LXXXVI 15; MATGESSON: 
secondly, and this is of the greatest importance, these stems are measured 
with different accuracy. A stem of the height of ten metres, for example, is 
measured only at half as many places as a stem of the height of twenty metres. 
The accuracy with which the form of two stems is settled is evidently propor- 
tionate to the square root of the number of measurements on the two stems. 
The taper-series now mentioned only illustrates the form of the stem inside 
the bark and above the diameter breast-high. Concerning the stem under 
breast-height, the investigation has given relatively small results, the measure- 
ments at this part of the stems being too few to allow a fairly careful study 
of the form. It looks, however, as though the distension by the root should 
be very closely correlated with the height of the stem and possibly with the 
diameter-quotient dp. : di. - 
As already mentioned, the stems are divided into form-classes by means of 
the absolute form-quotient, arrived at by dividing the diameter at the middle of 
stem above breast-height by the diameter breast-high. When trying to get a 
view of the variations of the stems within a stand according to the ab- 
solute form-quotient, and the variations of the mean absolute form-quotient 
for stands of different ages, and so on, this method is far from being suf- 
ficiently accurate. A single measurement of a diameter may possibly be in 
a high degree inaccurate, owing to accidental mistakes in measuring or to 
variations of the stem. That the single diameter-measurements really are subject 
to such errors, which are sometimes rather great, is most plainly to be seen by 
marking out the measured stem-curve graphically. These errors, however, are 
easily eliminated by reading off the diameters desired on the levelling curve.: 
This method, moreover, is combined with another advantage, the influence of 
the distension by the root being easily eliminated. 
The form-quotient thus obtained must be the most correct expression for 
the form of a stem, and is therefore suitable for use in close examinations of 
the form-variations within and between stands of different shape. 
Concerning the bark, the investigation is concentrated firstly upon the 
thickness of the bark at breast-height and secondly upon the taper of the 
bark with increasing height on stem. 
The thickness of the bark at breast-height varied within very wide limits, 
even in a wood uniform in age, density and other characters of the wood. 
Notwithstanding these great variations, a general type of the bark-curve of 
the different woods is to be observed, the bark-percentages decreasing with 
increasing diameter breast-high. This type of the bark-curve is to be found 
also in larch-groups with diameters breast-high up to one metre. 
The general thickness of the bark on stems from various woods is, how- 
ever, very different (compare, for example, the figures from sample-plot 281 
in table 12 with those from sample-plot 292). In trying to obtain a general 
average series for the bark breast-high, this fact must be taken into account, 
lest the run of the series should be quite erroneous. The calculation of the 
average series, therefore, was made as follows. Such a diameter-class was 
chosen as was to be found in all, or at least in the majority, of the woods 
measured. The value of the bark-percentage for this class was fixed graphi- 
cally, if necessary, and the values of the bark-percentages for all the other 
diameter-classes were divided by the former. Thus we obtain series of rela- 
tive bark-percentages breast-high (table 13); and from these an average series 
