Variation in the seeds and pulp-vesicles of 



Citrus Atirantium L. subsp. nobilis 



Mak. var. TacMhaua Mak. 



{Continued from p. 67). 



By 



H. N a k a 11 0. 



(With Plate II and one Figure in the Text.) 



In the foregoing pages I have often pointed out that the 

 weight of the fruit in the lower branches is always exceeded by 

 that of the upper branches, while the number of vesicles does 

 not always follow suit. This fact shows that the correlation 

 between the weight of the fruit and the number of vesicles is not 

 remarkably great. For this reason I have tried to find out the 

 degree of correlation between them. For purposes of investiga- 

 tion, the fruit of Fukuremikan which was collected in A habitat, 

 was carefully cut off from its axes, but the embedded part of 

 the axes in the rind was not taken off. The weighing was 

 carried out to one decimal gram and the round numbers were 

 obtained by approximation. 



From the following table it may be well explained that the 

 weight of the fruit and the number of vesicles do not vary 

 correlatively to a marked degree, although they run each other 

 close. When the 11— vesicled fruit is rich in the upper part,^ 

 while 10 — vesicled one is rich in the lower part, the result is 

 that the fruit of the former is greater than the latter in the 

 number of vesicles, but less in weight. The same case can also 

 occur between two vesicles 7 and 8, and 12 and 13. 



