22 Proceedings of the Royal Irish Academy. 



feature of the fortress. The abattis consists of a closely set mass of little 

 piUars, usually 3 to 4 feet high, girding the whole middle walls in a band 

 from 30 to 80 feet wide, more open between the north-west and northern 

 gates, but nearly impassable to the north-west and to the east, at which 

 latter side they are set with wonderful pains up a steep rock-slope below the 

 rampart The tops of the pillars, as noted by Dr. CoUey March,' are greatly 

 worn and furrowed by the weather, like those at Ballyldnvarga, and give a 

 more con\Tneing proof of age than is aftbrded by the facing of the wall, which 

 is less fretted, though probably already weather-worn when raised from the 

 crag ; for in many forts on the mainland we have seen evidence of such wear 

 on faces embedded in the wall. We must bear in mind, however, that 

 similar e\idence of weather-wear is found on the upper parts of sculptured 

 crosses of the ninth to the twelfth century ; and the base of the twelfth- 

 century cross of Dysert O'Dea is also deeply fretted. The rock at Uun 

 Aengusa did not afford such convenient crannies to form sockets for the 

 pillars as were found at the Black Fort ; nor was the soil sufficiently deep 

 (as at Ballykinvarga) to fix them ; so, in many cases, they simply lean against 

 each other or fall over in picturesque confusion.* There is no outer kerbing 

 or later annexe to the abattis, as at Ballykinvarga or the somewhat similar 

 fort at Mohne in the Baltic ; nor, like the former, liave they lesser spikes 

 between the pillars (spikes sharp enough to cut through the side of a boot) ; 

 but they are jagged and sharp indeed. O'Douovan exaggerates when he 

 writes that " many of them are so sharp that, if one fell against them, they 

 would run him through " ; but they are verj- perilous to pass, even when 

 undefended. He pictures<iuely compares them to an army petrified in act of 

 attack.' The band measures about 700 feet from the west to the north-east 

 gate, and over 200 feet more from it to the clifif eastward. 



The feature is very rare ; it occurs at Dun Aengusa, the Black Fort, and 

 Ballykinvarga, and was once found at Dunnamoe promontory fort in Mayo, 

 the pillars from which were used for house-building in Belmullet.' In Great 

 Britain patches of such stones are set to form obstacles at the more accessible 



' " X«. fif., roL IT., atx. ii., p. 226. 



« S« Pltte I., fig. 2 ; nnd Pi«te III., fig. I. 



' Mr. Burke chooaes the lesa dignified and ntlieT miilesding simile of "almonda in a pudding," 

 for the stones are not set out apart as diavn by Cheync. Tfaeae inaccurate vieirs led to the theory 

 noted by Mr. Wakeman, that they were tombstones of those buried round the fort; or by Dr. March, 

 that they were to protect cattle from alingen. 



'Journal K.S.A.I., toI. xij. (consec.),p. 182 ; Trotter's " Walks in Ireland," pp.503, 504. He 

 calls them "stone stakes of great sixe and height." Ber. Caesar Otvray says that O'Donovan 

 remembered them " more numerous and much larger " than in 1S41 ; but they had been *' removed 

 for sills and lintels." See " Erri< and Tyrawley," p. 69. 



