CHAPTER II

The Archaeological Evidence below Ground

The removal of the superstructure of Brooke House left the area unencumbered, with the bases of walls exposed and providing an outline plan of the building as it had come down to the twentieth century. In some places (and particularly on the south and east sides) there were clear indications of variations in the plan; and a dual attack on the problems of the site was therefore possible: a programme of surface-clearing which would complete the layout by adding superficially buried features to those already visible, and a sequence of trenches to determine the relationships of the different elements to one another both structurally and in regard to the accumulated deposits (Plate 42).

Before describing the outcome of these investigations several general observations are necessary. In the first place the stratigraphical evidence from the site was so meagre as to be practically valueless. Over much of the area of the building and its courts there was a varying depth of made-up ground which was quite without definite floor levels and which, in any case, rarely produced datable material of significance. A considerable quantity of pottery was found, but the bulk of it was of eighteenth- or nineteenth-century date, relating to the later occupation of the house and therefore of no value for the earlier phases of its history which were the chief target of this excavation. As will be seen below even twentieth-century pottery was found well down in the deposits, which were sometimes contaminated throughout their full depth by objects of late date. And here the second factor operated to handicap attempts to establish the chronological and structural succession on the site. Where stratigraphy and associated finds are lacking, differences and resemblances in building practice, in the use of distinctive materials, mortars and the like, may shed at least a limited light on the relationships of the different parts. At Brooke House, however, the foundations were entirely of brick, and apart from the obvious work of the eighteenth and nineteenth centuries, the general character of the brickwork and the mortar used with it were so uniform that they provided no help in dating, and sometimes even added their own complications.

As already stated, the examination of the buildings was carried out by two methods. In the early stages the surface-outlines were exposed by clearing the wall-tops. This operation in itself had the effect of adding new details to the plan with comparatively little effort, for in many places only a shallow deposit covered older walls of whose presence there were various hints. Side by side with this it was necessary to establish levels; and for this purpose two cuttings were made at right angles along the main axes of Court I and its surrounding ranges which were carried down to the surface of the natural sand throughout the greater part of their length.

These cuttings showed a depth of about 4 feet of material within Court I—an accumulation which, as already noted, was without significant features and showed signs of postTudor disturbance down to considerable depths. In the open areas, both in the courts and outside the limits of the house, the build-up of the surface appeared to have been due in part to normal accumulations in the earlier phases—apart from traces of brick paving at footing-level on the east side of Court I there were no signs of structural floors—in part probably to the deliberate raising of the surface in Phase II (see below). The soil had been very much deepened in later years by garden cultivation. The lower levels were remarkably devoid of finds wherever they were tested. The site produced a very small quantity of late medieval or Tudor pottery as compared with the abundance of eighteenth- and nineteenth-century wares; and none of this was found in positions that were significant in relation to the rest of the structures.

The extension of the main east-to-west cuttings beyond the limits of the buildings was carried out as part of the normal technique of excavation, aimed at achieving a full understanding of the site and its surroundings. Here the outcome was to reveal a ditch in a position compatible with the description in the survey of 1547 referred to on page 59. The full width of this ditch was seen only in the one cutting, A (fig. 9; the section there shown is a composite to which other cuttings have contributed). Its existence was tested in other cuttings both to north and south. As a defensive feature the ditch could not be said to be very impressive. Though at its deepest it was almost 10 feet below the modern surface, nearly 5 feet of this was due to recent accumulation, dumping and cultivation; its original depth had been only just over 5 feet and the effectiveness of this was nullified by its weak profile. The ditch had poorly defined edges and a long inward slope to a 6 foot berm at the base of the west wall of the house.

The filling of the ditch varied from one section to another. A silty mud in places suggested the presence of water, but this was not everywhere present and it cannot be certainly maintained that the ditch was wet; a patch of silt could indicate that there was standing water in places. There were other semi-natural deposits with a high sand content deriving from the sandy nature of the gravel into which the ditch had been dug. Some tipped and sagging layers, however, contained a good deal of tile, mortar and building débris, suggesting that after a time the ditch was not merely being neglected but was actually being filled in. Fragments of Delft and stoneware from near the floor of the ditch at its deepest part provided evidence that it must have remained open until into the seventeenth century, but the processes of filling and silting up seem to have been completed during that century. By this time a stable surface was established 3–5 feet below the present level. Though it varied in character (being in one place overlaid by a gravel layer suggesting a path) this surface was consistently present in all the sections. A large fragment of a bellarmine was found immediately beneath this gravel layer in Cutting D, and there were fragments of Delft at intermediate levels. The considerable depth of soil overlying this sealing layer was due in part to further building up of the surface. Some building débris was incorporated in it and it had been subjected to prolonged gardencultivation. From this accumulation came eighteenth- and nineteenth-century sherds.

No attempt was made to determine the complete course of the ditch. Observations subsequently made during building excavations showed that the ditch continued northwards along the same line for a distance of at least 80 feet from the north-west angle of the building. On the south side it turned eastwards through a right-angle about 25 feet from the south front.

Fig. 9. Composite section across ditch on west side of building

In anticipation of what follows it should be said here that, matters of detail apart, the result of the excavation was to reveal that Brooke House had at its fullest development (fig. 12) consisted of two courts, of which that on the south (henceforth referred to as Court I) was the earlier. The northern court (II) was an extension from the simple quadrangle of the first building, its latest version being a rationalization of a series of buildings which appear to have been grouped round an open space, though they were erected separately and were not always structurally linked. Court II was not as regular as Court I. Its eastern range was set obliquely to the main north-to-south axis, no doubt to conform with the property-boundary on this side of the site. This boundary had similarly controlled the placing of the isolated buildings which first occupied the area.

There were indications, however, of building, also in brick, earlier than Court I. Clear evidence of this was met with in one place: there was a second possible occurrence in the north-east angle of the court.

The early wall lay beneath the east (outer) wall of the east range, its north end being 4 feet south of the entrance porch (Plate 42): it was 26 feet 6 inches long, and ended on a transverse wall, not quite at right-angles to it, which projected to east and west beyond the later wall for a short distance. To the east (that is, outside the range) the transverse wall, set on an irregular brick offset, was 16 inches wide. Only about 5 feet of it survived on this side: it had been cut by a more recent drain and the whole area had been so much disturbed that nothing had survived beyond. On the west side (within the east range) the transverse wall extended westwards for about 6 feet in a curiously irregular way, with a surviving fair face only on the south side: here too there was no indication that it had been prolonged westwards. The north-to-south wall had several offsets on each side, in consequence of which it projected beyond the later wall, and its course was completely seen. At the north an expanded foundation and a finished end on the west side marked a corner (Plate 34a), while on the east a blunt stump survived to mark the beginning of another wall at right-angles to it. Again there was no trace of this return in the small area of much disturbed ground that was available for examination. The evidence points to a rectangular building parallel with the main axis of the east range, just over 26 feet long and of uncertain width, lying between Brooke House and the present road. The projection at the south-west angle remains unexplained. This building was dismantled almost to its foundations and its reduced west wall then served as part of the foundation for the outer east wall of Brooke House. The lower parts of the new wall were carried to and butted against the ends of the existing wall, leaving straight joints between the old and the new work. A levelling course of tiles was inserted to carry the new wall along the old—a characteristic piece of construction met with elsewhere on the site. The date and purpose of this first building are not known: it cannot be said whether it relates in any way to the subsequent first main 'period' of Brooke House, though there is no marked difference between the brickwork of the two structures.