13 



sible variations of this sort are so numerous, and a precise exam- 

 ination of their effects upon the remedy would require such an 

 interminable series of laborious and costly experiments, that I 

 have thought it best to leave this subject to the practical fruit 

 grower, for determination on the larger scale of orchard practice. 



It is also especially to be noted that the final summaries of in- 

 jury are much greater in our experiments than they would have 

 been if all the trees in the orchard had been treated, as in the 

 practical application of this method in the field. Since we pur- 

 posely left as many trees untouched as we sprayed, and since in 

 our experiments of this year the trees under observation were 

 surrounded by other trees heavily laden with apples, to which we 

 gave no attention, there were necessarily bred in these trees an 

 abundance of codling moths of the first brood, which, spreading to 

 our experimental trees, must have greatly increased the ratio of 

 injury by the later broods. 



In orchard use, however, where all the trees would be poisoned, 

 the later broods could re23resent only the small percentage of ti e 

 last generation which had escaped the treatment. 



TIME OF APPLICATION. 



We have next to determine the time of the year at which poi- 

 soning is the most effective; whether, in fact, it takes principal 

 effect upon the first brood or the later ones. A moment's reflection 

 will show that if only the first brood of the larvae was directly 

 diminished in a certain ratio, the second brood should show a 

 similarly diminished ratio, since these descend from the first; 

 whereas if both first and second broods are directly poisoned, then 

 the ratio of damage to the second brood should be greater than 

 that to the first; or, in other words, the percentage of benefit to 

 the picked apples should be greater than that to the fallen. Our 

 data for the present season do not apply to this question, since 

 all the sprayings were made in May and early in June, while the 

 apples were still very small; and it is incredible that the poison 

 should have remained upon the fruit through all the vicissitudes 

 of weather and time for the two months and more that elapsed 

 before the appearance of the second brood of larvse. The results 

 of last year, however, when the spraying was continued until Sep- 

 tember 3, give us^ important information. Kef erring to Table I., 

 it will be seen that the benefit to the picked apples instead of 

 being greater than that to the fallen fruit was last year about 

 twenty-seven per cent, less, so that certainly no appreciable effect 

 was produced by spraying during the life of the second brood. 

 The greater injury to the picked fruit is readily accounted for by 

 a circumstance to which I have already alluded; viz., that our 'ex- 

 perimental trees were surrounded by others upon which no exper- 

 iment was tried, and were consequently subject to invasion by 

 codling moths of the second brood, reared upon these unpoisoned 



