in front of the pile of coins, and is supposed to clip the tang-ge' ro from under the pile, carry it some inches or some feet distant, and come to rest when and where it may happen.

Such is Philippine tang'-ga a game developing a low order of skill which seems valueless in any worthier pursuit. As a game for healthful Filipino amusement it answers its purpose, in that it has enough of chance to arouse the player's interest, and yet such interest manifests itself in so unexplosive a way that it is not trying or exhausting. The game seems in every way the legitimate child of a people of the tropics; its natural habitat is the shady side of a building blistering in the southern sun.

ELROY,

WISCONSIN.

THE SUN'S INFLUENCE ON THE FORM

OF HOPI PUEBLOS1

By J. WALTER FEWKES

In a valuable memoir on Pueblo architecture2 Mr Victor Mindeleff first called attention to the arrangement of houses in Hopi pueblos in parallel rows separated by courts or plazas. He might have added that these rows of houses, as their parallelism infers, are oriented in a uniform direction without regard to the configuration of the mesa on which they are situated. The object of this article is to suggest the cause of this uniform arrangement and orientation, and to discuss its influence on clan localization. I shall also consider historic modifications of pueblos of this tribe produced by the accession of new clans as recounted in legends. My discussion will be limited to the villages on the East Mesa, called Walpi, Sichomovi, and Hano, restricting the consideration to the time they have stood on their present sites, or, roughly speaking, to the last three centuries. The thesis that I seek to defend is as follows: The arrangement and orientation of houses in Hopi pueblos are largely due to an attempt to secure sunny exposures for entrances and terraces and consequent protection from cold and wind. The facts discussed, like many others before the ethnogeographer, illustrate the influence of climatic or environmental conditions on human culture development.

It may lead to a better understanding of the discussion if I point out in the beginning that a modification has taken place in certain architectural features of Hopi houses since contact with Europeans. Little change has occurred in the forms of the buildings or the mode of construction of their walls-indeed the pueblo has profoundly influenced the Mexican house builder in that particular- but the relative position of entrances, especially those of the lower story, is

1 Read before the Association of Geographers, New York, December, 1905.

2 Eighth Annual Report of the Bureau of Ethnology, Washington, 1891.

radically different in old and new Hopi houses. An examination of ruins that antedate the arrival of Europeans shows a total absence of doorways in the walls of basal rooms, the entrances being universally hatchways or openings in the roof to which one mounted by a ladder. The chamber of the second story, however, was entered through the side wall from the roof of the lowest story. In both old and new houses lateral entrances are essential features of higher rooms, an arrangement that imparts to a four-storied pueblo, like Walpi, a terraced form. The roofs or terraces of a pueblo habitation are customarily used by the inhabitants in daily occupations almost as much as the rooms themselves, for exposure to the sun and protection from cold winds are especially desirable in these places.

These roof terraces and lateral doorways, ancient and modern, as a rule are situated on the same side of the houses: their orientation is generally south or east or somewhere between these two directions. The axis of a row of such houses is naturally at right angles to this orientation, or approximately north by south. Let us analyze the probable cause that has led to the union of houses in rows and aim to discover the origin of their growth in this direction rather than in any other.

Starting with a single habitation housing one clan or family as a nucleus, suppose that this clan by marriage of eligible daughters is rapidly increasing and the maternal house is not ample to accommodate the increased family. In other words, the family has outgrown its original house, and it has become necessary to build new rooms to the old rather than to construct new dwellings. These rooms may be built on the sides, or, if space allows, on the roof of the mother's house. It is evident that there are limitations to the capacity of the roof, and additions to two of the four sides are undesirable for the reason that a room constructed on the east or south walls would exclude the sun with its warmth from the maternal house, while one on the opposite (west or north) sides would be equally undesirable, as the sun would be shut out, thus exposing the dwelling to the cold. The northwest and southeast walls are advantageous for additions to the parental abode, since they permit the new habitations to have heliotropic exposures without interfering

with that of buildings already standing. Similar restrictions also governed the addition of subsequent rooms made necessary by family increase, these increments always tending to enlarge the row in a northwest or southeast direction and to restrict growth at right angles to this axis. As time went on the topography of the mesa may have necessitated a new site and another house nucleus. Commonly this happened when new clans joined the pueblo. Each incoming family was assigned a site for its dwelling, but this site seldom adjoined houses already standing. After this addition had erected its first house the law of heliotropism regulating the position of terraces and growth of rows of houses became operative, eventually leading to parallelism in the rows of rooms already existing. It will thus be seen that the arrangement of houses in rows extending north and south, or approximately in these directions, was not fortuitous but was due to the position of the sun and to human effort in obtaining a heliotropic exposure for the maximum number of terraces. It is instructive to consider the bearing of legends and the localization of clans in the modern pueblos on this theory, and for this purpose we will begin our studies with the Tewa pueblo of Hano.

GROWTH OF HANO

Hano was founded on its present site about the beginning of the eighteenth century by Tewa clans from Chewadi, a pueblo in the Rio Grande valley of New Mexico. Legends gathered from the present inhabitants declare that Hano (pl. x1) was developed from three originally independent building centers or nuclei of growth that later grew together by natural extension. These centers of growth may be designated, from the clans that built and first inhabited them, the Tobacco-Corn, the Cloud-Sand, and the Katcina houses. In the beginning the Tobacco-Corn habitation housed three clans in as many rooms placed side by side with terraces facing eastward. These rooms were inhabited by the Tobacco, Corn, and Bear families. The form of this triple-room house was changed in the first or second generation by an unequal growth of these three clans. The Tobacco and Corn families were vigorous, increasing rapidly in numbers, while the Bear people remained stationary or declined. The present localization of rooms in this