"They have," remarks Mr. Miller, "one curious custom which has always been observed in the pueblo. It is for some one (sometimes several simultaneously) to seclude themselves entirely from the outer world, abstaining absolutely from all personal communication with others, and devoting themselves solely to prayer for the pueblo and its inhabitants. This seclusion lasts eighteen months, during which they are furnished daily, by a confidential messenger, with a little food, just enough to preserve life, and during which time they may not even inquire about their wives or children or be told anything of them though the messenger may know that some of them are sick or have died. The food the recluse is permitted to use is corn, beans, squashes, and buffalo and deer meat; that is, such food as was used before the coming of the Spaniards. This religious seclusion is in honor of the Sun. It is one of the rites of the ancient religion of the Pueblo, preserved and practiced now. One of the old men I talked with said that he had himself the previous year emerged from this hermitage; three others were now in, they having retired to exile in February, 1877, and will emerge in August, 1878, then to learn the news of the previous year and a half."

CHAPTER VII.

RUINS OF HOUSES OF THE SEDENTARY INDIANS OF THE SAN JUAN RIVER AND ITS TRIBUTARIES.

The finest structures of the Village Indians in New Mexico, and northward of its present boundary line, are found on the San Juan and its tributaries, unoccupied and in ruins. Even the regions in which they are principally situated are not now occupied by this class of Indians, but are roamed over by wild tribes of the Apaches and the Utes. The most conspicuous cluster of these ruined and deserted pueblos are in the canyon or valley of the Rio Chaco, which stream is an affluent of the San Juan, a tributary of the Colorado. Similar ruins of stone pueblos are also found in the valley of the Animas River, and also in the region of the Ute Mountain in Southwestern Colorado. Ruins of clusters of small single houses built of cobble-stone and adobe mortar, and of large pueblos of the same material, are to be seen in the La Plata Valley, and in the Montezuma Valley, west of the Mancos River. On the Mancos River are a large number of cliff houses of stone, and also round towers of stone, of which the uses are not at present known. Cliff houses are also found on the Dolores River. Other ruins are found in the canyon of the Rio de Chelly.

The supposition is reasonable that the Village Indians north of Mexico had attained their highest culture and development where these stone structures are found. They are similar in style and plan to the present occupied pueblos in New Mexico, but superior in construction, as stone is superior to adobe or to cobble-stone and adobe mortar. They are also equal, if not superior, in size and in the extent of their accommodations, to any Indian pueblos ever constructed in North America. This fact gives additional interest to these ruins, which are here to be considered.

Two separate explorations and reports upon the Chaco ruins have been made. The first was by Lieut. J. H. Simpson, who examined them in 1849 and first brought them to notice, and the second was a re-examination by William H. Jackson in 1877. He was connected with Prof. F. V. Hayden's Geological and Geographical Survey of the Territories, and his report is in that of Professor Hayden, published in 1878, p. 411.

The canyon of the Chaco, which commences about one hundred and ten miles northwest from Santo Domingo, on the Rio Grande, is quite remarkable. It has enough of the characteristics of the canyon to justify the application of this peculiar term. But it differs from the great canyons in the lowness of the bordering walls and in the great breadth of the space between. Neither Simpson nor Jackson describe the canyon or valley with as much particularity as could be desired, but Mr. Jackson has furnished a map, Fig. 29, showing the course of the stream with the walls of the canyon shaded in, and with the breaks or gullies through these walls reduced to a scale. This shows that the level plain between the encompassing walls ranges from half a mile to a mile in places. The walls of the canyon are composed of friable sandstone, and are usually vertical. Their height is not given with precision. The engraving also shows the outline forms and comparative size of the several structures, with specimens of three varieties of masonry used in the walls. No. 2 shows an alternation of courses of stone from four to six inches thick and from eight to twelve inches long, with intervening courses of several thin stones. The same alternation of courses reappears in the pueblos in ruins on the Animas River, about sixty miles north. The canyon commences very much like the McElmo Canyon in Southwestern Colorado, whose vertical walls are at first about three feet high, with a level space between from three hundred to five hundred feet in width; its walls rising slowly as you descend. Without a present running stream, and bordered with open prairie land, it makes a novel appearance to the eye. Lieutenant Simpson remarks that after leaving the pueblo Pintado, which is above the commencement of the canyon, "two miles over a slightly rolling country, our general course still being to the northwest, brought us to the commencement of the Canyon de Chaco, its width here being about two hundred yards. Friable sandstone rocks, massive above, stratified below, constitute its enclosing walls." [Footnote: Lieutenant Simpson's Report, p. 77.]

And Mr. Jackson, who entered it from the same point, remarks that "two miles from the river we descended into the canyon of the Chaco. It is here only about fifty feet in depth, with vertical walls of yellowish gray sandstone." [Footnote: Hayden's Report, p. 436.]

At a point twelve miles down, at the Pueblo Una Vida, he remarks that "the canyon is here about five hundred yards wide, and is perfectly level from one side to the other."

[Footnote: ib., p. 437.] Farther down the walls of the canyon rise about a hundred feet, as appears in the restorations of the Pueblo Bonito and of the Pueblo of Hungo Pavie. Whether the canyon is accessible or not from the table-land above over against the several pueblos, by means of the arroyos which break through the walls and enter the canyon, does not appear from these reports; but it seems probable, Mr. Jackson says, that near the Pueblo Bonito he ascended to the top of the bluff by means of a stairway partly cut in the face of the rock. [Footnote: ib., p. 448.]