10 
concentrates with yellow, red, or blue, or a mixture depending upon 
a natural and regular phaenomenon : (viz.) That as we have seen in 
a, b, c, that blue is on the inner side of the dark circle, if there is 
much proportional breadth, so commonly a white or light space suc- 
ceeds, passing into yellow, orange, and red ; and sometimes the ar- 
rano-ements in succession approach nearer and nearer, and the white 
is obliterated, and yellow is facing or nearer the light blue : again, 
the yellow and blue pass into each other in narrow or nearer circles, 
and produce green very often at the expense of the marginal tints, 
if I may so call the blue on one hand, and the red on the otlier. 
Something like this, or duller and deeper, appears at fig. g, and the 
semicircles surrounding show it more distinctly, although placed 
here for a no less curious example. 
I have perhaps enlarged too much on this subject here. I, how- 
ever, thought it essential not only to begin with these general pre- 
mises, but to show some common varieties in the arrangement of 
prismatic colours, by which we ma^^ perceive that there is a perfect 
rule for proportion in it, however unequal it may at first appear, 
which seems to me to have been hitherto a desideratum ; and as I 
expect to prove much from the nature of this particular appearance, 
I have dwelt the longer upon it ; which is the more necessary, seeing 
how orderly sir Isaac Newton produced certain colours by the pres- 
sure of two glasses, and how naturally and truly he enumerated 
them without quite attaining the leading and true cause or principle 
of that arrangement, which he was so near discovering, and which 
would imdoubtedly have been of the most essential consequence in 
such good hands. The whole agrees so well with what is said of the 
experiments of that great man, that, although it is well known, 
I cannot forbear relating it here. Sir Isaac Newton himself had 
observed, that as he was compressing two prisms hard together, in 
order to make their sides (which happened to be a little convex) to 
touch one ano'.her, in the place of contact they were both perfectly 
tran' parent, as if they had been but one continued piece of glass. 
Koimd the point of contact, where the glasses were a little separated 
