at each stage like the Boulogne Tour d’Ordre, and formerly rising, we may infer, to a total height of about 80 ft. Its walls, 12 ft. thick at the bottom and 7 ft. at the top, have a core of rubble and white mortar and a facing of green sandstone and tufa held by pink brick-dust mortar and levelled at regular intervals of seven courses with double (rarely single or triple) courses of brick. The arches of the doorway and windows are decorated by the alternate use of stone and brick voussoirs in a common Roman fashion (pp. 67, 74). It was apparently divided into four stories, each with three or four semicircular-topped windows, which vary from 2 ft. wide at the bottom of the tower to 4 ft. wide at the top. A curious feature of some of these windows is that the actual external opening was originally reduced to a small spy-hole only 2 ft. high and 1 ft. wide. This external skin has in most cases  disappeared completely with the

FIG. 9. DOVER: UPPER WINDOW OF THE PHAROS ON CASTLE HILL SHOWING INNER WALL AND SPY HOLE
(From a drawing in Miss Jessie Mothersole’s The Saxon Shore, p. 123, made from a photograph by Mr. Seers)

 facing of the main structure, but survives in the eastern window of the third stage (Fig. 9). Presumably, in the upper story or stories, fires or torches shone out through more ample openings, but of this no trace now exists, except perhaps in the larger size of the surviving upper windows. As to the date of the structure, we can only guess. Suetonius tells us that in A.D. 40 Caligula commemorated his abortive advance upon Britain by erecting a pharos upon the opposite shores of Gaul; and an early date for the establishment of its counterpart on the Dover cliffs is, on a priori grounds, likely. The site itself has obvious advantages for the purpose. Here a tower, standing some 380 ft. above sea-level, would be visible all the way from the French coast, if the weather were clear and the light strong enough.⁷⁷  Moreover, the Dover sea-fogs not infrequently cling to the lower levels and leave the heights visible

from the sea above a bank of mist, and the position of the pharos suits this well. It is a further question whether the tower stood alone or with other buildings. Traces of foundations are vaguely stated to have been observed in the vicinity. But the church of St. Mary’s, which is immediately east of it, is Saxon and indeed late Saxon. The earthworks round it show no sign of Roman origin in their present state; nor is the situation likely to have been chosen for a Roman fort, and the whole idea that the Castle was begun by the Romans seems a medieval or
   ⁷⁷ Mathematicians calculate that, so far as the curvature of the earth is concerned, the Dover pharos would have been visible from sea-level for about 26 or 27 miles. Dover to Grisnez is 21 miles, to Boulogne 30 miles.