in the Burnt Isles, in the Kyles of Bute. 81 



of the rocks, it is impossible that any single conflagration could 

 have produced such eiFects. 



In an article in the ninth volume of the Edinburgh Ency- 

 clopaedia, written, I believe, by Sir GEORGE MACKENZIE, these 

 effects are attributed " to making signals by fires," chiefly be- 

 cause those hitherto known have been placed in commanding si- 

 tuations. I apprehend, however, that this will not account for 

 the fort in question, because, in the first place, the situation, in 

 a flat, surrounded on all sides by hills of considerable elevation, 

 does not appear at all calculated for such a purpose ; and, in the 

 next place, the regularity of its form seems still more inconsist- 

 ent with the effects of any accidental cause. 



We must, therefore, I think, conclude, that, in whatever man- 

 ner these singular buildings were constructed, or for whatever 

 purpose, they are the effects of design. They were probably 

 constructed at a period before the country was cleared of its ori- 

 ginal forests, when abundance of fuel, and ignorance of other 

 modes of cementing stone, had induced the inhabitants to resort 

 to the expedient of joining them by fusion. 



VOL. x. P. i. 



