20 Proceedings of the Royal Irish Academy. 



up the slope, wliere it is better preserved. The masonry is coarser to the 

 east when compared with the upper reaches and the inner walls. To the 

 north, we find its most interesting feature, the nearly perfect gateway, 

 first described by Lord Dunraven, before 1875, as being 4 feet wide and 



3 feet high (over the debris), with a lintel 9 feet long. This is virtually 

 correct, as we found it to measure 49 inches wide above, but only 45 inches 

 below, owing to a projecting block at the base. The height, as now cleared, 

 is 4 feet 10 inches to the east and 4 feet to the west, being on a steep, 

 ridgy slope. The wall is 6 feet 7 inches to 7 feet thick on top, and nearly 

 8 feet at the base, the passage being covered by a huge inner lintel, 8 feet 



4 inches long, 6 to 10 inches thick, and 15 inches deep and two similar outer 

 lintels. The space between them and the inner one is covered by short 

 " cross-bearere," which is also the ease in the north gate of the inner wall, 

 and is a common feature in soutcrrains, but rare in gateways. From it 

 westward the wall has been greatly and unnecessarily rebuilt, raised to a 

 level top, about 5 feet high, till we reach the upper ridge ; there it has been 

 repaired with a terrace, and is 2 or 3 feet higher. The ridge crosses it, and is 

 from 12 to 18 feet high. 



. _: .V.., NORTH GATE. 

 Fio. 3. — Dun Aeogusa, Outer Gate. 



O'Donovan notes that it had tw^o sections. They are not apparent in the 

 untouched wall. It (verj' likely) had a ttiTace, though none appeared (so far 

 as I remember) in 1878. He was fond of asserting at that time (it was very 

 early in his field-work) that the sections in fort-walls were to provide another 

 solid face agsdnst the enemy when the walls were sapped and fell in a siege. 

 Most of the succeeding writers have adopted that view. We have, however, 

 no suggestion of so elaborate siege-work in Ireland in those early times; 

 and I believe the system was adopted rather to allow the more equal settle- 

 ment of the dry-stone walls, which, when of any gieat tliickness, naturally 

 (especially if the filling be small) bulge out, and even burst the faces of the 

 wall. Another possible reason was that the terraces and outer sections 



