THE RITUAL OF nARROWS AND CIRCLES. 243 



Stones thrown loosely together in a circular ridge ; at the entrance 

 one large stone " remained standing. Borlase also described 

 and figured *' an oval enclosure at Kerris, made of a continuous 

 ridge of stones of all sizes, ^hich contained an area of fifty-two 

 paces by thirty-four. In South Brittany the circles are composed, 

 not of rude blocks like the stones of the alignments, but of thin 

 slabs which, says Mr. Lukis, f " are designedly contiguous, as 

 though to make a perfect enclosure impervious from the outside, 

 and only accessible from some point where they are in contact 

 with the head-stones of the avenue lines." 



Sometimes the circle is composed, to all outward appearance, 

 of a rampart of earth only ; as the Giants' Ring near Belfast. 

 Here, the wall, probably not less than twenty feet high, is 

 eighty feet thick at the base, and encloses an area of ten 

 acres, where stand an uncovered dolmen and a few remaining 

 monoliths. 



And there are multitudes of earth- circles, all over Britain, the 

 remains, no doubt, of substantial enclosures. They are often 

 marked "camps" on the ordnance maps, but for such a purpose 

 they are too small. 



Lastly, there are the stone circles that are themselves ringed in 

 by vallum and fosse. Examples can be seen at Arbor Low in 

 Derbyshire, at Avebury in Wilts, at Blisland in Cornwall, where 

 the ditch is eleven feet and the wall ten feet wide ; at Stennis in 

 the Orkneys, where the wall is three feet high ; and at the 

 neighbouring Ring of Brogar where the ditch remains but the 

 wall, which may have been only a stockade, has vanished. In 

 by far the greater number of such cases, the fosse is on the 

 inside of the vallum, and it is obvious that such a work had no 

 military intention. But an inside ditch would be good against 

 the predatory wolf. For if the wolf scaled the wall, pushed 

 through the frieze of thorns, crossed the ditch and seized his 

 prey, in the ditch he would have to leave it on his retreat, and in 



* Plate XV. fig. 2. 

 t Chambered BaiTovvs, &c., of S. Brittauy, p. 35. 



