24 Proceedings of the Rojjal Irish Academy. 



describes the wonderful flask that kept hot drinks warm and cold drinks 

 cool. Old fiction, however, usually based its non-magical surroimdings on 

 " things seen," and it is CAndent that spikes, if not of stone, " brass," or 

 " iron," at least of wood, girt many a fort in ancient Ireland. 



The abattis of Dim Aengusa has been i-emoved for a short distance at each 

 end by idlers who love to hurl stones down the precipice into the sea below. 

 There is, as we noted, an avenue nearly 80 feet long through the pillars to 

 the north-east gateway; it is probably ancient, being similar to that at 

 Ballykinvarga. O'Donovan overlooked the accessibility of all tlie cliii-edge 

 when lie supposed it a modem work to give access to the sea-face.' No 

 avenues lie to the north and north-western " gaps " ; the former (as we 

 tried to prove) is in the later reach of wall ; the latter was a shapeless gap, 

 with no trace of piere, in 1878. It is not necessary even to regard such an 

 avenue as made in ancient, but more peaceful, times later tlian the actual 

 foundation of the fort. The gangways left in the rock-cut fosses at Doon 

 Fort= probably the reputed scat of the brother of Acnghus tlio Firl>olg) and 

 Lisduft', near Kilkee, sliow tliat tlie old forl-dwellers little regarded this 

 undiiubtcd weakening of their defence. This was not from thoughtlessness 

 in early times ; for in the " Book of Leinster " the danger was noticed. " It 

 is a peril to be upon the fort unfortified; and the .shout of the person in its 

 door that lias conquered it."' The only fairly defensible gateway of an Irish 

 fort known to me is at Dunbeg in Kerry ; perhaps, too, at Dunnamoe, the 

 entrance was capable of more than mere passive resistance.' 



The Fracmext. — To the north and west of the central fort is a fragment 

 (as I believe; of the old outer wall. It is 7 feet 6 inches thick and about the 

 same height at the east end, and is about 250 feet long with a terrace 4J feet 

 high and \vide. It lies 54 feet from the middle wall at that end, but 

 approaches it to within 20 feet to the west. It is entirely levelled near the 

 cliff and from the upper ridge to the north-east gate, while much of its western 

 reach is very low. 



The Middle Wall. — The eastern part was evidently the old outer wall, 

 and the western, from the " liastion " westward, the older middle one. The 

 intermediate part was an aftertliought. It is tenaced throughout and varies 

 much in height, being from 5 or 6 feet at the north-east to nearly 12 feet high 

 at the upper ridge. In 1878 long reaches of the terrace were extant, but no 



" " 0. S. Letten," p. 213. 



'Journal B.S.A.I., vol. xiTii., p. 126. 



'" Book of Leinjtcr," p. 37. 



* Miss Stokes notes the advance implied by the gateways of Irish stone-forts over the gaps in 

 British forts ("Christian Architecture of Ireland," p. 26). Perhaps the "gaps" bad wooden 

 gateways : some Irish ones have lining slabs. 



