^7 
stance. Velvet has various appearances, from its construction, distinct 
from colour. 
Emerald green is good ; but f;reen is very properly a comj)ound, 
and the emerald may be more or less perfect, thick or thin, &c. 
" Reddish white" might do very well without the explanation, which 
makes it quite absurd: it is "snow white with a little crimson and 
ash gray." If any philosopher or other person were to endeavour to 
compound this confounded mess, they will see no natural means of 
doing it. First, they must have, I suppose, a j)erfect white, for snow 
white; then some white mixed with black for ash gray; and lastly, 
some "carmine red with a considerable proportion of blue, for 
crimson. If 1 can guess at the sequel, it will prove a very muddy 
affair. " White with a little red" would have done better, and would 
have been both natural and simple. 
'I'his I think a sufficient example* It might be deemed invidious to 
point out more, and the whole speaks for itself. I have done thus 
much as a duty; others may take the task of judging and cri- 
ticizing. What I have done v\ill show them that thej^ may pro- 
duce any variety of tints, even beyond all calculation, if the three 
simple primitives, as I call them, are used. I am only aiming at simpli- 
city, the foundation of every thing natural. 
When I wished to have concluded, which I should have done Ion"" 
since on the original plan, I happened occasionally to meet with 
something which many persons might think should not be passed 
over. However, as I might seem negligent if I did not make, some 
ajjplicable observations, I say a few words on the Decomposition of 
Light, by C. A. Prieur*. 
He observes that " the disk of the moon, when far from the horizon, 
with a single prism displays upon her lengthened image several 
circles, which indicate separations of colours perfectly unblended." I 
take the liberty of observing, that the more the image is lengthened, 
* Annales de Chimie, torn, lix. p. 227. Phil. Mag. vol. p. 162. 
