July IS, 1919 Relation between the Flowers and Fruits of the Lemon 1 59 



was greatest — namely, in the spring months, or at other times at which 

 the activity of the tree is at its height. This conclusion is in agreement 

 with the repeated observation that thrifty trees most commonly bear 

 lemons in clusters. 



The next question to be investigated was one of considerable physio- 

 logical interest: What is the correlation between the number of buds per 

 inflorescence and the numbers of fruits matured per inflorescence? 



A positive correlation approaching i is to be expected in case buds 

 on all sizes of inflorescences have equal chances of development ; a value 

 much below this indicates that a bud on a larger inflorescence has a 

 poorer chance. It is certain from the nature of the case that there must 

 be some relationship between the two, since an inflorescence having only 

 I flower could not produce more than i fruit, but an inflorescence pos- 

 sessing 20 flowers may or may not mature a proportional number of 

 fruits. If we take the fruits whose history could be definitely ascertained 

 and arrange them with regard to the size of the inflorescence from which 

 they developed, we get the arrangement shown in Table V. 



Table V. — Correlation between numbers of buds and fruit on inflorescences of various 



sizes 



Fruits matured per inflorescence 



r=o.i78±o.oi7 



