40 The American Geologist, July, i898 
resulted in so blocking the main channel as to cavtse the 
stream to cut through this solid mesa. May it have been that 
the flow of mud due to the water from the volcano dammed 
the stream and caused this side erosion, or, what is more 
probable, that the filling up of the old valley by the loess and 
then the gravel lifted the level of the entire valley so high that 
the water foimd its way over the lava area and then, when the 
gradual removal of the glacial detritus below reduced the 
level, the water wore through this sheet of lava and excavated 
the new channel before that removal had opened up its old 
channel. 
In this connection it may be noticed that upon the sum- 
mit of this insular mesa there is what seems to be the rem- 
nant of an aboriginal fortification consisting of two walls of 
stone meeting at a right angle to the northwest, the longer 
extending south about loo yards, the other east about sev- 
enty-five yards. At present the walls are loose ridges of lava 
blocks about six feet high and ten to fifteen feet wide. There 
is some evidence that the ground on the inside near the wall 
has been excavated several feet. The outline of only one 
pueblo of small size could be seen within the area thus partly 
protected, though chipped flints and potsherds are numerous. 
It seems hardly likely that the embankment was raised against 
water, and, in fact, there is at the angle an opening in the west 
wall like that of a sally port or door. The basaltic walls of 
the mesa and fragments at its base are covered with aboriginal 
tracings in which the figure of the turtle and the four pointed 
star seem predominant. 
III. The Bernalillo Volcano. 
The town of Bernalillo is sixteen miles north of Albu- 
querque on the east bank of the Rio Grande. At this point 
the Cretaceous emerges boldly from beneath the river gravels 
and loess. All the river deposits betray the influence of the 
Cretaceous below some time before the latter actually out- 
crops. Opposite the town and for some miles to the north- 
ward the river is bordered by long horizontal bluffs capped 
by basaltic flows of no great thickness, but of many miles ex- 
tent. These blufifs afford the most instructive illustration of 
base-leveling that could be imagined. (Fig. 5.) The basalt re- 
