April, '12J BALL: DRI\IXG .SPRAV 149 



end, it may not represent the average of the orchard. If in the center, 

 it may endanger the value of surrounding experiments. This sort 

 of check most nearly represents the value of spraying as against no 

 spraying, but cannot be used when the value of a given spray on a 

 given brood is to be measured, as it greatly magnifies the value of inef- 

 ficient spraying and of all second brood results. Under this system, 

 it is impossible to continue experiments on the same orchard through 

 a series of years, as there will be too great inequality in distribution. 

 2d, the single check, so-called, distributed throughout the orchard. 

 This method gives the most accurate indication of real orchard con- 

 ditions in the first brood possible, and by banding these checks and 

 thus destroying one-half to two-thirds of the first brood of worms, 

 part of the remainder will scatter so that these trees will represent 

 very nearly the condition of the orchard in the second brood. This 

 system will not show the value of spraying as compared with no spray- 

 ing. The scattering worms from the checks will tend to decrease the 

 real efficiency of the sprays on the second brood, while the few extra 

 worms these trees may carry more than a sprayed tree, will serve to 

 heighten the apparent value of the spray, so that the second brood 

 results will not be far from correct. This system also allows of the 

 continuous use of the same orchard. 



6th. Percentage of wormy fruit is a much less reliable basis of 

 comparison than worms per tree, though neither one is entirely accurate 

 and both should be given. Percentage of efficiency is the only method 

 of statement by which different experiments can be compared and 

 such efficiency can only be accurately measured by the single check 

 system. 



7th. That records founded upon picked fruit only are untrust- 

 worthy and practically valueless for scientific purposes. That postal 

 card canvasses, etc., to determine efficiency are absolutely worthless 

 and often misleading. 



8th. That, in all tests of efficiency, absolutely accurate account 

 should be kept of every fruit that sets on the tree and that the work 

 of the different broods of worms should be kept separate and compared. 



9th. That the number of broods and the relative importance of 

 each brood in the total injury will greatly influence the efficiency of 

 a given method or the importance of a given spray. 



Plan of the Experiment. Each orchard as a whole was given two 

 early sprayings, and one or two late ones, with the exception of the 

 experimental trees. Each orchard contained at least three varieties, 

 and on each variety two different spraj'ing compounds were tested, 

 each in two different strengths, making twelve duplicate tests in each 

 orchard of each spray. The averages presented are therefore based 



