116 



THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



these are enough to give approximately correct results, although greater accuracy is of 

 coui'se highly desirable. In order to test the degree of accuracy to be obtained from a 

 small number of trees, a comparison was made between large groups and small. After the 

 entire group of rings in each section, some 6,300 in all, had been identified and numbered, 

 the sections were tabulated in order of age, with the oldest first. They were then separated 

 into groups of five, and in a convenient mamier averages were obtained of the oldest five, 

 going back about 400 j-ears; the oldest ten, 350 years; the oldest fifteen, 300 years, 

 and the entire nineteen reaching back only 200 years. Finallj-, at the ancient end of the 

 oldest five, the oldest two were carried back to fully 500 years. On plotting the groups 



Fig. 19. — Annual Growth of Trees at Flagstaff since 1385 a. d. 



of fifteen, ten, and five with its extension of two, it became immediately evident that five 

 trees gave almost the same growth as fifteen, even to small details. So in the work dis- 

 cussed below, the five are used to give the record from 1503 to 1908; so also for the same 

 reason a comparison was made between these five, and the two oldest taken by themselves. 

 In this the agreement was not quite so perfect, yet was so close that errors thus introduced 

 will not at all affect the curves referred to below. However, the two oldest were very 

 slow growers, and 5 mni. were added to all their records where only these two were used, 

 in order to make their curve continuous with that of the whole five. Thus the tree record 

 is made to extend from 1411 to 1908, as is shown in figure 19. Unfortunately, it must not 

 be taken for granted that this remarkable agreement between very small groups of trees 

 is true necessarily for other trees, or even for this yellow pine tree under all conditions. 

 It is without doubt due to the fact that this tree under semi-arid conditions is extremely 

 sensitive to varying moisture supplj\ 



This extreme sensitiveness causes one fault in the record, namely, the frequent omission 

 of rings between 1891 and 1896, as is evident in the list already given; the complete omission 

 of a ring is an exaggeration which should be guarded against. Accordingly, these years 

 were speciall}^ investigated in both rapid-growing and slow-growing trees, and a series of 

 growth values estimated from the trees which did not omit the rings. These interpolated 

 values have been used in the figure. 



As has already been said, a correction is needed to offset the faster growth of trees in 

 youth than in old age. This has been made empirically by drawing a long, nearlj' straight 

 line throughout the plotted curve for 500 years. The slope in this line shows very closely 



