CONTROL OF CODLING MOTH IN ARID REGIONS. 69 



METHODS ITSED. 



In each experiment the trees selected for the different tests have 

 been of the same variety, the same size, and bearing as nearly the 

 same number of apples as possible. In every test the orchard has 

 been sprayed twice and the si)raying done before the trees Avere 

 selected, except for those trees on which the given spraying was 

 omitted, so that the spraying represents the average condition of a 

 commercial orchard. Check trees have been scattered through the 

 blocks so that one check woidd not influence another. All imj)ortant 

 tests have been made in parallel series on at least three winter varie- 

 ties, and all have been or will be repeated through a series of years, 

 and in the warm and cold sections of the State. Unsprayed trees 

 have been left very sparingly, as one badly infested tree will notice- 

 ably influence half an acre. The check li-ees have been banded, the 

 bands examined every three days, and the worms placed in cages and 

 bred for the life history work. The remaining trees of the orchard 

 have not been banded, and the worms caught on the check trees do 

 not more than offset the larger number escaping from the unsprayed 

 checks, thus leaving the orchard in about the same condition as if 

 there were no check trees present. 



All apples that set on the check trees were accounted for. Windfalls 

 were picked up every three days and sorted into " sound," " calyx 

 wormy " and '' side wormy."" All that were the least questionable 

 were sliced, open to determine whether the calyx had been entered. 

 At the proper time to divide l)roods all the apjiles on the trees were 

 examined and all wormy ones of the first l)r()od recorded. These, 

 with the Avindfalls to date, made up the damage from the first brood 

 of worms. 



Complete records were kept on 28 trees in 1904, CA trees in 1905. 

 and no trees in 190('). The (54 trees used in the comparison of early 

 si)rayings in 190.'") cai'iMcd 2i!().0()0 apples, nearly all of which wei"e 

 handled three times (juring the investigation. 



COMPARISON OF RESISTS. 



In :dl tlie experiments tlie lesults have lieen given in the actual 

 number of wormy api>les found in each brood under each different 

 test, and the percentage given shows the actual killing powei" of the 

 spray for that brood. In one case, for example, the unsprayed trees 

 averaged in the first brood 1'2 wormy apples, while the twice-sprayed 

 ones averaged only 4 wormy, showing that (iS worms, or 94 per cent 

 of those that appeared, were killed. There can be no question of the 

 accuracy of this method for the first brood, because in this brood the 

 moths will l)e eveidy distril)uted througliout the block witliout refer- 

 ence to sprayed or unsprayed trees. In the second brood the problem 



