New York Agricultural Experiment Station. 297 



twenty-five years, and two Hubbardston trees about forty years 

 old. Trees of the same variety as nearly alike in all respects as 

 could be found were paired for comparison, one of each pair 

 being thinned and the other left without thinning. With the 

 exception of the operations of thinning the fruit, all trees under 

 'pcperiment were in all respects similarly treated. Three ways of 

 thinning were to be tried. 



First method. — All wormy, knotty or otherwise undesirable 

 specimens removed and each cluster thinned to one fruit. 



Second method. — The same as the first and in addition the fruit 

 thinned to not less than four inches apart. 



Third method. — The same as the first and in addition the fruit 

 thinned to not less than six inches apart. 



The original intention was to continue the treatment for a 

 series of years on the plan above described in order to learn its 

 effect not only on the fruit of the current season, but on the yield 

 in following years; but because in some seasons there was little 

 fruit or none on some of the trees originally selected for thinning, 

 the plan was modified as hereafter stated. 



FIRST METHOD. 



The first method, in which imperfect specimens were removed 

 and clusters thinned to one fruit each, was tried on Baldwin only. 

 Two trees which were heavily loaded in 1896 were selected for 

 this experiment. The one on which the thinning was to be done 

 was designated " 13," the other " 14." 



Record for i8p6. — The thinning was done June 27th, the largest 

 fruits being at that time about one and one-half inches in diam- 

 eter. The fruit on tree 13 was thinned by taking of¥ all knotty, 

 wormy or otherwise inferior fruit, and all clusters were thinned 

 to one fruit. It took four hours to do the thinning and four for 

 picking, making altogether eight hours. It took five hours to 

 pick the tree that was not thinned. The marketable fruit graded 

 as shown in Table I (p. 299). 



. From this we see that while the tree which had its fruit thinned 

 gave 16 per ct. less fruit fit for barreling than the unthinned tree, 

 10 per ct. more of it ranked No. i, so that it really yielded as 



