32 BUREAU OF AMERICAN ETHNOLOGY [bull. 54 



is now under cultivation and irrigated from the small stream. This 

 wider portion of the canyon, less than two miles in length, was occu- 

 pied by a considerable population in prehistoric times, and to some 

 extent in more recent times, and contains the only agricultural land. 

 This opening is the site of one large pueblo and several smaller 

 houses, while cliff dwellings line the cliff on the northern side for a 

 long distance. As this cliff recurs in all the canyons of the district, 

 always on the northern side, and is of great archeologic consequence, 

 it is more fully discussed in subsequent paragraphs. Still farther 

 down, where the stream has cut into the basalt and sandstones, the 

 canyon is again narrow. Strangely, instead of following the contact 

 or cutting through the tufa, where the canyon reaches the hard 

 basalt, it cuts through the basalt to great depth in order to reach 

 the Rio Grande. In this vicinity the latter stream itself, instead of 

 following the contact of the basalt and tufa or cutting the latter, 

 has eroded a deep, narrow gorge though the basalt. The phenomena 

 at the upper falls of the Rito de los Frijoles suggest that the basalt 

 did not reach the top of the tufa when the canyon-cutting began or 

 the stream would probably have cut into the tufa or along the con- 

 tact ; time did not permit, however, full investigation of this matter. 

 If the gorge were well established when it reached sufficient depth to 

 encounter the basalt, the stream would then have continued to cut 

 downward along the same line instead of swerving from its course, 

 according to the law which has been well named ' ' the persistence of 

 rivers." Just before reaching the Rio Grande the Rito de los Fri- 

 joles forms two cataracts of 60-foot and 90-foot fall, respectively, 

 by plunging over two precipices of basalt in the bottom of the 

 canyon. 



This toj>ography certainly existed virtually as at present during 

 the occupancy of the canyon by the ancient pueblo and cliff dwellers, 

 affording them a secluded retreat in a region difficult of travel. It is 

 wholly impracticable to travel across the plateau in such direction as 

 to intersect the canyons. The trails mostly foUow the mesas which 

 separate the caiiyons, or pass around their heads. Travel in the 

 Frijoles canyon above or below the open valley is possible but diffi- 

 cult even on foot, which of course was the only method of travel 

 available to the ancient inhabitants. Access from the south to the 

 open part of the canyon is possible almost anywhere, though the 

 slope is steep. On the north the vertical or nearly vertical cliff pre- 

 vents access except in a very few places, makmg this side easy to 

 defend. 



The whole aspect of the region is that of immature topography. 

 The land forms are abrupt. Steep cliffs abound. Angles have not 

 even begun to give way to rounded outlines. The cutting of canyons 

 has proceeded rapidly, from the geologic point of yiew, though that 



