VOL. VIII.] PHILOSOPHICAL TRANSACTIONS. 87 



see the two colours parted from one another in the fashion of two images of the 

 paper, as they are represented at € and y, where suppose p the scarlet, and y the 

 blue, without green or any other colour between them. 



Now from the aforesaid position I deduce these two conclusions. 1 . That if 

 there were found out a way to compound white of two simple colours only, that 

 white would be again resolvable into no more than two. 2. That if other whites, 

 as that of the sun's light, &c. be resolvable into more than two simple colours, 

 as I find by experiment that they are, then they must be compounded of more 

 than two. 



To make this plainer, suppose that A represents a white body, illuminated by 

 a direct beam of the sun, transmitted through a small hole into a dark room, 

 and a. such another body illuminated by a mixture of two simple colours, which 

 if possible may make it also appear of a white colour, exactly like A. Then at a 

 convenient distance view these two whites through a prism, and A will be 

 changed into a series of all colours, red, yellow, green, blue, purple, with their 

 intermediate degrees succeeding in order from B to C. But «, according to the 

 aforesaid experiment, will only yield those two colours of which it was com- 

 pounded, and those not conterminate like the colours at BC, but separate from 

 one another as at i3 and y, by means of the different refrangibility of the rays to 

 which they belong. And thus by comparing these two whites, they would ap- 

 pear to be of a different constitution, and A to consist of more colours than «. 

 So that what Mons. N. contends for, would rather advance my theory by the 

 access of a new kind of white than conclude against it. But I see no hopes of 

 compounding such a white. 



As for Mons. N.'s expression, that I maintain my doctrine with some concern, 

 I confess it was a little ungrateful to me to meet with objections which had been 

 answered before, without having the least reason given me why those answers 

 were insufficient. The answers which I speak of, are in the Transactions from 

 p. 20 to p. 29 of this vol. And particularly in p. 22 ; to show that there are 

 other simple colours besides blue and yellow, I instance in a simple or homo- 

 geneal green, such as cannot be made by mixing blue and yellow, or any other 

 colours. And there also I show why, supposing that all colours might be pro- 

 duced out of two, yet it would not follow that those two are the only original 

 colours. The reasons I desire you would compare with what has been now said 

 of white. And so the necessity of all colours to produce white might have 

 appeared by the experiment p. 23, where I say, that if any colour at the lens be 

 intercepted, the whiteness which is compounded of them all will be changed 

 into (the result of) the other colours. 



However, since there seems to have happened some misunderstanding be- 



