150 



THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



From this it appears that if we would obtain correct results from our measm'ements of 

 trees, it is necessary to make a slight addition in those cases where only one reading is 

 available. The amount of this addition can be seen in tables C and D, on pages 302-307. 

 It varies from a negligible quantity with the youngest trees to as much as four or five decades 

 with the oldest trees. In making the computations upon which our final tree curve has been 

 based, additions of this sort have been made — that is, in cases where only a single measure- 

 ment of a tree was available there has been added a sufficient number of decades to bring 

 the age up to the maximum, according to the assumption just set forth. The extra decades 

 have been inserted at equal intervals from beginning to end of the tree's hfe. They have 

 in every case been given the mean value of the decades on either side of them. 



In addition to this, decades have been added to certain measurements for another 

 leason. Where two or more measm'ements of a single tree are available I have assumed 

 that the maximum reading is nearly correct, and have added enough decades to the other 

 reading or readings to bring them to the same age. The decades have been distributed evenly 

 from beginning to end of the tree's life, as in the other case. "Wliere the difference between 

 the readings amounts to more than 2 per cent, only the maximum reading has been used. 



Fig. 35. — Sequoia washingloniana : Corrective Factor for Age during first 250 Years of Life, 

 plotted by Decades. (See Table A, p. 301.) 



The object of all these changes is to give each tree its correct age as nearlj' as possible 

 and to prevent the curve from flattening out. Suppose that a period of rapid growth took 

 place in a tree 2,000 years ago and lasted two or three decades ; if one measurement of this 

 tree gives it an age of 2,500 years and another measurement gives only 2,450, or 2 per cent 

 less, it is clear that the period of rapid growth 2,000 years ago, as indicated by the two radii, 

 will come earlier in one case than in the other. The result will be that when the two are 

 averaged it will not be so evident as it ought, for by being spread over a long interval it 

 wiU be reduced by half its value, and will, in turn, raise places which ought to be low. 



From what has been said in regard to this correction for missed rings, it may appear as 

 if by applying it we should introduce large changes into our final curve; but as a matter 

 of fact the total number of decades added to all the 785 measurements of our 451 trees 

 amounts to only 405. The total number of decades indicated by all the measurements of 

 these trees is 111,700, so that the total change in the cm've amounts to less than 0.4 per cent. 

 The maximum effect is, of course, produced in the oldest trees. The part of the curve 

 less than 1,000 years of age is practically unaffected. From that point backward, the effect 

 increases, but even at 2,000 years of age it amounts to only 8 years; a maximum which, 

 according to the uncorrected curve, would occur at 100 b. c, according to the corrected 

 ciu-ve occurs at 108 b. c. 



The corrections for age and longevity are very different from those that we have just 

 been considering. Instead of producing a small and almost unnoticeable effect, they 

 greatly change the form of the curve. The methods used in obtaining them have already 



