472 A VOYAGE TO Book VL 



part reached quite round tlie wall, tliere was only 

 one ascent to it, which was adjoining to the oval 

 tower. The outward and inward walls are all of the 

 same kind of stone, ver}' hard and well-polished : 

 and disposed like those of Callo. The apartments 

 also were without ceiling or flooring, like those of 

 the above-mentioned palace. 



At Pomallacta, within the jurisdiction of the 

 town of Guasuntos, are some rudera of another 

 fortress like the former : and it is a common opinion 

 here, that there was a subterraneous communication 

 between these two fortifications ; but this does not 

 seem at all probable. For besides the distance of 

 six leagues, the ground is very uneven, and inter- 

 rupted by some of the smaller branches of the cor- 

 dilleras, breadles, and brooks. The inhabitants are, 

 however, very tenacious of their opinion : and some 

 affirm, that a few years before our arrival in the coun-r 

 try, a person entered this subterraneous passage at 

 the fort of Canar, but, his light going out, he was 

 obliged to return. They farther say, that the en- 

 trance is within the fort at the foot of the tower, 

 where indeed there is a small low door, but now 

 choaked up with earth; and was doubtless for some 

 use. But this does not imply that it led to the other 

 fortress, as, besides a great quantity of lights, there 

 must also have been here and there vent-holes or spi- 

 racles, which, considering the mountains, is utterly 

 impracticable. 



Many other walls and ruins are seen all over the 

 country, both in the plains, on the sides of the 

 hills, and on their summits; but most in desert 

 places, and without any vestige of a town or village 

 near them ; and except these three, they are either 

 of adoves or unknown stone, without any arrange- 

 ment. The more irregular are thought to be the 

 >Yorks of Indians before they were reduced by the 



yncas; 



