INSTRUMENTS AND TECHNIQUE 41 



done. So the practical method of standardizing or equalizing trees, 

 which has been used extensively for actual curve production (commonly 

 modified as in the next paragraph), is to divide individual values by 

 the mean value of the tree, so that the annual values of each tree will 

 enter the group table as percentage departures from its own mean. 

 Simple averages are then taken for each year in the group. This 

 avoids some of the exaggerated effect of extreme departures. It 

 places all the trees on an equality, but does not place all departures on 

 an equality. It is averaging by weight, in which the weight is inversely 

 proportional to the mean growth of the tree. 



Age correction — Young trees have to develop the trunk rapidly 

 in order to stand the strain of wind and snow. Hence the early rings 

 are larger and somewhat less sensitive to climatic effects. When the 

 tree curve is plotted, it usually rises at the early end, sometimes very 

 rapidly. A reduction to percentage departures does not correct this. 

 One can correct it by getting percentage departures from a type curve 

 developed mathematically, as Huntington did (1914), but it can be 

 done far more rapidly and with sufficient accuracy by drawing a curved 

 or broken standardizing line on the individual plot and getting the 

 percentage departures from this line. Such a line is usually straight 

 and horizontal for a large part of the record and slants upward at the 

 early end. A curve is more accurate than a broken fine, but there is 

 little real difference and the broken line is more easily described if it is 

 necessary to state its position in words. 



Other corrections — Huntington used a " flaring" correction for the 

 increased measured width of outer rings near the base of the big trees, 

 where the spread of the root system is felt and a horizontal measure- 

 ment is not perpendicular to the rings. Evidently, in drawing a 

 standardizing line this can be taken care of. It is evident that in 

 studies of cycles not exceeding half a century or so in length the flaring 

 effect is negligible. But in estimates of very long periods or of secular 

 values, this effect must be nicely gauged. 



It is much the same with his " longevity" effect. This effect simply 

 recognizes that a slow-growing tree has a different normal age-curve 

 from a quick-growing tree. The slow grower more quickly reaches 

 the normal slow growth. This, too, is important in getting early abso- 

 lute ring values, but plays little part in studies of periodic variation. 



Comment on standardizing — It is felt that standardizing serves 

 two purposes: first, correction for age, injury and flare, and second, it 

 compensates for few numbers in a group, so that 5 or 10 trees will give 

 practically the same results as 25 or 100. It is not thought important 

 to use it if the number of trees used in a group average is over 15 and 

 the age variations are small. 



