146 PAPEUS, ETC. 



Imndred acres rises abruptly from the marsh, terminating 

 at one extremlty in a higli and bold peak of similar cliarac- 

 ter to Glastonbury Tor, the summit of wlilch is crowned 

 by an eai'thwoi'k, whicli, from its position when seen from 

 a distance, appears to be a military work of great impor- 

 tance. This, however, will be found on closer inspection 

 not to be the case. The ramparts, though of considerable 

 magnitude, and enclosing what may possibly have been a 

 small village arranged on the threefold System, are of the 

 simplest construction, with hardly any attempt at outworks 

 beyond an escarpment on the steep side of the hill and a 

 few terraces commanding the principal entrance. The 

 whole, however, of the elevated piain before mentioned has 

 been enclosed with a low agger, and probably a palisade, 

 and must have resembled a large park. Within this 

 enclosure thcre is a fine spring, and a more favourable 

 Situation for cattle, when driven by high tides or stormy 

 weather from the marsh pastures, can hardly be imagined. 



There is one more of these stations to which I wish to 

 draw attention. This is situated on the first rise of the 

 ground, at Cannington Park, from the level of the marsh. 

 It consists of several large enclosures, with little, if any, 

 attempt at systematic fortification. It is held by sonie 

 authorities that the primoeval tribe of the Cangi, who are 

 Said to have inhabited the country between Quantock and 

 the sea, were rather a body of professional herdsmen than 

 deserving the appellation of a distinct people. If this was 

 the case, the fact of one of these stations being situated in 

 the immediate vicinity of Cannington — a name probably 

 derived from the eai-ly occupants of the district — consider- 

 ably strengthens the probability of my guess at the purpose 

 for which they were established. 



There is hardly an elevated spot on the whole mai'sh, or 



