BRITISH CATTLE STATIONS, 147 



Immediately near it, undisturbed by modern cultivation, 

 wliich does not bear the marks, more or less distinctly 

 defined, of early occupation. I imagine that most of tliese 

 mark the sites of British Cattle Stations, of which none 

 but, perhaps, the largest and most important were occu- 

 pied except during the dry months of summer, at other 

 times they must, from the natiire of the ground, have been, 

 before the construction of the Roman sea walls, almost 

 totally inaccessible. Nor would the marshes, during the 

 winter, have produced herbage of much value for bucolical 

 pm'poses. This may account for the absence of more dis- 

 tinct traces of permanent residence than I have discovered 

 at any of these stations. In confirmation of this opinion, I 

 may State that in the immediate viclnity of my own house, 

 at Bishop's Lydeard, a slight elevation of red sand runs 

 out, like a promontory, into the line of meadow, which, at 

 the time of which I am speaking, must have been a marsh, 

 resembling, on a small scale, that surrounding Glaston- 

 bury; and that the field which occupies its ridge, and 

 shows some faint marks of ancient works, is still known by 

 the name of Half Yard, which, I believe, would slgnify the 

 Summer enclosure. 



I have then now done my best to point out the different 

 types of primajval earthworks most commonly met with 

 in this district. They are, as I suppose, the aboriginal 

 type, marking the sites of permanent fortified towns, 

 distinguished by their threefold arrangement, somewhat 

 analogous to that of a Norman castle. The purely military, 

 or, as I suppose, Belgic type, distinguished by its concen- 

 tric arrangement and the Cattle Stations, differing from 

 both the others in the great size of their external inclo- 

 sures, and the absence of any very imjDortant or complex 

 military works, and, as far as I have been able to observe. 



