THE BOTANICAL MAGAZINE. 



[Vol. XXVI. No. .WS. 



modes on 11, while the fruits of Kojimikan can do so. Hence 

 two varieties are not always distinguishable in the number of 

 vesicles, though each holds a somewhat peculiar character. 



Finally I examined the vesicles of JJnshumikan. Though m}^ 

 investigations were restricted to a small number of fruits, I 

 became quite convinced that the fruits of Unsbu have not so 

 many vesicles, notwithstanding the larger volume. 



On 5th Nov. 1909, 143 fruits were picked without selection 

 from a tree of about 3 m high in B. habitat. 



The data obtained are : 



On 5th Nov. 1911 all the fruit was picked from a tree of 

 about 1.5 m high in A. habitat. 



The data are. 



If it is permissible to draw conclusions from such scantj' 

 material, we ma}'- say that the number of vesicles of TJnshu- 

 mikan very nearly coincides w'ith that of Fukuremikan and no 

 difference can be found between them, (to he continued.) 



