WicSTROPP — The Fort of Dun Aengusa m Inishmorc, Aran. 29 



was 11 feet lower than Iho uLliers. Wakemaii went so far as to draw an ideal 

 restoration of Dun Conor with a sunken way round the top. Some censured 

 the restorers for having obliterated this interesting feature, and refused to 

 believe that the wall had had an inner section till confronted with the third 

 Dunraven photograph. This feature (which I sketched and well remember) 

 first led me to reconsider O'Donovan's account in 1904, when I saw that by 

 the "internal " division he meant the banquette, being indeed tlie true inner 

 section, which removes the apparent inconsistency of his further statement 

 that the two outer divisions were of equal height, and explains his allusion to 

 the section 4 feet lower than the summit in all the other forts which have 

 banquettes behind the outer walls exactly as O'Donovan describes. 



The rampart has a slight batter (usually 1 to 5, or 1 to 7), but is usually 

 distorted and bulged out. It is 12 to 13 feet high at present, but rose 

 in parts to 18 feet high in 1839 ; resting on a low, and evidently scarped, 

 ledge of rock, 3 or 4 feet high, all round which, when covered by debris, may 

 have brought the old height to 18 feet. The third Dunraven photograph, 

 Mr. Cheyne's view in 1847, and two of my camera sketches in 1878, show 

 that a large patch of the facing opposite the north-west gate had fallen, 

 showing a second face inside ; and I recollect this condition in 1878. The 

 outer section had only one face, with filling between it and the next (or 

 middle) section ; but the latter had two faces. The inner sections were 

 terraces, the lower 4 to 7 feet high. The upper is, I think, a modern 

 development, 4 feet higher, as the outer sections were certainly of the same 

 height in 1839, and even in 1878. As we noted, a line of large foundation 

 blocks on the ledge outside, and to the south of the gate, implies a later 

 rebuilding. So do the joints, first noted by Mr. Lynch, but shown in 

 Burton's sketch of 1857 ; they lie 23 inches to the south, and 30 inches to the 

 north of the entrance outside,' and 14 inches to the north, and 16 inches to the 

 south inside. This suggests a rebuilding of the present door and outer wall in 

 early times, as does the useless ope to the north-west under the terrace. 

 The wall is of unusually good, though somewiiat small, coursed masonry, 

 with a facing of headers. A few larger blocks, or perhaps only stretchers, 

 3 and 4 feet long, are found in the lower courses. The masonry, as usual, 

 gets smaller about 8 or 9 feet up, owing probably to difficulties in lifting 

 the blocks. It is inconceivable that so many persons sliould ha\-e described 

 such masonry as " cyclopean." That of the neighbouring churches better 

 merits this oft-abused term. 



The gateway is perfect, facing slightly to tlie muth of east, and is a 



' See PI. III., fig. •->. 



