WEATHER RlOCOKl) AT ROCHESTER. 19 

 Table 1. — Weather record at Bochester, X. Y., May, ISDl, 



Day or 

 iiioutli. 



Temperature. 

 Maximum. | Minimum. 







Mean. 



Ilaiuliill. 





05 





Degrees F. 



Inches. 



1 



45 



55 





2 



6G 



40 



53 



0. 03 





58 



42 



50 



0. 44 



4 



49 



39 



44 



J. 00 



5 



;j9 



30 



34 



0.01 



6 



47 



3 J. 



39 



1. 00 



7 



51! 



39 



40 





8 



07 



44 



50 





9 



7(5 



54 



65 





10 



82 



50 



69 





n 



50 



47 



52 



1.00 



12 



01 





49 





13 



U4 



39 



52 





1 A 



14: 



57 



47 



52 





15 



77 



40 



58 





]6 



48 



44 



40 



0.01 



17 



54 



33 



44 





18 



(50 



40 



53 





19 



09 



41 



55 





20 



8;? 



54 



08 • 





21 



78 



04 



71 



0. 02 



22 



50 



50 



53 



0.01 



2o 



54 



40 



47 





24 



(58 



40 



54 





23 



75 



47 



01 





Warm weather dariii*^- the last lipw days of April brought out tlie 

 buds, so that wheu a cold wave caine on May 5 and G the white petals 

 could be seen. This cold wave Avas accompanied by a strong north- 

 west wind and driving snow, and made the outlook for the blossoms 

 rather discouraging. The thermometer fell below freezing, and water 

 standing in pails over night had a thin skim of ice in the morning. 

 When milder w^eather came, on May 7 and thereafter, the flowers were 

 examined, but no visible injury could be found. It may be, however, 

 that the cold weather did injure the flowers to some extent by chilling 

 them. The warm weather on May 10 caused the petals to open. The 

 week following was cool, dry, and sunny. When the sun shone brightly 

 it was ideasant to be out, but it was too cool to bring out many insects. 

 During the warmer parts of the days, however, insects were fairly 

 abundant. 



Results of the experiments. — The count was taken on May 30 and 31, 

 and the fruits were examined June 3 to ascertain if any had fallen ofl'. 

 This was rather earlier than desirable, but still late enough to show 

 the general result. In making the count the number of clusters inside 

 each bag was first determined. This was indicated by the peduncles 

 from which flowers had fallen, or by the young fruits. The number of 

 healthy fruits resulting from these clusters was also noted. In order 

 to determine the general fruitfulness of the clusters outside of the bags, 

 for comparison, a count was made of a number of clusters and result- 

 ing fruits on branches selected at random over the tree and of about 

 the same character as the covered branches. Not appreciating the 

 importance of the exx^eriments at the time, in a number of cases I only 

 guessed at the fruitfulness of the trees at large, without making a careful 



