In dealing with the first of the above groups of dyes, the direct dyes, the colourist is somewhat at a loss to explain in what manner the combination with the cotton fibre is brought about. The affinity of cellulose for dyes appears to be so small and its chemical activities so weak, that to assume the existence of a reaction between the dye-stuff and the fibre, tending to the formation of a colour lake, seems to be untenable. Then, again, the chemical composition and constitution of the dyes of this group are so varied that an explanation which would hold good for one might not do so for another. The relative fastness of the dyes against washing and soaping precludes the idea of a merely mechanical absorption of the dye by the fibre; on the other hand the great difference in the fastness to soaping and light between the same dyes on cotton and wool would show that there has not been a true formation of colour lake. The dyeing of cotton with the second group of dyes is more easily explained. The cotton fibre has some affinity for the tannic acid used in preparing it and absorbs it from the mordanting bath. The tannic acid has the property of combining with the basic constituents of these dyes and forms a true colour lake, which is firmly fixed on the fibre. The colour lake can be formed independently of the fibre by bringing the tannic acid and the dye into contact with one another. In the case of the dyes of the third group, the formation of a colour lake between the metallic oxide and the colouring matter can be readily demonstrated. In dyeing with these colours the cotton is first of all impregnated with the mordanting oxide, and afterwards placed in the dye-bath, the mordant already fixed on the fibre then reacts with the dye, and absorbs it, thus dyeing the cotton. To some extent the dyeing of cotton with the basic dyes of the second group and the mordant dyes of the third group is almost a mechanical one, the cotton fibre taking no part in it from a chemical point of view, but simply playing the part of a base or foundation on which the colour lake may be formed. In the case of the dyes of the fourth group, there being no chemical affinity of the cotton known for them, these dyes cannot be used in a successful manner; cotton will, if immersed in a bath containing them, more or less mechanically take up some of the colour from the liquor, but such colour can be almost completely washed out again, hence these dyes are not used in cotton dyeing, although many attempts have been made to render them available. Indigo is a dye-stuff that stands by itself. Its combination with the cotton fibre is chiefly of a physical rather than a chemical nature; it does not form colour lakes in the same way as Alizarine and Magenta do. Cellulose can be dissolved by certain metallic solutions and preparations:--