preparation, it would be loose.  In the case of the polygenetic class, the idea is necessarily twofold.  The dyeing materials are not colours, only colour generators.  Hence in all cases the fabric must be prepared with the twofold purpose—­first, of using a metallic or other agent, capable of yielding, with the dye material, the desired colour; and secondly, of yielding it on the fibre in an insoluble and permanent form.  Now, though I have gone so far into this mode of classification, because it does afford some information and light, yet I can go no farther without getting into a territory that presupposes a knowledge and acquaintance with the chemical structure of the colouring matters as organic substances, which would be, at present, beyond us.  I shall now turn to another mode of classification, which, if not so far-reaching as the other, is at least an exceedingly useful one.  The two methods may be combined to a considerable extent.  By the latter plan the colours may be conveniently divided into three groups:  I., substantive colours; II., adjective colours; III., mineral and pigment colours.

Substantive Dyestuffs.—­The substantive colours fix themselves readily and directly on animal fibres and substances, but only a few amongst them will dye vegetable fibres like cotton and linen directly.  Almost all substantive colours may, however, be fixed on cotton and linen by first preparing or mordanting those vegetable fibres.  Silk, wool, fur, etc., act like fibre and mordant together, for they absorb and fix the substantive colours firmly.  In our experiments we saw that turmeric is one of the few substantive colours fixing itself on both cotton and wool, without any aid from a mordant or fixing agent.  Magenta was also a substantive colour, but Alizarin was certainly not one of this class.

Adjective Dyestuffs.—­Some of these substances are definitely coloured bodies, but in some of them the colour is of no consequence or value, and is quite different and distinct from the colour eventually formed on the fibre, which colour only appears in conjunction with a special mordant; but, again, some of them are not coloured, and would not colour the fibre directly at all, only in conjunction with some mordant.  All the polygenetic colours are, of course, comprised in this class, for example Alizarin and logwood (Haematein), whilst such monogenetic colours as annatto and turmeric are substantive, for they will fix themselves without a mordant on cotton and wool.  The adjective colours can be conveniently subdivided into—­(a) those existing in nature, as logwood (Haematein) and Cochineal; (b) those artificially formed from coal-tar products, as Alizarin (madder), Gallein, etc.