134 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



as we have already seen, fires would occur during dry seasons, and hence any diminution 

 of growth due to them would simply accentuate the diminution due to the drought. In 

 spite of this, however, most of the sinuosities in the curves are apparently not due to fires. 

 The reason for this conclusion is twofold. 



In the first place, ]iractically every one of our curves is based upon trees which did 

 not grow in one restricted locality, but were spread over a wide area. For instance, the 

 yellow pines of New Mexico came from forests all over the State; the white oaks came from 

 four or five localities in the States of Tennessee, West Virginia, and Kentucky; and the 

 giant redwoods were from four localities in the Sierra Nevadas from 5 to 60 miles apart. 

 It is far from probable that a single fire would affect all these trees at once, and in most 

 cases the number of trees from a given locality is not sufficient to produce more than a 

 slight effect upon the general curve unless the fire were very widespread. 



The second reason for believing that fires have had no great effect in producing the 

 sinuosities of the curves is stronger. A fire causes an immediate decrease in the tree's rate 

 of growth. It is followed by years of gradual recovery. Therefore in the case of a fire 

 the curve ought to drop suddenly and then rise gradually. Sometimes this occurs, but in 

 most cases the drop in the curves is not confined to a single decade but continues through 

 periods of from 20 to 50 years. In a large number of cases, also, the rise in the curve is 

 more sudden than the succeeding or preceding fall, indicating that the growth of the trees 

 received a somewhat sudden impulse, which was followed by a long period of gradual 

 decline. Manifestly this is exactly the opposite of what would occur in the case of a fire. 



Another important cause of differences in the rate of growth of trees is the amount of 

 sunlight or shade to which they are subjected. With young trees this is undoubtedly a 

 matter of extreme importance, and if our curves were based wholly on immature trees 

 it would render them almost valueless. As a matter of fact, however, the curves are based 

 on the largest trees, those which for centuries have been dominant. In practically all 

 cases the first 10 j'^ears of the hfe of the trees are not used in our curves, and in many cases 

 a larger number of years is omitted. Nevertheless in their youth the trees which we have 

 employed were doubtless shaded by other trees. Through the greater part of their lives, 

 however, they towered to full height and were not overcrowded. Moreover, even if 

 shading did prevent normal growth in a young tree, after the trees which overshadowed 

 such an individual had died, other trees could scarcely grow up so fast as to overshadow 

 it again. In other words, the shading of one tree by another might make the curve of 

 growth very low in youth and high in old age, but it could not cause it to fluctuate back 

 and forth from high to low. Therefore we must conclude that while shading is an important 

 factor in j'outh and may largely influence the beginnings of the curve of growth of each 

 tree it is not an important factor after maturity is reached. 



It is not easy to estimate the ravages of insects. Doubtless they, too, hke fires or 

 shading, often check the growth of the forests. The same arguments, however, apply to 

 them as to fires. Their ravages are apt to be local and would not be likely to influence 

 trees so widely scattered as those which we have used. IVIoreover, when trees are attacked 

 seriously by insects or other parasites the chances are that the trees so affected will die, 

 but the trees which have been used in our curves are for the most part uncommonly large 

 and elderly individuals which show little sign of disease. 



As to the effect of man, little need be said. Most of our curves are derived from 

 regions where man's influence has not been felt until within a few years. Even in long- 

 settled regions the sinuosities of the part of each curve belonging to a period two or three 

 centuries ago do not indicate that conditions were then essentially different from those 

 which now prevail after the coming of the white man. 



In discussing the reasons for thinking that the sinuosities in the curves are not of 

 accidental origin, we have been dealing with the matter negatively. There are, however, 



