[F3] The frontal index is obtained by multiplying the greatest width above the orbits by 100, and dividing the result by the length of the face, as measured from the alveolar point to a line connecting the supra-orbital foramina. In the case of Egil[F4] the dwarfing of the limbs seems to have been more pronounced. Egil may be regarded as an unimproved 40-inch pony; he met his death some forty years ago by falling over a cliff near Hillswick, Shetland—_i.e._, before the Marquis of Londonderry set about making a short-limbed strain suitable for pit work. A comparison of the skeleton of Egil with that of an Exmoor pony in the British Museum shows that in the northern island pony the limbs were relatively shorter by one inch, and the face by half an inch, than in the southern moorland pony, but, notwithstanding the shortening of the limbs, the front cannon-bones in Egil, as in fine-limbed prehistoric races, are in length seven times the width. [F4] Egil (a four-year-old black stallion) belonged to Mr Anderson, Hillswick. As Egil probably belonged to an unimproved stock, it may be asked, Does his skeleton lend support to the view that modern pigmy horses reproduce, apart from their size, the characteristics of their remote ancestors? In Miocene, as in prehistoric times, there were light as well as heavy horses, but in the light as well as in the heavy each limb had three hoofs (fig. 18). In _Neohipparion_, a late Miocene three-toed horse, about 40 inches at the withers, the skull is longer by 12 mm., and the molar teeth are more specialised than in Egil, the 40-inch modern island pony. In the Equidæ the cannon-bones are especially interesting; strange as it may appear, the middle cannon-bones are relatively longer and more slender in the extinct _Neohipparion_ than in modern race-horses. In Egil the front cannon-bone is 175 mm. (6·75 inches) long and 25 mm. wide; in _Neohipparion_ the corresponding bone has a length of 210 mm. (8·25 inches)—is longer than in a 48-inch Exmoor pony,—and is so slender that the length is nearly nine times the width—_i.e._, relatively more slender than in desert Arabs. Further, in _Neohipparion_ the middle metatarsal (hind cannon-bone) is as long as the femur (thigh-bone), whereas in even the 16·2 hands race-horse Persimmon (fig. 5) the middle metatarsal is only three-fourths the length of the femur. But, though in the 40-inch Shetland pony the skull and the cannon-bones are actually shorter than in the ancient 40-inch Miocene horse (_Neohipparion_), the second and fourth digits are as rudimentary, are as much “splint” bones, in Egil as in Arabs and thoroughbreds. There is hence no evidence of reversion in Egil, no attempt to reproduce the second and fourth toes, which, though shorter, were as complete in _Neohipparion_ and his three-toed contemporaries (figs. 19 and 20) as the large functional middle toe,—in other words, in pigmy horses there is evidence of arrested growth but not of arrested development.