no indications of door-ways. The rooms were connected internally by openings, but the building was entered from the exterior by ladders placed against the walls. We named the place Casa del Eco (the house of the echo) from the discovery that words spoken below the building, at the mouth of the cave, were distinctly repeated, producing at first a most startling effect.

Near the Casa del Eco and on the northern bank of the Rio San Juan, another important group of buildings was discovered, the ground plan of which may be seen in fig. 2, plate vi. The walls had entirely disappeared, scarcely one stone standing on another, but the plan of the original structure could be readily traced.

Thus far I have confined my descriptions, with one exception, to some of the mouldering ruins which are to be found in the cañons and cliffs of the northern tributaries of the San Juan river. To the south of this stream and in the valley drained by it, these same remains occur numerously in New Mexico and Arizona. Along the Rio de Chelly, for instance (an intermittent arroyo which penetrates northwardly through Arizona and joins the San Juan at the head of its cañon), there are many ancient structures which present some novel and striking features in pre-historic architecture. About eight miles from the mouth of the Chelly is a collection of cliff-buildings which extends uninterruptedly for a distance of nearly six hundred feet along a rock-recess, fifty feet above the river-bed. The walls of some of the dwellings are still standing three stories high, and one of the houses is so perfectly preserved that its wooden roof, constructed of cedar poles, still remains intact. This Pueblo de Chelly is the most extensive cliff-settlement yet discovered in this region, and photographs of it could be only obtained in sections. Fig. 3 of plate vii represents the southern or right-hand end of the communal building, while figure 2 shows the northern extremity of the group. The structure consists of about seventy-six rooms or apartments on the ground floor, the plan of which may be seen in figure 1. Below the walls, seven large burial urns were unearthed, and while photographing the ruins, nearly a hundred beautifully fashioned arrow-points and several fine specimens of pottery were picked up. Across the channel of the stream, in the open valley, an extensive graveyard may be seen, in which the restingplaces of the departed have been marked out in the Pueblo manner by flat stones set on edge. In some instances the heads of the graves have been indicated by tall headstones.

In the preceding pages of this article I have simply attempted to give a general idea of the main features in the ancient Puéblo architecture. Many other equally interesting ruins might be described, but it is unnecessary to pursue the subject farther at present. These ruins are similar to those of the Chaco cañon to the south, in New Mexico, which were first brought to light by Lieut. Simpson in 1849. In writing of these latter, Mr. W. H. Jackson remarks: "In all the ruined pueblos, the most remarkable feature was the skill with which the stone walls were built, and which has enabled them to withstand, for hundreds of years, the ravages of human hands and the slower work of the elements. Beginning at the foundation with a width of thirty-two inches, each succeeding story was built a little less in thickness, until the walls of the fourth floor were about eighteen inches through, giving them a pyramidal shape, and of such solidity that in some cases, although the floor has been crushed down and the cross walls fallen, they yet remain firm and plumb nearly forty feet in height. They had three methods of laying the stones: by regular sandstone blocks of the size of two bricks, cut and ground uniformly; by alternate layers of these blocks with very small and thin pieces of sandstone, generally three courses of the thin to one of the thick; and last by laying the entire wall of these excessively small pieces of thin sandstone. As an example of this last kind, I measured off a square yard on the northern wall of the Pueblo of Chetho Kette, and counted the number of stones. forming the surface. There were 450, laid so closely together that a knife-blade could not be pushed between, and not a particle of mortar of any kind appeared at the surface. This entire wall was 400 feet long and originally fully forty feet high and averaged twenty-four inches thick. Imagine the industry and patience of such builders. Every doorway and window was framed with scrupulous exactness, and it would appear as if the plumb and square had been faithfully used in all their work."

In reviewing the above stated facts we would naturally be inclined to doubt the high antiquity of these works of art. But it must be remembered that the conditions were exceedingly favorable for the indefinite preservation of mural constructions. The equability of the atmosphere and the aridity of the soil must be taken into consideration. It is true that we find among these ruins traces of timber which, under ordinary circumstances, is

exceedingly perishable; but the wood which occurs is generally of the most durable sort, such as cedar and artemisia. The exsiccating properties of the desert atmosphere would be eminently conducive to its preservation, and it might remain intact for many centuries. On the other hand, we must not lose sight of the fact that in the majority of instances the wood-work of these structures has entirely disappeared. Only in exceptional cases do we find traces of vegetable fibre in the ancient remains. This fact alone would be sufficient to prove conclusively the great age of the buildings. In a country where scarcely a drop of rain falls from one year's end to another, and where the temperature varies but a few degrees throughout the 365 days, a great length of time would be required for the gradual disintegration of the most durable woods and particularly of solid rock. Yet there existed at one time thousands of stone and adobe structures throughout the San Juan valley, which can at present only be traced by inconsiderable mounds of dust. It is now generally held that the ancient Pueblos were a progressive branch of the so-called Mound-builders, forming a connecting link between the latter and the Nahuatl tribes of Mexico.

Frederick von Hellwald,' in an article on "The American Migration," says:

"If, then, it is permitted us to conclude from analogy, which is no uncertain guide when like causes operate under like conditions, the populations of the copper age of America; which had already dawned in the region of the lakes, would have followed the valleys of the Ohio and Mississippi, and then have directed their steps through the present States of Louisiana and Texas, probably along the edge of the gentle acclivity which, under the name of the Sierra Guadalupe, stretches from the Rio Grande to the Rio Brazos, towards the banks of the great Rio Grande del Norte.

"There are many indications, moreover, which lead us to believe that this was not the only route by which the northern tribes made their way to the south. A part of them seem to have become detached in the Mississippi valley, and to have projected themselves towards the south-east into Florida, the seat of a higher civilization, whence they eventually proceeded to Cuba and Yucatan; while a branch of them traversing the whole length of Cuba and the great arch of the Caribbean islands, descended finally to the banks of the Oronoco. This fraction of the migratory population may, of course, have 1 Smithsonian Report, 1866. Translated by C. A. Alexander.

been small and its impetus inconsiderable, since the necessity of maritime transport, though only from island to island, would naturally impair its force. In the opinions of others, and among them Humboldt, even the Rocky Mountains in their extension northward may have led similar branches of emigrants to adopt a different path in their progress towards the south. Whether these branches originally issued from the lake regions, though it is not impossible, is difficult to determine. They must at any rate, in departing from their homes, have taken a directly west or at least south-west direction. Although no substantial reasons can be assigned why any race of those latitudes should have given a preference to the toilsome defiles of the Rocky mountains, when the fair and commodious plains and prairies of the south lay before them, yet too many points of apparent connection present themselves to admit of our consigning their adoption of such a route to the category of impossibilities.

"It is from the Rio Gila that we are first enabled to perceive definite traces of the course of the migration into the regions of the south; the indications of the different stages of its progress increase with its entrance upon Mexican territory, but we yet possess only sparingly the means of identification. The first immigrants who appeared in the north of Mexico brought with them the so-called Toltecatl civilization, the work of the races of the great Nahoa family. * * *

*

"If we admit that the age of the civilization indicated in the region of the Mississippi reaches back 2,000 years, it is not impossible that the Nahoas were also the builders of the earthmounds in North America, or at least belonged to the race from which these works proceeded. As regards the stone structures of the great casas of El Zape and La Quemada, we cannot but infer that their builders must have been long permanently settled in those districts, which accords much better with later researches than the assumption of many that the immigrating tribes had merely halted in the places for a few years, perhaps a quarter of a century, and in that time had erected these monuments."

A. Marlot observes, in his “General Views on Archæology,” “It is finally worthy of remark that the mound-builders," as the Americans call the race of the copper-age, seem to have preceded and prepared the Mexican civilization, destroyed by the Spaniards; for in progressing southwards, a gradual transition is noticed from the ancient earthworks of the Mississippi valley to the more modern constructions of Mexico, as found by Cortez." (Translated by Philip Harry, Esq., for the Smithsonian Report for 1860.)

In concluding the subject, I wish to extend my thanks to Prof. F. V. Hayden, through whose courtesy the illustrations for these papers have been furnished.