VOL. LX.} PHILOSOPHICAL TRANSACTIONS. 71 



surface abcd; and the solid space contained between them, is very nearly equal 

 to a cylinder on a base equal to the spherical surface abcd, and of altitude equal 

 to ef. Therefore diminish the diameter of each particle of light, that is, diminish 

 ef, in any proportion whatever, and the solid space abcd X e/" diminishes in the 

 same proportion. And if the matter in that space were given, the density of it 

 would always be as —' But the matter of each particle of light, and consequently 

 the matter of the given number in the space abcd X ^ is always as e/"'. 

 Therefore the density of the substance formed by light in the space abcd X ef, 



is always as •^' or as ep. 



That is, the number of spherical particles of light, in the solid space abcd X ef, 

 being given, and the density of each particle, and the spherical surface abcd 

 being given, the square of the diameter of each particle will be as the density 

 of the substance they compose, on the surface of the sun. But it has been 

 found, that if the diameter of each spherule were one millionth of one millionth 

 of an inch, the greatest number of such spherules that the sun's surface can 

 contain at once, would, with the density which has been assigned to each separate 

 particle, form a substance upon it 12 times more dense than water. Therefore, 

 that the same number of such particles should form a substance about 173 times 

 less dense than water, the diameter of each must not exceed ^\- of one mil- 

 lionth of one millionth of an inch. But I have showed that the greatest 

 probable density of light on the sun's surface, does not amount to 173d part of 

 the density of common water. Therefore the diameter of each spherule does 

 not exceed VtVj or more accurately, Ci'i^'i of on^ millionth of one millionth of 

 an inch; and is probably still less.* More than 95 100 spherules of this size 

 go to make up one of the size first assumed. Therefore, though the sun 

 should lose as much matter as can be supposed to be contained in any single 

 emission of light, 95 1 00 times in every second, no sensible alterations in the 

 system could take place in millions of years. And perhaps light does not issue 

 from the sun so frequently as 95 100 times in a second. 



p. s. The late Dr. Pemberton, of Gresham College, to whom the foregoing 

 paper was communicated, about 12 months before his death, in a letter with 

 which he favoured the author, on the subject, remarked, that he had no material 

 objection to any part of the reasoning, except it was, that the particles of light 

 are too peremptorily spoken of as spherules. Their real figure is quite unknown; 

 but the probability is, that they are not spherical, since Sir Isaac Newton found 

 that their different sides have different properties. As the like objection may 

 occur to others, it may not be improper to add a few words to obviate it. 



• Instant. Product, p. 38—41. 



