I 334 1 



Vll. Co7nmzmication from Dr. BlACKBURNE refpeSilng 

 CalqriC) Lighty ajtd Colours. 



I 



SIR, Spring Garden, May 15, iSooi 



.T is now more than twelve months fince I announced 

 my intention of piiblifbing fome remarks concerning the ufe 

 of the word heat, and the conftitution of light. My avoca- 

 tions have not allowed me to fulfil my promife ; and being 

 flill uncertain when I may be able to finifli for publication, 

 I fhall be much obliged to you to infert in your refpeftable 

 Journal the following brief refults of my inveftigation con- 

 cerning light, which I conceive to be of more importance 

 than thofc concerning heat, which chiefly relate to verbal 

 criticifm. The proofs and reafoning on which thefe refults 

 depend, will, I hope, be foon fubmitted to the public. 



*' I am, &c. 

 «' Mr. Tilloch. WILLIAM BLACKBURNE." 



ift. Light is a compound refulting from a peculiar com- 

 bination of caloric and oxygen. 



adly, Li all thofe phenomena which have given occafion 

 to the idea that light is identical with, or a modification of, 

 caloric, the manifedation of the latter principle is to be re- 

 ferred to the difunion of the conftituting principles of light. 

 The caloric, therefore, which fo frequently refults from the 

 application of light to various fubftances, iffues from the de- 

 corapofition of light in various degrees.. 



3dly, The phenomena of colours are to be afcribed to the 

 different qualities of light, as containing caloric and oxygen 

 in different proportions. The different proportion? manifeft 

 themfelves in the circumftances both of the decompofition 

 and the formation of varioufly coloured light. 



4thly, The feparation of light by the prifm is to be regarded 

 as a chemical decompofition, not a phyfical or mechanical 

 divifion of light, 



5thly, The changes which take place in the colours of 

 different fubftances, as of plants in the procefs of vegeta- 

 tion, of metals in that of oxydation, are owing to corre- 



fpondent 



