Very interesting is the gradual artistic evolution of the features. Here, as in organic evolution, there is a process of specialisation, the primordial indefinite form taking on more of characteristic complexity. In the case of the eye, for example, we may often trace a gradual development, the dot being displaced by a small circle or ovoid, this last supplemented by a second circle outside the first,[244] or by one or by two arches, the former placed above, the latter above and below the circle. The form remains throughout an abstract outline or scheme, there being no attempt to draw even the lines—_e.g._, those of the lid-margins—correctly, or to indicate differences of light and dark, save in the case where a central black dot is used. In this schematic treatment so striking and interesting a feature as the eye-lash only very rarely finds a place. A similar schematic treatment of the eye in the use of a dot, a dot in a circle, and two circles, is observable in the drawings of savages and of Egyptian and other archaic art.[245] ----- Footnote 244: In one case I find the curious device of two dots or small circles, one above the other within a larger circle, and this form repeated in the eye of animals. Footnote 245: An example of circle within circle occurs in a drawing by a male Zulu in General Pitt-Rivers’ collection. ----- The evolution of the mouth is particularly interesting. It is wont to begin with a horizontal line (or what seems intended for such) which is frequently drawn right across the facial circle. But a transition soon takes place to a more distinctive representation. This is naturally enough carried out by the introduction of the characteristic and interesting detail, the teeth. This may be done, according to M. Perez, by keeping to the linear representation, the teeth being indicated by dots placed upon the horizontal line. In all the cases observed by me the teeth are introduced in a more realistic fashion in connexion with a contour to suggest the parted lips. The contour—especially the circular or ovoid—occasionally appears by itself without teeth, but the teeth seem to be soon added. The commonest forms of tooth-cavity I have met with are a narrow rectangular and a curved spindle-shaped slit with teeth appearing as vertical lines (see the two drawings by boys of six and five, Fig. 6 (_a_) and (_b_)). These two forms are improved upon and more likeness is introduced by making the dental lines shorter, as in Fig. 5 (_c_) (p. 340). With this may be compared a drawing by a boy of five (Fig. 6 (_c_)), where however we see a movement from realism in the direction of a freer decorative treatment. [Illustration: Fig. 6 (_a_).] [Illustration: Fig. 6 (_b_).] [Illustration: Fig. 6 (_c_).]