POLLINATIONS AT ROCHESTER. 



29 



were falling beliind in growth and becoming paler in color. Accord- 

 ingly the count was delayed till these fell. The record was taken June 

 13 and 14 and was verified June 21. 



Tlie emasculations and pollinations were performed in exactly the 

 same manner as at CJiestuut Farm. In all cases it Avas planned to carry 

 on enough experiments of one kind on each tree to cover the chances 

 of accident. Wherever the tree was of sufficient size to allow it, twenty- 

 five or more bags ay ere placed on it to cover the possible variations of 

 difierent branches and secure a reasonably correct general average. 

 In computing the per cents in the following tables the number of 

 clusters is first multiplied by 7.5, the average number of flowers i)er 

 cluster. The per cents are then computed on the basis of the num- 

 ber of flowers, not of the number of clusters. 



Taklk 4. — Anjon (dwarf) tree, with paper har/s. 





Covered . 



Uncovered. 



liecord 









No. 



Clusters. 



Fruits. 



Clusters. 



Fruits. 



2896 



3 







7 



9 



2897 



2 











2898 



5 







8 



3 



2899 



2 











1 2900 



5 







7 



8 



■ 2901 



4 











2902 



2 











5 



2903 



3 











2904 



1 







9 



12 



2905 



1 



1 







2906 



1 











2907 



4 











2908 



G 











2909 













2910 



2 











2911 













2912 



3 











2913 



7 











2914 



4 











2915 



2 











2916 



3 











2917 



1 











2918 



3 











2919 



2 











2920 



2 











Total. 





1 i 37 





Result: Covered, 0.01 per cent; uncovered, 13.3 per ceut. 



The above table shows the A^^ork on a dwarf Anjou tree. The twenty- 

 five bags on this tree contained from one to seven (dusters each and 

 from all these only one fruit Avas produced. A number of branches 

 outside the bags were selected and their clusters and fruits counted to 

 determine the natural fruitage of the flowers when exposed to insect 

 visits. The total number of clusters inclosed in bags AA^as seventy- three, 

 from Avhich one fruit set. Outside the bags thirty-seven clusters set 

 thirty-seven fruits. The data obtained from this table will be found on 

 X). 37, tree 15. All similar data in the synopsis reiiresont the totalis of 

 tables like this and the following; 



