of a thin line of red, bounded by blue on one side and yellow on the other. If these triple lines approach very near together, the intermediate space must appear green; a phenomenon which will often occur to us as we proceed.
377.
We frequently meet with these colours in nature. The colours of the spider's web might be considered exactly of the same class with those reflected from the steel wire, except that the non-translucent quality of the former is not so certain as in the case of steel; on which account some have been inclined to class the colours of the spider's web with the phenomena of refraction.
378.
In mother-of-pearl we perceive infinitely fine organic fibres and lamellæ in juxta-position, from which, as from the scratched silver before alluded to, varied colours, but especially red and green, may arise.
379.
The changing colours of the plumage of birds may also be mentioned here, although in all organic instances a chemical principle and an adaptation of the colour to the structure may be assumed; considerations to which we shall return in treating of chemical colours.
380.
That the appearances of objective halos also approximate catoptrical phenomena will be readily admitted, while we again do not deny that refraction as well may here come into the account. For the present we restrict ourselves to one or two observations; hereafter we may be enabled to make a fuller application of general principles to particular examples.