184 
JOURNAL OF ECONOMIC ENTOMOLOGY 
[Vol. 17 
orchards, and of measuring the homogeneity of the experimental areas is stressed. 
Ths usual methods of measuring variation are described and illustrated by experi¬ 
mental data, special attention being paid to the calculation of the probable error of a 
mean and standard deviation when the number of trees in a plat are less than 16. 
The influence of numbers on the confidence that can be placed in experimental plats 
is shown and the writer includes certain formulae that may be of use to investigators 
in entomology. 
The logic and mathematics of biometry is such a vast subject that, 
in the time allowed, it is possible to present only a few principles of 
elementary statistics as illustrated by some of the writer’s experiments 
for the control of the codling moth. These field tests were conducted, 
during 1920, in Niagara County New York. Incidentally, it might be 
mentioned that formulae commonly found in texts on statistics have 
been omitted but those that may be of use to codling moth workers 
and which are not commonly given have been added for convenience of 
reference. 
Biometry has three divisions: (1) exact measurement of biological 
phenomena, (2) classification of observations, and (3) refined mathemat¬ 
ical analysis of the data. With the exception of noting the use of 
double dichotomous and manifold classification in the arrangement of 
the data (Table 1), discussion of the first two divisions will be omitted. 
In order that mathematical analysis of experimental data may be 
precise and the conclusions applicable to a certain region, it is essential 
that the universe of investigation be defined; that the experimental 
areas be average samples selected from such a universe; and, that each 
series of plats be uniform, within the limits of normal variation, before 
any differences of treatment be made. The universe of investigation 
may be defined as a physiographical region in which the climate and 
soil is such that the growth habit of bearing orchards are similar for the 
same variety, and where the seasonal history and habits of the codling 
moth are fairly constant for the same reason. Owing to the expense 
involved, the number of field tests is limited, therefore, the experimental 
areas should be selected with the aim of securing conditions typical of 
the region. Since the data of experiments on one orchard may never be 
exactly duplicated in another, each planting may be considered as a 
random sample in a universe of experiments , hence, the importance of 
repetition of field tests before drawing definite conclusions for an ex¬ 
tended area. 
Homogeneity of Orchard Plats 
That plats upon which different treatments are to be tested may be 
comparable, it is fundamentally important that, previous to the field 
