Annual Meeting. 81 



the earliest period of occupation. As the foundations of the 

 island rotted and sank, the inhabitants found it necessary to 

 build the island higher by adding another course of clay, with 

 new hearths on top of the older. At a later period, a further 

 sinking of the foundation made another floor necessary, apparently 

 the last one in the history of the lake dwelling. 



Ashes, burned fragments of wood and bones, were numerous 

 in the immediate vicinity of the three hearths, and here and at 

 the various places where we excavated were found a considerable 

 number of fragments of broken pottery cooking vessels. It is 

 worthy of remark that these pottery fragments were all found 

 fairly near the surface, not at the level of the lov/est stratum of 

 occupation. The pottery is all of that distinct type which I have 

 in former papers described as souterrain pottery, and not of the 

 type usually found in crannogs. None of it showed signs of 

 having been turned on a wheel, though some of it had thumbnail 

 ornamentation. 



A few years ago Mr, F. J. Bigger made some slight excava- 

 tions in the form of small trenches in this crannog, and found 

 some small fragments of the same pots, which he presented to the 

 City Museum. These and what we found are exhibited together. 



Mr. Bigger, in a short account of his investigation (published 

 in the Ulster^ Journal of Archaeology, 2nd Series, Vol, VII, 

 p. 195) states that he also found a " Danes pipe" which, with 

 the pottery fragments, he had presented to the Public Museum. 

 The records of the City Museum show that the pottery was given, 

 but no pipe, and I am inclined to be sceptical about the latter. 

 Mr. Bigger makes the extraordinary error of describing this 

 crannog as a Dun. 



In our investigation we found no implements or fragments of 

 metal of any sort, and no flints ; we found a stone implement, a 

 flat oval, measuring 5| " x 4$ " x 2 ", slightly hollowed in the centre 

 of each face and chipped at the edges. I cannot suggest for what 

 purpose this implement served. 



From conclusions arrived at in former papers in investi- 



