356 



Kleinere Mitteilungen. 



precisely the same (fig. i of the first communication). We then obtain as 

 an cud curve a remarkable many-topped line which in the beginning pro- 

 duces the impression of being the product of too small a number of leaves. 

 This is not the case however. In the first place the average number of 

 leaves of the trees with which I experimented is about 2000. But the 

 essential point again is this, and as I have in all my cornmunications 

 always pointed out with the greatest insistence, that all the curves of that 

 analysed end-curve have exactly that same many-topped habitus. In this 

 case the different leaf-length properties or the activity of these properties 

 are very regularly distributed over the tree. Such examples I have often 

 met with, and they will later be set forth in extenso. Let us now consider 

 our new curve, we shall then make ourselves acquainted with something 

 quite different. 



I thought in the beginning to have an instance in which all the 

 curves exhibited precisely the same top as was the case with the first 

 four curves. Suddenly however half-way up the tree, the top thrust out 

 a large distance to the right side, and to my astonishment the consequent 

 curves as well as the definite end curve exhibited exactly the same top 

 as curve 5. It is noteworthy that this top-removing happened suddenly, 

 without transitiivi. Now it is known from my second communication that, 

 as in the analysis of all my other curves, the removing of the top occurs 

 suddenly without the least transition (see the second communication). 



One will regard this newly recorded case as a bud- variation. However 

 in any case it is indeed a particular sort of bud-variation. Really not a 

 bud-variation in the sense in which one is generally accustomed to conceive 



