284 Oiccount of antiquities in Scotland: Ftb."£i., 

 as to bind and connect them .fii-mly together, instead . 

 of bringing down the wall,' as would have inevitably 

 liappened without much care or fkill in the work- 

 men. Nor could I help admiring the judgement dis- 

 played in making choice of a stone of this form for 

 the purpose here intended ; as this is perhaps at the 

 same time more beautiful to look on, and pofsefses 

 - more strength, for the same bulk and weight, thaa 

 ' "^'any other form that could have been made choice 

 of. 



The outs^ide of the wall was quite smooth and com- 

 pact, without any appearance of windows, or any o- 

 ther apertures of any kind. The inside too was ■ 

 pretty uniform, only here and there we could pe»- 

 ceive square holes- in the wall, of no great depth, 

 somewhat like pigeon-holes, at irregular heights. 



I have been informed tliat there is in many of ' 

 these buildings a circular pafsage, about four feet , 

 wide, formed in the centre of the wall, that goes 

 quite round the whole, on a level with the floor. I 

 looked for it, but found no such thing in this place. 

 At one place, however, we discovered -a door enterireg - 

 from within, and leading to a kind of stair-case that 

 was carried up in the centre of the wall, and formed 

 a communication between the top and bottom of the 

 building, ascending upwards round it in a spiral form. . 



The steps- of this stair, like all the other stoi.es 

 Rere employed, discovered no marks of- a tool, but 

 seemed to have been chosen with great care of a pro- 

 per form for this purpose. At a convenient height 

 over head, the stair-case was roofed with long flat 

 i&'aaies, g<?ing quite acrola the opening,^ and this rcof . 



