Even among those skinks which have never broken their tails there is a wide range of variation in relative length of tail. This is partly a matter of relative growth since the proportions change during the course of development. Also there may be slight sexual difference and there is much individual variation. In fetuses still well below hatching size, the tail length is less than the snout-vent length. For instance, an egg in a natural nest 12 days short of hatching contained a fetus that had a snout-vent length of 14 mm. and tail length of 12 mm. (Figure 18). In the late stages of fetal development the tail growth is relatively rapid. At hatching, the tail is considerably more than half the total length. In a large series of young with snout-vent lengths from 30 mm. down to hatching size of 25 mm. or less, the tail length averaged 130.8 per cent of snout-vent length. In larger young, up to a snout-vent length of 40 mm. or more, the tail continues to lengthen more rapidly than the body. In skinks that are about two thirds grown, the tails average relatively longer than in either larger or smaller individuals. In the sample representing the size class 50-54 mm. snout-vent length, the tails average 163.3 per cent of the snout-vent lengths, whereas in groups of adults of various sizes and both sexes, the tail length is near 155 or 156 per cent of the snout-vent length. Sexual dimorphism in tail length is slight if it exists at all; in adult males, tails averaged a little longer than in adult females. [Illustration: FIG. 18. Diagram showing relative tail-length (as a percentage of snout-vent length) in skinks of different size groups that retain their original tails unbroken; in the early stages of growth the tail becomes relatively longer as size increases, but the trend is reversed before adult size is attained. For each series the mean, standard error, standard deviation, and extremes are shown.]