MINDELEFF] USE OF FARMING SHELTERS 643 
than they took away. If they found anyone in the fields, they would 
beat him, or perhaps kill him, merely for the amusement it seemed to 
afford. It was the Navaho method of ‘‘ sowing wild oats.” There is 
little doubt that the pressure which bore on the Pueblos for at least 
some centuries was of this nature, annoying rather than actually dan- 
gerous. No doubt there were also occasional invasions of the country 
of more than usual magnitude, when from various causes the nomadic 
tribes had either an abundance or a scarcity of food, and, knowing the 
character of the villages as storehouses of corn and other products, or 
impelled by old grudges growing out of former forays, a whole tribe 
might take part in the incursion, and perhaps try themselves by an 
assault on some village of considerable size. But such expeditions 
were rare; the pueblo tribes were annoyed rather than menaced. 
Eventually, however, they found it necessary to provide against the 
ever-present contingency of an invasion of their country, and the great 
valley pueblos were developed. 
As aggregation of the little settlements into villages and of villages 
into great valley pueblos continued, the use of farming shelters grew 
apace. No matter what the conditions might be, the crops must be 
grown and harvested, for the failure of the crops meant the utter 
annihilation of the people. They had no other resource. They were 
compelled to combine into large pueblos containing often a thousand 
or fifteen hundred souls, a condition which was at variance with their 
requirements and manner of life; but they were also compelled to till 
the soil or starve. The lands about the home villages were never 
sufficient for the needs of the people, and in consequence a consider- 
able portion of the population was compelled to work fields more or 
less distant from them. Thus, in the ultimate stage of pueblo devel- 
opment the use of farming shelters was as much or more in evidence, 
and as much a necessity to the people, as in the prior stages. 
This sketch of the development of pueblo architecture exhibits a 
sequence; but it isa cultural, not a chronologic, one. The data in hand 
will not permit the determination of the latter now, but within a given 
group sequence in culture and sequence in time are practically synony- 
mous. The time relations of the various groups, one to another, must 
be determined from other evidence. 
The use of farming shelters has been a most important factor in 
producing the thousands of ruins which dot the mesas and canyons of 
the Southwest, while another factor, the localization of clans, has 
worked with it and directed it, as it were, in certain channels. All 
the evidence which investigation has revealed, from traditions to the 
intrinsic evidence of the ruins themselves, concur in establishing the 
fact that the pueblo tribes were in slow but essentially constant move- 
ment; that movement has continued down to the present time and is 
even now in progress. Viewed across long periods of time it might 
