A HISTORY OF DEVONSHIRE 



top. When the stones were removed from this space and replaced on the walls, it was apparent 

 that these latter had never been more than 5 ft. high or 5^ ft. at the very outside. 



We thus have a double wall with a space between. This space, it would seem, is too narrow 

 to serve any useftil purpose. There is apparently an entrance into it from the interior of the pound 

 on the NW. side, but with one most doubtful exception no others were observed. The balance of 

 evidence is rather against this presumed entrance being a work of design. It is clear that the wall 

 has never been built solid throughout, except at the flanks of the gateway ; a vacant zone of irregu- 

 lar width being everywhere else left between the inner and outer facings, which are of very varying 

 thickness. The most reasonable supposition is that this space between the two facing walls was to 

 have been filled with a core of earth which in its turn might have carried a wooden fence or 

 palisade. But there is no trace of such earth-filling, and if it ever existed it could not have utterly 

 disappeared. It could not possibly have been washed away, for the walls would have retained it, and 

 ' vegetation would have grown up and held it fast. It seems not unlikely that the work was never 

 finished, for it is impossible to conceive that the builders should have spent immense labour, and 

 rested content on raising a double wall with a space between and a total thickness of some 10 ft. 

 and height of only about 5 ft. — too low to keep out either a hungry wolf or active human 

 assailants. A single solid wall carried to a greater height, with a banquette, would have been more 

 easily erected and much more effective. The unfinished character of the wall is strongly supported 

 by Mr. C. W. Dymond, F.S.A., who subjected it to a rigorous examination at the request of the 

 writer. 



We are not familiar with the conditions of warfare existing at the time when Grimspound was 

 designed and built, but it strikes one at once, on viewing the site, that it is commanded on all sides 

 but one by the fire of archers and slingers, and that the entrance is so placed that it is in the most 

 favourable position for a rush of assailants down the slope of Hameldon. 



However unsuitable the position may appear to us to-day in a strategic sense, we must not 

 conclude that this cyclopean defence was erected with no object other than to be a fence to fold 

 cattle. Some of the stones employed in the wall are of great size. On the west is a huge slab set 

 on edge measuring 10 ft. by 5 ft. and gin. to i ft. in thickness, and weighing from 3 to 4 tons. 

 Other stones laid in courses, if not so long, are not of less weight. Such a wall was not erected by 

 the occupants of a dozen huts and for the protection or convenience of a limited number of persons. 



Presumably Grimspound was not a fortified village any more than it was merely a cattle 

 pound, but was more probably the unfinished eppidum, the place of refuge for the scattered popula- 

 tion in the neighbourhood, and the dozen householders were the oppidani, the caretakers or guardians. 

 The lunette arrangements on the western side were probably cattle-pens. 



Grimslake, a small stream that dries up in very hot summers only, flows through the enclosure 

 at its northern extremity. It passes under the wall, flows through it for some way, and then 

 emerges three-quarters of the way down. It has been confidently asserted that the stream was 

 diverted from its proper course by the hands of man to force it to enter the pound. This, however, 

 is not the case. There is an outcrop of feldspathic granite in the midst of the furrow between 

 Hameldon and Hookner Tor, so that the drainage of these hills cannot unite, but flows parallel in 

 independent channels, and meets below the road from Chagford to Widecombe, at a distance of 

 400 yards from the enclosure. There is at present very little drainage from Hookner Tor : never- 

 theless, it must have been greater in former times, as the groove of the channel and its stream is not 

 only distinctly visible, but is more important than that of Grimslake. 



Raddick Hill near Princetown (O. S. cvi, SE). — This is a good example of a pound 

 enclosing hut circles. The wall of the enclosure is in an imperfect condition, but enough remains 

 to show that this was not a massive defensive wall like Grimspound, but was probably a mere corral 

 for harbouring cattle by night. The hut circles within the pound are eleven in number, and of 

 these seven gave unmistakable evidence of having served as habitations. They varied in internal 

 diameter from' 10 ft. to 25 ft. 



The most interesting of the series is one which has a diameter of a little over 1 5 ft. It is 

 paved throughout, the entrance ill-defined, but apparently facing SSW. Nearly opposite this 

 supposed entrance is the hearth, slightly raised, backed by a large stone in the wall. To the right of 

 this hearth is the cooking or fire^hole, under a large slab that has fallen forward out of the wall. 

 This slab, in falling, protected from destruction a fine vessel of rude hand-made pottery, which 

 occupied the cooking or fire-hole, and was fairly intact. It had the peculiar shape of the cinerary 

 urns used in the early Bronze Age, with a rib running as a hoop round it 2^ in. from the lip, and 

 the portion above this was ornamented by rude chevrons. The bottom of the pot was rounded. This 

 was distinctly seen when it was in situ, but when taken out the bottom, which was very rotten, went 

 to pieces. The vessel measures I o|- in. in height and was 1 in. in diameter at the top. The rib 

 was I in. thick. 



Although the utmost care was taken, considerable portions of the wall of the vessel crumbled 

 on air-drying, but these were restored, and this interesting specimen of a cooking pot was deposited 



352 



