120 



Kleinere Mitteilungen. 



with the numbers of the curves in Curve I. In this manner the internodal 

 lengths have been represented in curves of the 9*, 14 "1, 25"', 40 "\ 52'"', 

 öö'** branch-pairs (see curve II). It must be observed here that these curves 

 only represent the principal and most characteristic alternations in the 

 lengths of the internodes. Now look at curve II. 



We here perceive the following fact : in curve I we saw that the same 

 tops manifested themselves after branch-pairs 14, 40, and 66, so the red 

 coloured curves belong to each other. If we now compare with this the 

 curves of the lengths of the internodes of the 14''', 40''', and 66"^ branch- 

 pairs, also red-coloured, then we see that these latter curves practically 

 coincide. 



Exactly the same thing is to be seen with the violet coloured curves 25 

 and 52; Thus, what coincides in curve I also coincides in curve II. 



7 Zd456769f0 1J 

 rank- number of /he internodes per branch. 



IZ 13 « 15 IS 



Curve II. 



W 13 20 21 22 23 2t 2S 2e 27 2a 



In curve II we thus observe a grouplike position of the curves. In like 

 manner the lengths of the internodes of some branch-pairs both preceding 

 and following those (colored) selected are brought into curve by stippled 

 black lines. These lines are seen also to coincide with those (colored) 

 selected. With regard to these fine-dotted black lines it may be said that 

 no transition exists between the different curves, but that they suddenly 

 overlap from one habitus to another. Those parts of the tree are conse- 

 quently sharply defined ones; thus, the tree consists in sharply individu- 

 alisased parts without transition, without relation. I will therefore demon- 

 strate this in the same manner as I have previously done, and I will even 

 hold the results arrived at in this paragraph as a direct proof of my inter- 

 pretations. In C. Uganda there are more forms of leaf-length-properties, 

 and also a greater number of properties of internodal length; this greater 

 number of properties of leaf-length and also of internodal length is certainly 



