LOGARITHMIC CROSS-SECTION PAPER IN FOREST MENSURATION 341 



except at their lower ends. The reason for this is obscure, but that 

 it is not entirely an accident is indicated .by the fact that a number 

 of other tables which I have had occasion to plot have shown a similar 

 tendency, although not always in such a marked degree. This fact is 

 a two- fold advantage. First, the curves are easier to locate and 

 draw, and second, if extensions are necessary, they can be more 

 safely made. It may be objected that the diameter class curves of 

 figure 1-B present the same peculiarity, and hence the same advan- 

 tage. This is true in this particular case, but straight lines in the 

 diameter class curves are of much less benefit. It is decidedly 

 preferable to draw the height class curves first, since in this way 

 the data are concentrated on the smaller number of curves, which 

 will be more strongly located. It is in the location of these first 

 curves that judgment has the greatest play, that the highest skill is 

 required, and hence that the help of the straight line form will be 

 most valuable. 



There are many other cases where logarithmic paper will be found 

 helpful. In growth problems, however, something a little different is 

 demanded. Time is a variable which differs from such dimensions as 

 height or diameter, in that it is determined, in forestry at least, either 

 historically or by a process of counting instead of measurement. Long 

 periods of time can be determined to a far greater percentage degree 

 of accuracy than short. It seems illogical, therefore, to plot ages in 

 such a way that the second century occupies no more space than 

 the second decade. This situation suggests the use of a cross-section 

 paper in which the vertical graduations are logarithmic and the 

 horizontal are evenly spaced. Such paper can, as a matter of fact, 

 be obtained under the trade name of "arithlog" paper. On this paper, 

 as in the case of the logarithmic, certain types of equations appear as 

 straight lines. In these, as before, the forester is but little interested. 

 The only exception is that the compound interest formula 



A = C(1.0/^n) 



and certain allied equations, conies into this class, since it can be 

 expressed logarithmically as 



log A =\og C -\- n X log 1.0^ 



If ^ = n are the variables, this is a first degree equation which can 

 be plotted as a straight line. Tt \\ ill be noted that while A apjjears as 



