18 THE POLLINATION OF PEAK FLOWERS. 



sod and had been for several years, consequently had made but little 

 growth. Though not positiv ely unhealthy, they api)eared stunted. The 

 trees in the gardens in good vegetative condition gave a very satisfactory 

 set of fruit, while those in the grassy orchard set a liglit crop. 



The experiments consisted in covering the buds with bags a day or 

 two before the flowers opened and noting the eflects. The bags were 

 of three sorts, mosquito netting, clieese cloth, and pa])er. The mosquito- 

 net bags Avere selected because they would exclude all but the most 

 minute insects and at the same time necessitate but little change from 

 ]iormal conditions. The wind could readily blow through the meshes 

 of the net, Avhich were about ten to the inch, and if the i)ears were 

 adapted for the puri)()se pollen coidd doubtless be blown by the wind 

 on the stigmas. Piiper bags were used because they afforded the 

 most perfect exclusion of pollen and insects, Avhih^ cheese-cloth bags 

 Avere regardc^l as m mean between the other two. In two instances the 

 three kinds of bngs were ])laced on the same tree for comparison. 

 The bags, which were from 10 to 18 inches long, were of sufficient size, 

 to inchjse a large fruit spur, or a branch with from two to six or eight 

 small spurs on it, so that each bag contained from two to ten or more 

 clusters. The bags after being drawn carefully over the buds, Avere 

 gafhered closely to the Inanch and flrndy wired. It was found con- 

 venient to use shipping tags for labels, and to fasten the Avire into the 

 tag before going to the field. Each label received a number correspond- 

 ing to a serial number in a notebook, Avhere the facts about the exjjeri- 

 ment Avere recorded. 



The bags Avere i)ut on the trees on May 7, 8, and 9. By evening of 

 the last day a few flowers opened, on May 10 the flowers came out 

 rapidly, and on May 11 the trees Avere white with bloom. On May 14 

 the petals began to fall, and by May 17 the greater ijart of them Avere 

 on the ground. By the 24th a large number of the little pears had 

 fallen off, this being the noi inal shedding of the weak or unfecundated 

 flowers. 



Weather record. — Eochester, 18 miles from Brockport, is the nearest 

 point at which detailed Aveather observations are taken. Brockport is 

 about the same distance from Lake Ontario as Eochester and at very 

 nearly tlie same altitude, in a very level country, so that the foUoAving 

 record for Eochester must be nearly correct for Brockport ; 



