380 JOURNAL OF ECONOMIC ENTOMOLOGY [V'ol. 4 



10 ft. 20 ft. 



Theoretical 1 oz. 



Leakage factor 4 oz. 2^ = 8 



^4 = 2 



Product = dose 4 



16 



Since the amount of leakage has precisely the same effect on the 

 leakage factor as the size of tent the formula becomes 



dose = d'''x P'Vl 

 in which p = per cent of leakage with .25 as the unit. 



Applying this formula tables were calculated for the dose correspond- 

 ing to the size of the tree and the degrees of leakage indicated by the 

 straight lines of the chart and the curves indicating the dosages were 

 drawn in accordance with these calculations. 



The table is based as already indicated upon the actual practice 

 of fumigators as regards the dose for a 10 ft. tree with average leak- 

 age. The portions of the table nearest this average therefore can be 

 looked upon as having an accuracy independent of the calculations. 



The most probable error is in the determination of 4 as the leakage 

 factor; 4 is right if 16 oz. is right for a 20 ft. tree but most fumigators 

 in fact give a 20 ft. tree less than 16 oz. which would indicate a larger 

 factor. In case the leakage factor should be more than 4, the result 

 would be the slight widening of all the curves, making still more pro- 

 nounced the difference between the dosage of the tight and leaky tents 

 of the same size. 



It seems safe to say that the importance of leakage is not less than 

 that indicated by the chart. 



Of recent years there has been a very general appreciation of the 

 necessity of measuring the sizes of trees when scheduling. There is 

 no doubt that grave mistakes are made where measuring is not done 

 but no matter how accurate one may be in the measuring of a tree, 

 this chart shows that the fumigator may make the dose only a quarter 

 as large as it should be or three or four times as great as is necessary 

 in case the tents are leaky or tight. No mistakes anywhere near as 

 large as these were found in all of the thousands of measurements 

 made by us to test the relation of dosage to size of tree. We are 

 led to conclude, therefore, that it is very much more important to 

 measure the leakage than to measure the size. 



