THE ARGENTINE ANT IN RELATION TO CITRUS GROVES. 59 



resenting a reduction of 96.7 per cent by spraying. By May 27 the 

 leaves of the unsprayed trees had become quite badly infested with 

 white flies. A count of those on 100 leaves picked at random from 

 these trees gave 26,200 larvae and pupae, while on an equal number 

 from the sprayed trees there were only 73, a reduction, therefore, of 

 99.7 per cent. 



The rust mite began to appear on the fruit in June, and by the 

 23d of that month there were 50 to 60 mites per fruit on unsprayed 

 trees, while on the sprayed they could only be found on the row of 

 trees adjoining the unsprayed block and then only to the number 

 of 10 per fruit. 



On August 5, after the second insecticidal application, examination 

 of 100 fruits on the sprayed trees gave 987 scales, or an average of 

 about 9 per fruit, and 89 rust mites, the latter being so scarce that 

 they were difficult to find. The unsprayed fruits were so badly in- 

 fested that scales could be counted in the time available on only 10 

 fruits, on which there were 6,982, and the rust mites were quite too 

 numerous to count. Fully 75 per cent of the unsprayed fruit had 

 by that date become discolored by the rust mites. 



By September the trees in the experimental block had responded 

 beautifully to the treatment, and many persons commented on their 

 improved appearance. 



About ten times as much new growth occurred on these trees as 

 on the untreated trees. The fruit was larger, and a very large per- 

 centage of it entirely clean. The storm of September 28-29, however, 

 blew down and broke many of the trees and knocked approximately 

 87.2 per cent of the fruit to the ground, preventing bringing the 

 work to a completely satisfactory conclusion. 



It was possible, however, to count most of the fruit on the ground 

 and that on the trees and examine it for insect injuries. There were 

 practically twice as many fruits per producing tree left on the treated 

 as on the untreated trees. Owing to the morass of weeds in the 

 untreated block and to much of the fruit having been removed and 

 sold by the owner, it was impracticable to count the fallen fruit in 

 that block. In the treated block all the fruit which was not washed 

 out of the orchard was counted and examined for insect injury. 

 There were on the ground and on the trees 69,672 sweet oranges, 

 tangerines, and mandarins, which, averaging about 200 to the box, 

 made approximately 348 packed boxes of fruit. It was estimated 

 that nearly one-fourth of the fruit was not recovered. The pro- 

 duction was, therefore, about 435 boxes, or more than as much fruit 

 as the entire orchard had produced the previous year, as a result of 

 only one season's treatment. 



