In opposition to the hypothesis of local origin it might be stated that there is no evidence of forms intermediate in development. The oldest remains of pueblo architecture known are but little different from recent examples. But it must be borne in mind that pueblo architecture is of a very low order, so low that it hardly comes within a definition of architecture as an art, as opposed to a craft. Except for a few examples, some of which have already been mentioned, it was strictly utilitarian in character; the savage had certain needs to supply, and he supplied them in the easiest and most direct manner and with material immediately at hand. The whole pueblo country is covered with the remains of single rooms and groups of rooms, put up to meet some immediate necessity. Some of these may have been built centuries ago, some are only a few years or a few months old, yet the structures do not differ from one another; nor, on the other hand, does the similarity imply that the builder of the oldest example knew less or more than his descendant today--both utilized the material at hand and each accomplished his purpose in the easiest way. In both cases the result is so rude that no sound inference of sequence can be drawn from the study of individual examples, but in the study of large aggregations of rooms we find some clues. The aggregation of many single rooms into one great structure was produced by causes which have been discussed. It must not be forgotten that the unit of pueblo construction is the single room, even in the large, many-storied villages. This unit is often quite as rude in modern work as in ancient, and both modern and ancient examples are very close to the result which would be produced by any Indian tribe who came into the country and were left free to work out their own ideas. Starting with this unit the whole system of pueblo architecture is a natural product of the country in which it is found and the conditions of life known to have affected the people by whom it was practiced.