We have now, I think; obtained sufficient evidence that the disposition of differently coloured or composed bands in agate is in most cases the result of crystalline segregation.We shall find, also, that the order of this segregation is constant under given conditions; and that, with fixed proportions of elements and fixed rate of cooling and drying, the agate will necessarily produce itself in a riband of a fixed succession or pattern of stripes: a spectrum of substances, which, if we had observed data enough, we might read like a spectrum of light; inferring, not the nature of the elements from its bars of colour, but the: former conditions of solution from the bars of elements.