238 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



In both cases the number and intensity of cyclones increase and decrease in harmony 



with those of sun-spots, the greatest number occurring at the time of 



maximum spots, which appears also to be the time of minimum tern- Table u. 



perature. Wolf has compared the frequency of the cyclones with the 



number of sun-spots and gets the interesting result seen in table 14. 



Here we seem to have an unmistakable relationship; but when 

 other climatic elements are investigated, for instance, thunderstorms, 

 hail, movements of glaciers and the like, and especially when regions 

 far from the equator are considered, the indications of a sun-spot 

 cycle become so weak and conflicting that no true causal connection 

 has yet been established.* 



In view of the importance and complexity of this subject it seems advisable to test it 

 by means of our measurements of the growth of trees. This is especially desirable because, 

 although the trees in some respects fail to give a perfect record, their record has the great 

 advantage of being long and homogeneous. Professor Douglass has already made a be- 

 ginning in this subject. From the trees of Arizona he has found an apparent agreement 

 between the amount of growth and the sun-spot cycle. The average 11-year cycle for a 

 long period seems to show a double maximum in the growth of trees. Inasmuch, however, 

 as an arbitrary period of 11.4 years was used and no attention was paid to the fact that the 

 actual sun-spot cycle may range from 7 to 16 years, it is questionable whether much rehance 

 can be placed on these results. 



The other method of Professor Douglass, on the contrary, is reliable and conclusive. 

 In figure 25, page 120, he has simply compared the growth of 13 trees in Germany since 

 1820 with the standard sun-spot curve. The result is striking. Each curve shows seven 

 major maxima, and of these seven all but one occur at essentially the same time in both 

 curves. In the one exceptional case, the trees reach a maximum in 1901, while the sun- 

 spots do not reach their highest point till four years later. In view of the accidents to which 

 trees are liable and the extent to which the failure of the rains in one or two critical months 

 may check growth for several years, one single disagreement out of seven possible agree- 

 ments is no more than we should expect. The apparent relation between sun-spots and 

 tree growth thus found in Germany is especially significant because the German curves 

 of temperature and rainfall do not agree so closely with the sun-spots. This seems to 

 indicate that when temperature, total rainfall, and seasonal distribution of lainfall are all 

 integrated, as they are by the growth of the trees, the sun-spot cycle is more evident than 

 when individual climatic elements are concerned. This appears to he the reverse of what 

 is true in equatorial regions.* 



A test of the trees of California in this same fashion fails to show any such agreement 

 with sun-spots as is found in Germany. In the two curves of figure 75 a resemblance may 

 be traced at certain points, but it soon gives way to disagreement and appears to be purely 

 accidental. So far as these particular curves are concerned there seems to be no warrant 

 for beheving in any connection between solar changes and climate. The contrast in this 

 respect between the German and Cahfornia curves is a good illustration of the complexities 

 and apparent contradictions of this involved subject. We shall later inquire whether 

 such apparent inconsistency is incompatible with a solar theory of climate or is its ex- 

 pectable consequence. Meanwhile let us investigate the same subject by another method. 



The method here employed is simply to determine the average rate of growth at different 

 portions of the actual sun-spot cycle, and see whether it varies in harmony with the cycle. 

 Beginning with 1610 A. d., when the dates of maximum and minimum sun-spots first begin 

 to be known with certaintj^, each sun-spot cycle has been taken by itself and divided into 

 parts according to table 16. 



* See notes on pp. 205 and 253. 



