April, ’24] 
hartzell: dosage estimation 
281 
the amount of time required for counting the thousands of dead psylla 
and for re-dusting the “count” trees to determine the survivors of the 
original application. This gave the data from which to calculate the 
percentage efficiency of the operation. In addition, generally, the 
following data were secured on each count tree: size and shape (the 
latter by means of photographs), planting distances, amount of dust 
applied, and temperature. A total of 46 trees, treated with 2 per cent 
lime-nicotine dust of practically the same toxicity, had simultaneous 
measurements of the following variables: (1) Per cent efficiency of the 
operation; (2) dosage, (3) temperature, and (4) percentage of open 
space. Some of the statistical constants, calculated from the data, are 
given in Table 1. 
Table 1. Means, Standard Deviations, Coefficients of Variability, and 
Range of Variables in Pear Psylla Experiments 
Variable 
Mean 
Standard 
Coefficient of 
Range 
Deviation 
variability 
Percentage of effi- 
ciency of opera¬ 
tion . 
72.78 ± 1.69 
17.03 ± 1.20 
23.40 ±2.45 
32.8—99.0 
Dosage on basis of 
tree volume (cu. 
ft.). 
2931.22 ±122.67 
1233.46± 86.74 
42.08 ±4.87 
1218—5359 
Dosage on basis of 
tree space (cu. 
ft.). 
4529.78 ±266.78 
2682.51 ±188.63 
59.22 ±5.91 
1688—11000 
Temperature Fahr. 
Percentage of open 
71.44± 0.69 
6.93 ± 0.49 
9.70 ±0.97 
61—84 
space. 
28.59 ± 1.70 
17.07 ± 1.20 
59.71 ±5.96 
0—60.4 
Correlation 
The first three problems presented are the following: (1) effect of 
dosage on efficiency, (2) the influence of temperature on efficiency, 
(3) the relation of open space to control. The simplest method of 
measuring the effect of one variable upon another is to determine the 
correlation coefficient (r), the variables involved being indicated by 
subscripts. These are known as zero order coefficients and mean that 
the relation between two variables has been measured without regarding 
the influence of other variables. The mathematics of correlation will 
not be presented because they are explained in detail in a number of 
elementary texts on statistics. The subscripts denoting the several 
variables in the investigation are as follows: 
