Page:VCH Derbyshire 1.djvu/281

 ROMANO-BRITISH DERBYSHIRE arched passages, some voussoirs of which were found lying near (fig. 18). The excavators thought that the two archways were unequal in size one being large for wheeled traffic and one small for foot passengers. The western gate was found in 1905 to be exactly similar, even in the in- equality of the two entrances. The south gate, on the other hand, had a single archway 1 1 or 12 feet wide, with a guard-chamber on each side. The rampart had also corner turrets, and close to the south- eastern turret an oven or hearth was found in 1899. In other forts frequent traces of cooking and kitchen refuse have been discovered inside the turrets. Of the interior of the fort little is known. The headquarters building was partially cleared in 1899 (fig. 19). It measured 76 by 87 feet, was entered by a door not quite in the middle of the north side, and contained at the back three rooms, which may possibly have I ' ~|?-;'p. ""' once been divided by wooden partitions into the more usual five rooms. But no vault was found in it, nor have any certain traces been noticed of the usual colonnade and double court. It may have been simpler in plan than the ordinary headquarters building, and this, combined with the absence of a vault, may be thought to suggest an early date. Of other buildings only faint traces have been uncovered. On the west of the central building two tiled pilee of a hypocaust, some flooring of broken roof-tiles and a doorstep found in 1899 and 1905, may indicate a considerable structure. Near the north gate a floor of burnt clay (or something resembling clay) and some stout oakstakes sunk below the floor, found in 1905, suggest wooden barrack huts. For the rest we must await the teaching of the spade. Remains without the fort are equally little known. Mr. Watson in 1771 noted worked stones above and below ground, and a subterraneous stream of water below the north-east angle of the fort near the Glossop Brook, which may possibly represent the bath-house, and another building has been suspected near the north-west gate. Of the cemetery hardly a vestige survives. In 1841, when a mill-dam was altered 100 yards east of the fort near Glossop Brook, a stone coffin was found, and near it a bronze coin of Domitian, but the two are probably not coeval. 1 1 Hamnett, Deri. Arch. Journ. xxi. 10. Mr. Hamnett also mentions 'a sepulchral urn of red earthenware,' found about 1800 at Woolley Bridge, half a mile north of Melandra, but now destroyed. 213 FIG. 19. HEADQUARTERS, MELANDRA.