borough have been dug up. In this part of Peterborough the houses are small, and no deep excavations such as might have disclosed the existence of old buildings have been made, except those along the main streets for the deep drainage of the town.

Up to the year 1879, there had been no more important find than that of an occasional coin or specimen of pottery; but at that time an unusually well-read and intelligent workman, employed with others in excavating limestone at a point on the north-west side of the borough rescued this vase, which I exhibit, from a labourer, who had driven the point of his pick into it, and brought it to me. The site of the excavations which have led to these discoveries is marked by an asterisk on the plan of the district, Plate I, and is more clearly shown in Plate II, which is reduced from the Ordnance Map 6 in. to the mile. From that time, for some years, I continually visited these stonepits, watching for any discovery that might be made and securing such objects as I could. The result of my own observation, and that of others, is the conclusion that at this point, very early in the occupation of this island, the Romans established a summer camp-Castrum Estivum-probably an outpost from Durobrive; this military post would probably only be garrisoned when the dryer weather of the summer, rendering the marshes more passable, opened the way for attack by the tribes inhabiting the more elevated tracts of land which exist as islands or peninsulas in the great plain of the fens. know from history that these tribes, the Gyrvii, with their neighbours the Iceni and Corytavi, long resisted the Roman dominion. The road to Denver, to which as it left Castor I directed your attention, would pass very near this spot, running on through the north-eastern and eastern part of the old town round the present Abbey by the east side of Bridge Street to the river edge, probably to a tongue of land to the east of the present bridge, which was found a few years since to have been raised and strengthened by ancient fagot-work. On the opposite bank, when excavations were being made in 1890 for the building of an engine-house for Messrs. English's saw-mills, at a considerable depth below the surface an

We

tained, I may state that a superficial examination showed the existence of a low embankment (Plates III and IV), which on section is seen to be 40 ft. wide with a trench on one side only, the southern side; this is still traceable for

ancient paved road was cut through, which was probably the continuation of this road, although a gap of some miles exists before its course can be traced across the fens. Returning to the site whence these objects were ob

some 600 yards it runs in a north-easterly direction to a point where, as I have been informed by the workmen who quarried the limestone in this locality, it turned at a sharp angle to the south in the direction of the river, which runs east and west about a mile to the south of the supposed camp. The extent of this last embankment cannot be determined, as the ground between it and the river had been levelled and covered with buildings and railroads many years before attention was directed to it, there are now only a few yards of it remaining. The two banks formed together an angular inclosure, one side looking north, the other east; the angle projecting northeast, the ditch or fossa being on the inner side of the bank. This would be the natural form of the walls of a camp made to protect the interior from attack from the north and the east.

It is a point of interest that, within the angle formed by the two banks there is a spring which yielded an abundant supply of excellent water, and which was up to quite recent times the favourite source of drinking-water for those residing in its neighbourhood; this was probably a factor in the selection of this site for a camp more than 1800 years ago. At some distance from either embankment, but still within the space included by them, is what is shown on section to have been an extensive shallow pit (inset in fig. III), 4 ft. deep, excavated out of the top layer of the rock. It is 36 ft. across in one direction, the extent in the opposite direction has not been determined; it is filled with dark vegetable mould, which for a part of the extent is divided horizontally by a layer of rough stone; in the vegetable soil both above and below this stratum of stone are many fragments of Roman pottery, some ordinary Castor ware, others more highlyfinished Samian ware.

Mr. Irvine has made drawings of sections and a plan of this excavation: it may have been a shallow stone-pit, which was afterwards filled with rubbish, the imperishable earthenware remaining while the organic matter returned to dust; or it may have been a pond for watering horses or cattle, the layer of stone being thrown in to improve the bottom when, after years of use, it had be