circular image of 

 the fun. But it is manifeft, that feven circles could nbt compofe an oblong fpe£lrum, 

 vrith rediiineal fides. Therefore the rays of the fame denomination of colour muft be 

 diiFerently refrangible. Which is alfo made ftill farther evident by obfervation of the fpec- 

 trum, fince in it we perceive, that the prifmatic colours are difTufed over fpaces, which 

 are, on the fides, terminated by right lines, and therefore the centers of the circles of the 

 fame denomination of colour are difFufed over lines equal to thefe fegments of the re£lili- 

 neal fides of the fpe£lrum. Newton has (hewn, prop. 4. B. i. Optics, how to feparate 

 from one another the heterogeneous rays of compound light, by diminifliing the breadth of 

 the fpeftrum, its length remaining unchanged ; and when the length of the fpeftrum is to 

 its breadth, as 72 to i, the light of the image is feventy-one times lefs compound than the 

 fun's dire£t light. In the middle of a black paper he made a round hole, about a fifth or a 

 Cxth part of an inch in diameter, upon which he caufed this fpe£trum fo to fall, that fome 

 part of the light might pafs through the whole of the paper ; this tranfmitted part of the 

 light he refrafted with a prifm placed behind the paper, and letting the light fall perpen- 

 dicularly upon a white paper, he found that the fpedirum formed by it was perfeftly cir- 

 cular. Hence, therefore, it follows, that the equally refrangible rays occupy a fpace on 

 the re£lilineal fides of the fpe6lrum equal at lead to the fifth or Cxth part of an inch ; that 

 is, the rays of the fame colour are differently refrangible. 



The different quantity of the homogeneous rays of different colours will not account for 

 tlie different fpaces they occupy in the fpe£trum ; for this difference in quantity would 

 affcdt only the intenfity of the colour, not the magnitude of the fpace which it would oc- 

 cupy. All the red light therefore is not homogeneous ; but confifts of rays of innumera- 

 ble, different degrees of refrangibility •, and fo of the other colours. 



Now fince the rays which are of the fame denomination of colour are differently refran- 

 gible, they will either form oblong fpe£trums detached from each other ; or they will in 

 part lap over, and fall on each other. The former pofition is manifeftly falfe : therefore 

 the original prifmatic colours will partly lap over and fall on each other, and therefore ne- 

 cefTarily generate the intermediate colours. And fo Sir I. Newton obferves, where he faysj 

 that the original, prifmatic colours will not be diflurbed by the intermixture of the con- 

 terminous rays, which are intermixed together. This overlapping however, which Newton 

 fpeaks of, arifes only from the fun's having a fenfible diameter, and does not neceflarily 

 imply an equal refrangibility in any differently coloured rays. If there be but three original 

 prifmatic colours, red, yellow, and blue, and that the red and yellow lap over, fo as that 

 there fliall be a certain fpace in the fides of the fpectrum equally occupied by yellow and 

 red circles, then will thefe circles by their intermixture compound an orange colour ; and 

 this colour as to refrangibility will be homogeneous, becaufe the coincident rays of dif- 

 ferent colours are equally refrangible. In like manner green may be compounded by the 



3 D z- mixture 



