r let ]

portion as appears mofl likely to fuit the fubjecl to be
treated. A third row ftill nearer to the fubjecl:, may
eafily be formed from the former.

To be fomewhat more particular : We fuppofe the
original colours placed in their order on the upper part of
the pallette : To compofe a fecond row of colours ; take
for example, lake and Naples yellow, to thefe we ima-
gine white to be added ; this composition we place in
the fecond row, and fo of the reft. If we wifh for a
lighter teint of the fame colours, this is produced by an
addition of white ; or perhaps we want a red fomewhat
different, varying to light red, or to vermilion ; thefe unit-
ed, form a compound, admitted lower on the pallette: or,
this fecondary colour, inftead of red, requires yellow, or
a different kind of yellow ; or, in fhort, any other colour
which may be a departure from the firff mixture. This
combination is termed breaking the colours : its ufe is,
to prevent glare and rawnefs, and its judicious exercife
is a very principal token of a mafterly hand.

We may here obferve, that nature never prefents a
pure, unmixed, ftrong colour of any caff : but every
colour has an inclination to, or a tinge of fome other.
A blue fky feems at firft an exception, yet is not fo in
fact; nor the colour of the fea, nor the green of a
field ; for an innumerable variety of lights, reflexes, &c.
are perpetually playing upon it, and thereby diversifying,
debafing, and confufing the original colour. Even dra-
peries vary in the colours of their lights and fhades: the
yellow light of the fun, the red light of a fire, a white
light, &c. alter their appearances : and the deprivation
of light in the fhades, as the parts are more or lefs
4 retired^