io lieat and Illumtndte OljeHs: 32J 



t am well aware that the refults I have drawn from th*; 

 foregoing experiments, both with regard to the heating and 

 illuminating powers of differently coloured rays, muft be af- 

 fe6led by fome little inaccuracies. The prifm, under the 

 circumftances in which I have lifed it, could not effccl a 

 complete reparation of the coloursj on account of the ap- 

 pat-ent diameter of the fun, and the confiderable breadth 

 of the prifm itfclf, through which the rays were tranf- 

 niitted. 



Perhaps an arrangement like that in Fig, 16. of the New- 

 tonian experiments mioht be employed, if inftruments of 

 fufficient fenfibility, fuch as air thermometers, can be pro- 

 cured, that may be affcfted by the enfeebled illumination of 

 rays that have undergone four tranfmiffions and eight refrac- 

 tions ; and efpecially when their incipient quanlity has beeii 

 fo greatly reduced in their limited paflage through a fmall 

 hole at the firll incidence. 



But it appeared moft expedient for nie, at prefent, tci neg- 

 left all further refinements, which may be attempted here- 

 after at leifure. It may even be prefumed that, had there 

 not been fome fmall admixture of the red rays in the othet 

 colours, the refult would have been ftill more decifive with 

 regard to the power of heating vefted in the red rays. And 

 it is likewife evident, that at leaft the red light of the prif- 

 matic fpeftrum was much lefs adulterated than any of the; 

 other colours ; their refractions tending all to throw theni 

 from the red. That the fame rays which occafion the 

 greateft heat, have not the power of illumination in any 

 Ilrong degree, ftands on as good a foundation. For, fince here 

 alfo they have undergone the faireft trial, as being moft free 

 from other colours, it is equally proved that they illuminate 

 obje6ls but imperfcftly. There is fome probability that a 

 ray, purified in the Newtonian manner above quoted, efpe- 

 cially in a well darkened room, may remain bright enough 

 to ferve the purpofe of microfcnpic illumination, in which 

 cafe more precifion can eafily be obtained; 



The greateft caufefor a mixture of colours, howevef, which 

 is the breadth of the prifm, I faw might eafily be removed ; 

 therefore, on account of the coloured points, which have 



Vol. VII. Tt bceM 



