Marching for several days through oaks and mesquites, over hills and rising country, we reached Nacori, a poor village in the foot-hills of the Sierra Madre.  It is scarcely forty miles from Granados, and lies at an elevation of 3,700 feet.  Our camp, about two miles outside of the village, was permeated with a delicious odour of acacia blossoms, and water in the neighbouring mountains, though strongly impregnated with iron, was quite palatable.

In this region Mr. Hartman found a new form of agave with delicate stripes of white on the lanceolate leaves that constitute the basal rosette of the plant.  The flower stalk is only twelve or thirteen inches high, and I should not wonder if this diminutive and beautiful century plant some day became fashionable in greenhouses.  It grows in large numbers in the crevices of the rocks, the perpendicular walls of canons often being studded with the bright little rosettes when the drought has withered all herbaceous vegetation.

From here I made an excursion to an ancient pueblo site.  As usual, there were traces of small dwellings, huts of undressed stone, and fragments of pottery.  We found three mortars and one pestle, a remarkable number of metates (the stone on which corn is ground), and the corresponding grinding stones, showing that a large population must have once lived here, huddled together in a small space.

But the most striking feature of antiquity met thus far on our journey were curious stone terraces built across the small gullies.  They are called trincheras (trenches).  Some of them do not appear to be very old, and many present the appearance of tumble-down walls, but the stones of which they are constructed were plainly used in their natural state.  Although many of the boulders are huge and irregular in shape, they were used just as they were found.  The building material always conformed to the surroundings:  in places where conglomerate containing water-worn boulders abounded, this was used; where porphyry was prevalent, blocks of that material were employed.  There is no trace of dressing or cutting, but in the mason work considerable skill is evident.  The walls are not vertical, but incline somewhat toward the slope on which they are erected.  The terrace thus formed is often filled with soil to the height of the wall-top for a space of from fifteen to twenty feet.  Earth taken from them does not show any colours.  Some of these trincheras measure thirty feet in length by four feet in height, while the smallest ones I saw were only five feet long and three feet high.  Naturally enough, the largest ones are in the lower part of the gullies; then, some twenty-five feet back and above, others almost as large may be found.  As the arroyo rises and narrows, the walls, each placed a little higher up the slope than the preceding one, are necessarily smaller.

In the mountains near Nacori, especially on their eastern and southeastern sides, trincheras were encountered in every gulch as high up as six thousand feet, though steep crests and the mountain tops bear no traces of them.  In one arroyo, which was about a thousand feet in length and of comparatively gentle slope, twenty-nine trincheras were counted from the bed of the main drainage to the summit of the mountain.  Some of them were quite close together, three being within eighteen feet of one another.