CLIMATIC RECORDS IN THE TRUNKS OF TREES 



733 



accidents of competition, injury and so forth and that 

 other factors such as fires, drainage and sometimes pests, 

 which themselves depend on weather, actually exaggerate 

 climatic effects. Hence if in rigorous surroundings we 

 can show empirically a relation between tree growth and 

 terrestial or cosmic conditions, we are justified in re- 

 garding it as a genuine case of cause and effect. 



Long residence in the great yellow pine forest of 

 Northern Arizona led me to the study of that tree espe- 

 cially. In 1907 I had made and reduced ten thousand 



trees the cross-identification was more essential, for in 

 that region two causes operate to produce errors in ring 

 counting; first, the strongly marked double rainy season 

 (winter and summer) producing rarely an extra ring 

 which resembles the annual rings ; and second, the occa- 

 sional series of deficient years causing some trees, in part 

 of their growth at least, to stop ring production for one 

 or more years. The error there of straightaway count- 

 ing was found to average four per cent in the last two 

 hundred years. By cross identifying all rings this error 



(Figure i lower part.) 



COMPARISON BETWEEN 57 NORTH EUROPE PINE TREES (smoothed) AND SUNSPOT NUMBERS 

 The trees are froih England, Norway, Sweden and North Germany. 



(Figure 2 upper part.) 



measures upon 

 twenty -five 

 1 o n g-1 i V e d 

 trees. Four 

 years later 

 three or four 

 thousand verv 

 careful meas- 

 ures upon the 

 last fifty years 

 of nearly sev- 

 enty diflFerent 

 trees were add- 

 ed. And now I 

 have nine 

 thousand more 

 upon eighty different samples of the European Pinus 

 sylvestris or common pine of North Europe. The coni- 

 fers, by the great regions they cover, the great variety 

 of climates they endure, and especially by the promi- 

 nence of their rings, seem best adapted to this purpose. 

 Apart from care in measuring the rings, the details of 

 which have largely been described (Monthly Weather 

 Review, June, 1909, and Bull. Am. Geog. Soc, May, 1914, 

 Carnegie Publications, No. 192, Chapter XL), the most 

 fundamental and essential feature of the method is the 

 cross-identification of rings among a group of trees. The 

 ease and accuracy with which this can be done in a fairly 

 homogeneous forest is remarkable. A group of thirteen 

 tree sections collected along a distance of a quarter of a 

 mile in the forest of Eberswalde, near Berlin, show 

 almost identical records. Two to ten rings in every decade 

 had enough individuality to make them recognizable in 

 every tree. A group of twelve sections from Central 

 Sweden, cut, however, from logs at the sawmill at Gefle. 

 show such agreement that there is not a single question- 

 able ring in the last hundred years or more. Especially 

 marked combinations of rings could occasionally be 

 traced across Europe between the groups hereinaftei' 

 mentioned. In Arizona, identification across seventy 

 miles of country was unquestioned and even at two hun- 

 dred miles resemblance was apparent. But in Arizona 



90 Yr 



l25Yrs. 



was reduced to 

 half of one per 

 cent or perhaps 

 to zero. Re- 

 cently I have 

 made an inter- 

 esting check on 

 the accuracy of 

 the accepted 

 identificat i o n 

 by noting 

 every state- 

 ment of weath- 

 er, freshets or 

 crop failures 

 mentioned b y 

 the historian Bancroft in his accounts of the settlements 

 of Arizona and New Mexico. I find fourteen cases in 

 which the noted feature of the year agrees with the tree 

 record, and but one doubtful disagreement. The most 

 striking correspondences occur with reference to the flood 

 on the Rio Grande in 1680, the famines between 1680 

 and 1690, and the droughts in Arizona in 1748, 1780 and 

 1821. 



The accuracy with which the pine trees near Presr 

 cott, Arizona, represent the rainfall recorded in that city 

 for forty-three years, is, without correction, about sev- 

 enty per cent (Figure 1.) By a provisional correction 

 for conservation of moisture by the soil, this accuracy 

 rises to about eighty-two per cent. The nature of this 

 conservation correction is very simple, it is practically 

 the "accumulated moisture" of the meteorologists. It 

 signifies that the rings in these dry climate trees vary not 

 merely in proportion to the rainfall of the year but also 

 in proportion to the sum of the profits and losses of the 

 preceding years. The "credit balance" in their books 

 at the beginning of the year has only somewhat less im- 

 portance than the income during the current year. One 

 must remember that conservation in this dry climate may 

 be very different in its action from that in wet climates 

 where the ground is continually moist or water-soaked. 

 In reckoning the amount of moisture for the year 



