2 20 Proceedings of the Royal Irish Academy. 



was I disappointed, as the remains of what were probably six forts were found. 

 Ln Cliara, as in Mayo generally, the rock-platform type is well represented, 

 but two examples of the stone-walled headland, and one of the normal 

 entrenched earthwork, occur. It is, of course, to be expected that others 

 remained till the sea cut them away. Probably, save in a sheltered position, 

 all the oldest forts have long since been so destroyed. Even of those that 

 remain in Co. Mayo, the evidence of sea-cutting in the later ages is every- 

 where — the stupendous cutting-away of Dun Briste and its buildings from 

 the fortified promontory of Downpatriek Head, the cutting-away of the 

 island and hut at Cashlaunicrobin, of much of the garth and even half 

 the late peel tower at Dookeeghan, of most of Durmri nulla, and I think of 

 the neck of Horse Island if (as is apparently certain) the latter was fortified. 

 Even in history we read of the breaking of the neck of Duross peninsula by 

 the sea in 1394 as recorded by Mac Firbis. I have so recently discussed this 

 question of erosion and the general question of promontory forts that I need 

 do no more than refer to the previous papers. 1 



To the south-east of the lighthouse a natural depression and a bold 

 ridge behind it suggest a site, but no trace remains ; probably a landslip in 

 very early times broke away the upper part and left no space behind 

 the otherwise most suitable natural fosse. 



Dunallia and Duxtraxeex. — (Ordnance Survey Map, 6 inches to the 

 mile, No. 75). In Ballytoohymore the map shows two islets off the coast 

 called " Doonallia " and " Doontraneen." I expected that they were the 

 remains of promontories the necks of which had been cut through by the 

 sea, and was not a little surprised to find their true nature. For the first 

 and northern is really a peninsula, and the other is accessible when the tide 

 goes out. 



The names mean either Cliff fort, or O'Malley's fort, and fort of the 

 withered grass stems. 



Dunallia resembles on a very much smaller scale such cliff forts as 

 Banes' Island, in Co. Waterford, Illaunadoon in Co. Clare, 2 Darby's Island 

 near Kilconly in Co. Kerry, and Dunaneanir in the Mullet, all very fine 

 examples. I have never seen any shore forts exhibiting such great 

 ramparts as the fortified rocks inland to which they are akin, such as 

 Cashlaun Gar, in Co. Clare. Such forts usually are only very slightly 

 fenced as a protection against falling over. Dunallia is no exception. 

 The narrow neck and gangway are reached down a watercourse and a 



1 Ancient Forts of Ireland, section 120. Journal Roy. Soc. Antiq.. Ir., xxxvii, p. 239, xxxviii, 

 p. 28, xl, pp. 6, 99, 179. Trans. R. I. Acad., xxix, pp. 11, 221, 265. 



- Where the last traces of the fence (extant iu 1875j haTe recently fallen away. 



