special attention to two groups of ditch and bank work, and suggested that in one case (see diagram i) the banks probably represented field divisions. He pointed out, what indeed is a significant fact, that although about forty hut-floors occur just outside these enclosures, none arefound within them. Had Mr. Petrie's theory as to field divisions been proved by further examination to be tenable, this circumstance might have been very naturally explained by sup- posing that all traces of any floors within the enclosures had been destroyed during the process of cultivation. Careful examination of the site, however, enables the present writer to state with confidence that the soil, at this part of the Common, at any rate, has never been cultivated. The absence of floors within the enclosures is therefore of considerable value as tending to show that the latter were for the purpose of enclosing cattle at night, or perhaps at other times when wolves were in the vicinity. Doubtless the banks were surmounted by hedges or fences. The banks may be described as protective rather than strictly defensive earthworks. Although the forms of the enclosures are somewhat irregular, there is a pronounced tendency towards a square or oblong. This will be seen in the diagram No. i, representing the en- closures containing no hut-floors, already mentioned. These enclosures, which are doubtless coeval with the neolithic dwellings, probably represent successive stockaded enclosures, be- i i o A o o o o o 1 1 i '^ 0=0. j i I ( Diagram No. 2. Enclosures at Hayes, Kent. Neolithic Flakes from Millfield, near Hayes Common. (scale : ^ linear). cause the ditch from which the material for the bank was derived occurs in some cases on the inside of the enclosure, and this points pretty clearly to intervals between the construction of different enclosures. In the diagram J shows an earlier enclosure than B, and C is earlier than D. In D, how-ever, there seems to have been originally a dividing line cutting the oblong into two nearly equal shapes. All traces of the dividing line are lost, and the thin dotted line in the diagram is merely conjectural and intended to explain the peculiarities of the enclosing ditches and banks. Another group of enclosures, probably of earlier type, is shown in diagram No. 2. In this case the rectilinear work A is probably earlier than the oval enclosure B, because the former was partly destroyed when the latter was constructed. Both square and oval en- closures were probably disused at a very early time, as traces of neolithic dwellings are found both within and without the banks, and indeed in one case a hut-floor has been cut somewhat into the actual bank of the oval work. Keston. At Millfield, Kestcm,' adjoining Hayes Common, the site of a factory of neolithic ' Proc. Soc. Antiq. (ser. 2) xvii. 216-21. 316
