Kleinere Mitteilungen. 



119 



I shall indeed presently try to make out whether, following these 

 analytical methods, there exists a direct connection between the frequency 

 of leaf-length and that of internodal length. 



In the curve immediately following below (Curve I) the frequency- 

 curve of the leaf-length of an example of C. Uganda has been analysed in 

 the same manner that has already occurred. The measurements began at 

 the base and advanced towards the top. At different lengths of the tree 

 the results already obtained have been brought to expression by a curve. 



3 -i S 6 7 e d W II 12 13 U IS IS n IS IS 20 21 22 23 24 25 26 27 



length of leaves in c. m 



Curve I. 



Although perhaps not so nicely as in the former communication, it 

 is however very well to be seen here that at different heights of the tree 

 definite tops again recur. — Curves with the same tops are represented 

 by the same definite colour. The numbers in the curves relate to the 

 order of succession of the branch-pairs. The tree had 66 branch-pairs; 

 the curve with the number 66 is consequently the definitive frequency- 

 curve of the leaf-lengths of the whole tree. The curves are thus following 

 after 9, 14, 25, 40, and 52 branch-pairs. Now the internodal lengths of 

 this self-same tree. These also have been represented by curves, in which 

 the internodes of those branch-pairs have been selected which correspond 



