148 Kleinere Mitteilungen. 
Uganda Robusta 
3457 er, 
39,3 = 
42,1 Pen 
43.4 — 
Also with regard to these points above mentioned, we find this to be 
confirmed: with Uganda the numbers of a single tree are in general very 
variable, whilst in Aodusta a certain uniformity is perceptible over 
the whole tree, just as I mentioned before. The question is, if the causes 
of these two phenomena can be traced. 
It might be possible to find the reason in the phenomenon that of 
the different qualities a greater number of each quality is present in 
the plant, but, that these qualities are not distributed over the whole 
plant in the same manner. 
In order to demonstrate this I give below (curve IV) an analysis of a 
curve of frequentation of the leaf-length of Uganda in the same way as 
occurred with the curve of Kobusta. (See curve I.) 
If we now carefully compare this curve of Uganda with that of the 
above mentioned one of the Rodusta then we shall see that in Kobusta at 
different heights on the tree, the character of the curve is generally the 
same. The different length-qualities, indicated by the apices, all occur 
at different heights and also in the same manner, in the same predo- 
mination, in the same proportions in relation to each other. 
With Uganda, on the contrary, the constellation of the properties 
at different heights of the tree is very different. At each height the 
curves are different; and it is other properties which predominate. We 
see at each height shiftings of the tips. In some cases the curve is of 
many apices, or again of a single apex. Again we see (and this is again 
a help to the conception, that the tops are not merely accidental but in- 
deed of some real signification in the plant) the periodically reappearance 
of a same top: top a, returns in top a»; top b, returns in b, and bs etc. 
From this it appears that there is a definite number of apices present, so 
that this regularity and persistence in the reappearance of the same apices 
may be considered as an indication of the presence of a greater number 
oflength-properties as has also been accepted in the whole of this research. 
So the possibility exists that the different ways in which in Rodusta 
and Uganda the further representatives of a property are divided over the 
tree is the reason of the similarity of the numbers in one and the same 
Robusta tree, while it is on the contrary of the variability with one and 
the same Uganda tree. 
