It is by this supposition only that the comparatively modern form of the square and terraced pueblos built contiguously to the latest abandoned of the cliff towns may be explained. For when the cliff dwellers had become numerous enough to be able to maintain themselves to some extent out on the open plains, it has been seen that they did not consider their villages safe and convenient or quite right unless builded strictly, in both general form and the relative arrangement of parts, as had been for many generations their towns in the cliffs--did not, it is reasonable to suppose, know at once how to build villages of any other form. Thus we may confidently regard these round towns as the earliest built by the cliff dwellers after they first left the cliffs. The direction in which these cavoid or cliff-form or rounded village ruins may be farthest and most abundantly traced, is, as has been said before, to the southward into and through the land of Zuñi as far as the cliffless valleys bordering the Rito Quemado region in southerly central New Mexico, wherein lies the inexhaustible Lake of Salt, which the early Spanish chronicles mention as the possession and source of supply of the "salt in kernels" of the Zuñi-Cibolans. Not only did a trail (used for such long ages that I have found it brokenly traceable for hundreds of miles) lead down from the cliff-town country to this broad valley of the Lake of Salt, but also there have been found in nearly all the cliff dwellings of the Mancos and San Juan section, whence this trail descends, salt in the characteristic kernels and colors found in this same source of the Zuñi supply. This salt, as occurring in the cliff ruins, is commonly discovered wrapped in receptacles of corn husk, neatly tied into a trough-like form or pouch by bands of corn-leaf or yucca fiber. These pouches are precisely like the "wraps of the ancients," or packs of corn husk in which the sacred salt is ceremoniously brought home in advance of the cargoes of common salt by the Zuñi priests on each occasion of their annual, and especially of their greater quadrennial, pilgrimages (in June, after the planting) to the Lake of Salt. And it is not difficult to believe that both the packs and the pilgrimages--which latter offer many suggestive features not to be considered here--are survivals of the time when the remoter cliff-dwelling ancestry of the Zuñi Corn tribes ventured once in a period of years to go forth, in parties large enough for mutual protection, to the far-off source of their supply of salt. Except this view be taken it is difficult to conceive why the "time after planting" should have become so established by the Zuñis (who are but two days' foot-journey from the lake, and visit its neighborhood at other periods of the year on hunting and other excursions) as the only period for the taking of the salt--to take which, indeed, by them or others at any other season, is held to be dire sacrilege.