The orbits varied in width and height, the cranial ridge was either single or double, either much or little developed, and the zygomatic aperture varied considerably in size.

I noted particularly that these variations bore no necessary relation to each other, so that a large temporal muscle and zygomatic aperture might exist either with a large or a small cranium; and thus was explained the curious difference between the single-crested and the double-crested skulls, which had been supposed to characterise distinct species.

As an instance of the amount of variation in the skulls of fully adult male orangs, I found the width between the orbits externally to be only 4 inches in one specimen and fully 5 inches in another.
Exact measurements of large series of comparable skulls of the mammalia are not easily found, but from those available I have prepared three diagrams (Figs.

14, 15, and 16), in order to exhibit the facts of variation in this very important organ.

The first shows the variation in ten specimens of the common wolf (Canis lupus) from one district in North America, and we see that it is not only large in amount, but that each part exhibits a considerable independent variability.[23] In Diagram 15 we have the variations of eight skulls of the Indian Honey-bear (Ursus labiatus), as tabulated by the late Dr.J.E.Gray of the British Museum.