42 
to add calculations for the more exact proportions and distances, but 
to show what is the utmost extent of its utility. 
The Rev. W. Kirby informs me that Lamarck is said by Latreille 
to have given a chromatic scale of colours ; and from a sketch which he 
was so kind as to give me, just as I was about to send this to the ])ress, 
1 tind it was very conclusive as far as it went, and is truly ingeuiuu;'. 
He observes tiiat the colours of the rainbow or prism are nut in the 
natural order, but exhibit two reversed branches of that order joined 
together, white and black being suppressed, v\ hich he affirms are the 
extremities of the natural series of colours: by this reversion the 
yellow is laid next to the blue, and from the mixture of yellow and 
blue rays results green, observable in the rainbow. Yellow, red, and 
blue, he considers as three principal points of coloration ; for the type 
of these colours he takes the middle part of the yellow, red, and blue 
bands of the rainbow. He observes that although it is very easy 
to determine the colour of the middle part of the red band, because 
that band has its full width, yet it is more difficult to determine 
that of the blue and yellow bands, because these bands have half 
their width taken up to form the green band. He then forms by 
proportion binaries, and mixtures, of which he gives an instructive 
table, and an arrangement which clearly points out the difficulties 
that attend accuracy in this subject. It shows also that it is na- 
tural to apply to the prismatic colours for an original source, and 
by that application how various the ideas and modes of improve- 
ment have been, while all have had a similar confidence of being 
most perfect. 
Perhaps there may have been ages of perfection in this subject, 
which are lost, like other things that do not depend upon durable 
materials ; as the statues of Praxiteles, which being handed down to 
us give evident proofs of great perfection in those days. Indeed I 
can scarce help thinking that I have seen something in antiquity that 
belongs to a compounded wedge, such as I have constructed for a 
measurer of colours : if so, it may help to unravel a mystery ; and I 
