Part III.] 
Beeson : Beehole Borer of Tealc. 
61 
PART IV. 
Statistical Methods. 
The general principles on which the collection of statistical data 
on the incidence of the beehole borer is based, have been evoived gradu¬ 
ally during the past two seasons’ work. Without going into details 
of rejections and failures the following discussion of the method 
adopted, is presented and it is hoped that it will be of use in future 
investigations. 
Sample Plots. 
In order to compare the relative degrees of attack in different areas 
of teak forest it is necessary to reduce observations to a common basis. 
The chief variables are (a) the age of the stand, (6) the number of sample 
trees taken, and (c) the distribution of the sample trees in the girth 
classes of the stand sampled. The most convenient means of com¬ 
parison is found to be the mean sample tree, both for the total beehole 
incidence and for the annual incidence of attack. Firstly, the stand 
to be sampled should be enumerated in 3 inch girth-classes to deter¬ 
mine its composition, and, secondly, the sample trees selected for 
analysis should be distributed proportionately throughout the girth- 
classes, so that the arithmetic mean sample tree of either series of 
measurements is approximately the same. The minimum number of 
trees for analysis, that will give a fairly close indication of the 
incidence, appears to be 10* for crops over about 20 years of age. 
In stands of less age, and in localities where the abundance of 
beeholes is low, a higher minimum number is required. 
Analysis. 
The sample trees are girthed at 4' 6", and sawn up in lengths of 9 
inches from soil-level upwards. This length is found to be the most 
convenient for splitting ; in small poles double billets of 18" may be 
used. The billets are serially numbered from the base upwards, 
quartered and split into small wedges until every trace of the work 
# In the Yanaungmyin Sample Plot, the incidences of trees in groups of 5 selected 
at random from the total number of sample trees (67) was compared, and it was found that 
as far as the annual incidence was concerned 5 trees gave sufficiently accurate results 
but with regard to the girth-beehole value the results were not uniform. 
[ 279 ] f 2 
