The study of colouration among existing animals has led to the conclusion that in mammals the primitive colour-pattern was that of stripes, either longitudinal or transverse and more probably the former. In the second stage these bands break up into spots, which still show the longitudinal arrangement and may be either light on a dark ground, or dark on a light ground. In a third stage the spots may again coalesce into stripes, the course of which is at right angles to that of the original stripes, or the spots may disappear, leaving a uniform body-colour, lighter or white on the belly. These changes of colour-pattern have not proceeded at a uniform rate in the various mammalian groups, or even within the same group, for an all-important factor is the mode of life of the particular animal. In general, it may be said that the scheme of colour is such as to render its possessor inconspicuous, or even invisible, and many a creature that seems to be very conspicuous and striking in a museum case can hardly be seen at all when in its natural surroundings. Thus, Arctic mammals and birds, in their winter dress, are white; desert animals are tawny or sandy-brown; forest animals are frequently striped or spotted; while those that live on open plains are more commonly of uniform colouration. There are exceptions to these rules, but they hold good for the most part. From careful comparative study of the teeth and skeletons a clew may be gained as to the habits of animals and from the habits something may be inferred as to the colouration. [Illustration: FIG. 6.—_Tapirus terrestris_, 3 days old. Compare with Fig. 137, p. 320. (By permission of W. S. Berridge, London.)]