146 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



greater, 8.2 years on the average, and he gets slightly less rings than they, as is shown by 

 the fact that his average divergence when plus, minus, and zero years are added, comes 

 out a minus quantity. In the case of Z, greater carefulness than with either of the others 

 is indicated by the fact that on the average his recounts are 0.25 per cent more than 

 he and his companions obtained in the original counts. In all cases the errors are so 

 small that they can be neglected without serious consequences. 



So far as mere errors of counting are concerned, it is probable that in the long run 

 where a number of observers are concerned they balance one another; but there is still a 

 certain degree of error on account of the fact that certain rings are so small that thej' 

 almost coalesce and are not differentiated, but counted as one. This is evident from the 

 fact that in general the more carefully a radius is prepared and the more minutely it is 

 scrutinized with a lens, the greater the number of rings. As a partial offset to this may 

 be put the fact that occasionally two rings are formed in a single year. Such cases are rare, 

 however — far more rare than in the yellow pines examined by Professor Douglass^and 

 it is almost invariably easj- to distinguish them by the lack of firm, hard fiber in the outer 

 portion of the extra ring, that is, the part which grows late in the season. On the whole 

 we may conclude that, so far as purely human errors of counting are concerned, individual 

 cases of very bad trees may occasionally run as high as an error of 5 or 6 per cent, but on 

 the whole they are less than 0.333 per cent, and tend to balance one another, even with a 

 single observer, and still more where several are concerned; yet there remains a certain 

 error, sUght, but constant, due to the fact that rings which actually exist are not counted. 

 We have no means as yet of knowing how great this is, but it must be less than 1 per cent 

 and probably not over 0.1 per cent, for in many trees all the rings are so large that one 

 can not possibly fail to see them. 



The next source of error, the actual absence of rings, is a much more serious matter 

 than the errors of counting. When radii on different sides of a tree are counted they are 

 in many cases found to be unequal. The table of recounts on page 145 show's that in the 

 23 trees there recorded the differences between the fii'st and second readings range all the 

 way from zero to 141, with an average of 19.1, or 1.3 per cent of the average age. In one 

 ( xceptionally bad case a radius on one side of an old tree when first counted gave an age 

 of 3,067, which when recounted was reduced to 2,996, while a radius nearly at right angles 

 to the first gave an age of 2,587, which a recount reduced to 2,526. Such a divergence 

 of 480 between the pair of readings made at the first attempt and 470 between the second 

 pair is most unusual. It indicates great irregularity, and this u-regularity, in turn, partially 

 explains the large differences of 2 or 3 per cent between the original counts and the recounts, 

 differences also due partly to the pecuUar way in which the stump had been cut into the 

 form of a rough cone and had since decayed to an uncommon degree. An inspection of 

 the tables at the end of the volume shows that differences of 1 to 5 per cent between different 

 sides of the same tree are not uncommon. Fortunately they are far from being the rule, 

 since more than half of the trees show apparent differences of less than 1 per cent, which 

 may be due in many cases to errors in counting. In the cases where there is a genuine 

 difference it appears to be due to the fact that during bad years one side of a tree may not 

 make any growth. This may be due simply to the fact that that side is shaded by other 

 trees or is prevented from getting much nourishment because of the presence of great 

 numbers of roots of other plants; or it may be due to injuries such as the breaking off of 

 branches or the ravages of fire. 



Plate 6 illustrates an extreme phase of the matter. WTien the young sequoia there 

 shown was about 23 years old the portion about an inch to the right of the letter A ceased 

 to grow for thi-ee or four years, while the opposite side grew as rapidly as ever. Then the 

 stunted side revived and made a good growth during the last three years before the tree 

 was cut. If the tree had continued to live it would doubtless have put on wood on all 



