ON THE AGE OK STONE CIRCLES. 429 



levels on fixed lines giving contours of 1 or 2 feet (30'5 or 61 cm.) vertical 

 height, and to fill in intermediate 15-2 cm. (6-inch) contours by the eye; 

 but to ensure absolute precision, to show the little knolls and depressions 

 on the vallum at the south-west, north, and east, to mark the irregu- 

 larities made by Bateman on the summit of the tumulus, to indicate the 

 little dyke running in a southerly direction from the vallum — it was 

 desirable that all the 6-inch contours should be surveyed severally, which 

 entailed the necessity of taking some eighteen hundred levels ! 



The periphery of the crest of the vallum constitutes almost a true 

 circle, with a diameter of exactly 76™'25, as shown by the outer circle 

 described on the plan. The centre of this circle comes near the middle of 

 the south-western side of stone, No. III. of the central group. The crest 

 of the vallum deviates very slightly in any part from the true circle 

 excepting on the north-west, where it bulges out. The bottom of the 

 fosse, as seen on the surface of the silting, declines from the line of the 

 true circle far more than the rampart, as shown by the inner circle 

 described on the pJan, with a diameter of 58 metres ; the only segment of 

 this circle that can be said to be true is on the south, south-west, and 

 west. The ditch is thrown out far more than the rampart to the north 

 and north-west ; but it would not be expected to find that the fosse silted 

 up regularly and symmetrically all round, whereas the crest of the ram- 

 part, of course, is much about in the same position as it was at the age of 

 construction. 



The ditch was marked by a depression from the original surface all 

 round averaging 1™'37, and it is surprising that in the course of all 

 these ages it should not have silted up to a greater extent ; had the monu- 

 ment been situated in a chalk district, the ditch would probably have 

 been indicated by a much shallower depression on the surface. 



The average height of the vallum above the general surrounding turf- 

 level is 1^-83 (6 feet)— Dr. Brushfield states 16 feet, i.e., i-^-SS.' Judging 

 from those jDortions of the ditch already excavated, the material obtained 

 from the fosse when it was first excavated was not enough to form the 

 vallum, but the construction of the latter will be mentioned later on 

 when dealing with the excavations. The confines of the rampart are 

 bounded at various points by ten small Governmental stones. The fosse 

 and vallum are interrupted on the north-west and south-east by the 

 entrance causeways, which are not in line with the central group of 

 stones. The causeways are on the same general level as the area occupied 

 by the megaliths and the surrounding land. The circumference of the 

 rampart, including the entrances, is about 246 metres. 



Txhe vallum is joined on the south-west by a slightly raised bank — 

 about 30 cm. in height — and an almost imperceptible ditch, which runs 

 for some distance in a southerly direction. It would be desirable to cut 

 a section or two across this so-called ' serpent,' to ascertain if it is of the 

 same date of construction as Arbor Low itself, or more recent. 



On the south-east, adjoining the external face of the vallum and partly 

 resting on it, a tumulus stands, the summit some 2™*13 above the sur- 

 rounding tui-f level. 'Between 1770 and 1824 three unsuccessful 

 attempts had been'^made to discover an interment, but a fourth, made by 

 Mr. T. Bateman on May 23, 1845, resulted in its discovery. About 

 46;cm. above the natural soil a large slab, l'"-52 broad by 91-5 cm. wide,/ 



' Journ, Brit. Arch. Assoc, 1900, p. 129. 



