the nails inserted through it. Nearly all the soles were strongly made and studded with nails. The light shoes made from a single piece were not uncommon, in these the leather being brought together at the heel, where alone there was any stitching (Plate XX., Figs. 1 and 3). Laces passing through the loops on the sides kept the shoe on the foot. The variety just described had no nails. Such shoes should probably be termed carbatinae. A similar shoe is still in use in Shetland and in Eastern Europe.[1]

Generally speaking, shoes were much more common than boots, but it could not be said that there was any predominating type which marked itself out as the soldiers' ordinary wear; it was probably one of the coarser forms illustrated in Plate XX., such as Fig. 4. The soles with their heavy tackets were abundant. Sometimes the strong counter of the heel remained in its original position, even when the upper had entirely disappeared. Again, in many cases the shoes were obviously those of women and children rather than of grown men. There was no evidence that the sandal was in common use although one or two examples were noted.

Fig. 4 is probably a second-century type. It was found with others resembling it in the inner ditch of the East Annexe. Fig. 5, which came from Pit XXV, represents the calceus, the close-fitting type of shoe. It probably belonged to a youth. It is remarkably well preserved, even to the little tag at the heel for pulling it on to the foot. The nails form a decorative pattern on the sole, a feature not uncommon in the shoes of the period. Many of the shoes are remarkably fine examples of the skill of the leather worker, who exercised the greatest ingenuity and skill in cutting out the uppers, and occasionally in stamping them with lines of small decorative punch work. One of the most remarkable specimens (Plate XX., Fig. 6) came from Pit XVI. It was probably made for a woman, judging by its size and shape. Notwithstanding the lightness of its uppers, its sole is heavily covered with nails. Its light openwork associates it in the same class as the more simple caligae from Mainz, and like them it must belong to the first century. A child's shoe of the same sort was also found, but in less perfect preservation. Like the larger shoe, it had nails in the sole, and the same was the case in a tiny shoe which must have belonged to a child about four years old. In Fig. 7 we have another example of finely cut leather work. Others, again, may be seen in the fragments illustrated as Figs. 11, 12, 15, and 17 of Plate XIX., where also we have

1 Haverfield, The Classical Review, Vol. xii. p. 142.