across it without perceiving its existence, though I was looking for it, and no one I fancy would remark it if his attention were not turned to it. Its boundary mounds never could have been 3 feet high, and now in many places are very nearly obliterated.
That these alignments were once race-courses, appears to me one of the most improbable of the various conjectures which have been hazarded with regard even to Stonehenge. No Roman race-course, that we know of, omitted to provide for the horses returning at least once past the place they started from, and no course was even a mile, much less a mile and three-quarters long. What sort of horse-races the British indulged in before the Conquest I don't know, nor will I hazard an opinion on the subject; but if they wanted the races to be seen, there are several beautiful and appropriate spots close at hand where they could have laid out a longer course along one of the bottoms, where tens of thousands might conveniently have witnessed the sport from the sloping- banks on either hand, whereas here only the front rank could have seen the race at all, and that imperfectly. It may also be remarked that the east end of the cursus is closed by a mound which must have been a singularly awkward position for the judges, though that is the place assigned to them by Sir Richard; and the west end is cut off also by an embankment, behind which are several tumuli on the course, which seems a very unlikely racing arrangement.
But if not race-courses, what were they? If any one will turn back to woodcut No. 12, p. 55, representing the alignments at Merivale bridge, and compare them with the cursus as shown in woodcut No. 26, p. 102, representing the ground about Stonehenge, I think he must perceive that the two cursus, if complete, would occupy exactly the same relative position with regard to Stonehenge—on a much larger scale of course—as those at Dartmoor do to the circle there. The arrangements are so similar that the purposes can hardly be different. At first sight this seems to tell against the battle theory. We know of no battle fought on Salisbury Plain. This, however, is the merest negative assumption possible. We know that the massacre at Amesbury was followed by a four years' war, between Ambrosius
