[241 ] 
XLI. On Colours.—In Answer to Mr. T. Harcreaves’s Stric- 
tures on the Work entitled ** Chromatics; or, An Essay on the 
Analogy and Harmony of Colours.’ By Tur Autuor. 
To Mr. Tilloch. 
Sir, — i answer to the observations of Mr. T. Hargreaves on 
my Essay entitled ‘‘ Chromatics,” &c. in your last Number, I beg 
to state that the pigments chosen to illustrate the various deno- 
minations of colours therein, have been selected from the most 
eminent for durability and beauty, and that J am not acquainted 
with a any blue, red, or yellow, superior in these respects, to the 
three pig ments, albveminings rubiate or madder red, and Indian 
yellow, used in exemplification of the primary colours. 
An eye critically nice will discern in every colour a tendency 
to some other colour, according as it is influenced by light, shade, 
depth or diluteness ; nor is “this the case only in the inherent 
colours of pigments, &e. but it is so also-in the transient colours 
of the prism, &c. Hence blue in its depth inclines to purple; 
deep-yellow to orange, &c.; nor is it practicable to realize these 
colours to the satisfaction of the critical eye,—since perfect co- 
lours, like perfect geometrical figures, are pure ideals. My ex- 
amples of colours are therefore quite as adequate to their office 
of illustrating and distinguishing, as the figure of an angle in- 
clining to the acute or obtuse, instead of a perfect right angle, 
or middle form, would be in illustrating the conception of an 
angle in general. 
Mr. H.’s objections to the examples of secondary and tertiary 
colours rest upon similar ground. ‘Thus purple, composed of 
blue and red, (which in its perfect hue should neutralize or ex- 
tinguish a perfect yellow,) denotes, in the example referred to, 
not any particular or individual tint, but a class of tints, bounded 
on the one extreme by blue, and on the other by red: and thus 
also of the other colours. The secondaries, purple, green, and 
orange, have accordingly been exemplified by intermediate tints 
composed of two of the primaries alternately; and the ferdiaries, 
russet, citrine, and olive, by like iabeemediatee of these secon- 
rs ; for all these Menianniinatinnt of colours, as above instanced, 
are indications of classes or genera, and not significant of in- 
variable hues or tints of colour. 
The remarks of Mr. H. are however perfectly just with re- 
_ spect to Example X. of the Essay, in which the neutralizing co- 
Jours are contrasted, and consequently require such individuality 
of the opposed tints as may render them reciprocally neutra- 
lizing. 
The foregoing remarks upon the particular relations of colours 
Vol. 50. No.234. Oct. 1817. Q apply 
