¹ There are, of course, a number of minor banks and ditches to which the above definition does not apply. Some of these are probably the boundaries of prehistoric settlements; some delimit agricultural land from moor—or down—land; others define medieval parks, or coppices; such are not considered in this paper.

² A fifth dyke, Black Ditches, to the north of the Devil’s Ditch, should perhaps be regarded as part of the system (see p. 139).

³ Apart from a doubtful extension of the Fleam, shown on the map by broken lines, recently revealed by air-photography.

⁴ Hereabouts the scarp, from crest of rampart to floor of ditch, measures 60 feet.

⁵ I am indebted to the Editor of ANTIQUITY for additions to my list. Those interested in the subject will find references to others in Allcroft’s Earthwork of England, PP- 494-522.

⁶ There is reason to suppose that the continuity of the Cambridgeshire ditches was unbroken by road-gaps. Travellers along the Icknield Way probably crossed, under surveillance, a wooden gangway. In the case of other dykes the differentiation between original and later crossing-points should be demonstrable by excavation; in respect to the original openings a causeway of undisturbed subsoil should be present on the line of the ditch; in the case of later gaps cut by road-maker or farmer, the ditch would be filled in— ‘made ground’.

^{page 150 note *} Mr Lethbridge’s report will appear in the Camb. Antiq. Soc. Comm.

⁷ Tacitus, , Annals, n. 19.Google Scholar

⁸ Fox, Arch. Camb. Reg., p. 90.

⁹ Archaeologia Cambrensis 1928, p. 103.Google Scholar

^{page 154 note *} The Editor desires to acknowledge the courtesy of the Cambrian Archaeological Association, the Cambridge Antiquarian Society, and the Cambridge University Press in lending blocks to illustrate this paper.