THE CLIMATIC THEORY OF TERRACES. 35 



could not have occurred without leaving evidences in the way of fault scarps and gorges. 

 CUmatic changes, on the contrary, are positively known to have taken place, for no one 

 doubts that when New York was covered with ice the climate of Arizona was different 

 from what it is to-day. Furthermore, climatic changes would act in exact agreement with 

 the topography. That is, if a mountain range were high and massive, and a change of 

 climate toward aridity were to kill a large amount of vegetation upon its sides, much 

 soil and gravel would be washed down and deposited at the base. A retm-n of moist con- 

 ditions, on the contrary, would lengthen the streams and cause channeling. Thus high, 

 steep mountains would be accompanied by lofty terraces. In the case of low, gently rounded 

 hills, on the other hand, aridity might cause the death of vegetation, but the soil would not 

 be washed away with anything like such great rapidity as upon mountains of steeper slope. 

 Thus no terraces, or only low ones, would be formed. Every part of a region would 

 be acted upon equally, without respect to the topography or to the direction or size of the 

 streams, and the effect would everywhere be of the same kind, yet the results would vary 

 in harmony with the topography. Altogether it seems as if the climatic theory fitted all 

 the facts so far as they are yet known, while the theory of earth movements meets obstacles 

 at every point. The matter still needs a vast amount of study, however, especially along 

 the lines of a careful mapping and measuring of the terraces of a few chosen regions.^ Much 

 light might also be obtained by a careful investigation of the many channels which have 

 been cut by various rivers since the opening of the Southwest to settlement by the white 

 man. 



THE CORRELATION OF TERRACES. 



If the theory of the climatic origin of terraces be accepted, we are at once confronted 

 by the problem of the correlation of those in various parts of the world. As yet it is too 

 early to attempt much along this line. Nevertheless, it is worthy of note that in number, 

 arrangement, degree of weathering, and other characteristics, there seems to be a fairly 

 close agreement between those of America and of Asia. Undoubtedly the number of 

 terraces varies considerably, but the variation apparently follows well-defined rules. In 

 general the number is greatest among lofty mountains in regions of pronounced aridity. 

 Where the relief diminishes, or where the climate is less arid, either because of greater 

 precipitation or because of lower temperature and less evaporation, the number of terraces 

 becomes less. In the districts of maximum development, such as the higher mountains 

 of southern Arizona or the lofty mountains of Central Asia, the common number is five. 

 Elsewhere it diminishes to two. Such a discrepancy does not mean a different periodicity 

 in the various regions. It simply indicates that from the geological point of view terraces 

 are ephemeral. In regions where the mountains are low the amount of deposition may be 

 so small that terraces of different ages appear as parts of a single formation, or the entire 

 body of some terraces may be carried away by erosion. In regions of only slight aridity 

 erosion may be so active during the erosive portion of a terrace cycle that all the deposits 

 are carried away. That such occurrences take place is proved by the fact that even 

 where the maximum number of terraces is found in some valleys, others in close proximity 

 show only two or three. Moreover, in a single valley some portions have five terraces, 

 and others only two, three, or four. One can often trace a terrace upstream and actually 

 see it disappear, either coming to an abrupt head, coalescing with an adjacent terrace, or 

 simply being lost by erosion. 



The alluvial terraces, both of Asia and America, are evidently due to a series of changes 

 of decreasing intensity. The first suggestion that presents itself is that they may repre- 

 sent the various epochs of the glacial period. This does not seem probable, however. In 

 the first place, the youngest terraces are far too new to have anything to do with the 

 last glacial epoch. In the second place, there is too much difference in size between the 



