If there be any truth in the supposition, the colouration of every animal must be due to the development of one or more of these molecules. Colouration may be expression of the arrangement of all the molecules in the fertilised egg, or it may be due to the development of a number of molecules whose function is to determine the colouring of an organism, or it may be the result of the development of one such molecule, which perhaps splits up in such a way that a portion attaches itself to each of the other molecules. But it is idle to speculate on this point. As we have already insisted, the tendency to build up elaborate theories on very slender foundations is a too frequent failing of zoologists. We desire merely to emphasise the fact that the phenomena of animal colouration almost force us to the conclusion that the colouring of each organism is the result of the development of a number of units. It may be objected that, if this be the case, the number of the units which contribute to the colour of any organism must be exceedingly large, since we see in nature an almost limitless number of different schemes of colouring. If the colour of each animal be the result of the development of a few units, it might be thought, firstly, that the diversity of schemes of colouration which we observe in nature could not possibly occur; and secondly, that, under such circumstances, the colour pattern of a bird or beast should be of the nature of a mosaic, each colour being sharply defined and separated from every other colour, instead of the colours shading one into the other, as is so frequently the case. Such objections would be based on a misconception as to the nature of the units which combine to produce the colouration of an organism. _These units show themselves as centres of development of colour_, as points from which the colour or colouring they represent spreads, until it meets and mingles with other patches of colour which are being developed from other centres. The colour produced at one centre may spread more rapidly than that which forms at another; this, of course, will result in a preponderance in the organism of the colour which is produced at the former centre. Further, we must bear in mind that the development of each colour-producing unit is largely affected by conditions external to it, as we shall see when dealing with Sexual Dimorphism. More than one naturalist, who has paid careful attention to the subject of animal colouration, has perceived that through the apparently endless diversity of the colouring of organisms something like order runs. Mr Tylor Quoted Over thirty years ago Mr Alfred Tylor called attention to this important fact. That observer, whose views met with the approval of Wallace, was of opinion that colour follows structure, and that in a many-hued animal it changes at points where the function changes.