179^' antiquities in Scotland. ' I03 



firmly in the wall, beginning at the outside of the 

 wall, both on the convex and concave surface, so as 

 to mak-e these surfaces tolerably smooth ; the stones 

 having been left to terminate in the cavity of these 

 galleries as chance occasioned. The floors also were 

 ill laid, and so open as to admit of seeing through 

 every seam. 



The next remarkable particular that struck me, as a 

 very great singularity, which had never been taken no- 

 tice of by any one who had described these structui'es, 

 was, that these galleries grew narrower and narrower, 

 as yoit ascended upwards, till they closed at the top al- 

 together, as is represented in the elevation ; a good 

 many of these galleries, therefore, must have been so 

 narrow as not to have admitted accefs to any hu- 

 man being ; yet still we find openings from within, 

 into these narrow galleri*?s, as well as into those that 

 were wider. 



It deserves in the next place to be remarked, that 

 it is impofsible to get accefs from below to a supe- 

 rior story of these galleries by means of the stair, 

 without going entirely round the whole circuit of the 

 building in the gallery ; so that if these had been 

 employed as apartments for sleeping in, or for any 

 othar purpose, one person going between the top 

 and the bottom, must have displaced every person, 

 or every thing of half the bulk of a man, in every 

 ^art of the whole. This consideration, alone, Ihows 

 the striking absurdity of supposing that these gal- 

 leries had ever been appropriated for the use of rhan, 

 as apartments for lodging in. But the most decisive 

 proof that these structures never could have been 

 iinnloyed as habitations for men of any sort, is still 



