Kleinere Mitteilungen. jcy 



it. We have here the case then that not merely one branch, nor a single 

 small part of the tree exhibits a strong deviation, but that suddenly the 

 m/!re further part of the tree deviates, and this deviation in constant. 

 About the lower half of the tree there is an average leaf-length of i6 Vi c. M., 

 whilst the upper half of the tree has an average leaf-length of 19 c. M. 

 Both parts are eaf/i in its-cnvn way constant in character. One would be 

 much inclined in my case to speak of a vegetative mutation. Yet relying on 

 the principal of the Sympodial construction of the tree we can speak of a 

 bud-variation. 



Let us now refer back to both of the previous investigations. We 

 then once more observe all those analysed curves. Is there then a difference 

 in principle between this newly recorded case and all the others? Is there 

 a difference in principle in the question whether it is only once that a top 

 removing of the curves occurs within the end curve (as in our present 

 case) or that several times top-removing takes place (as is the case in the 

 two previous communications). Certainly not. In all these cases we have 

 to deal with precisely the same phenomenon. In all the curves analysed 

 in both of the previous communications we have had to deal with repeated 

 bud-variation, whereby we again draw attention to the fact that variations 

 do not so much take place in all sorts of arbitrary directions, but that 

 definite variations are repeated in the same way. The number of variations is 

 limited. The 7nodes of variation are limited. One variation suddenly passes 

 into the other, witliout any transitional intenal, just as we have explained in 

 extenso in the second communication. This sudden transition of one 

 variation into another, this individualization of the variations is indeed a proof 

 that we have here to deal with an expression of definite clearly marked 

 properties. No further doubt is possible here. In different parts of the 

 tree different properties manifest themselves. This phenomenon is precisely 

 the same as the occurrence of "the different modifications" in the very 

 same sort of plants on the same land, quite according to their position, 

 quite according to the way in which external influences affect them. With 

 the coffee trees treated of by me, that phenomenon occurs in a minor 

 degree in one and the same tree. The different curves within such an analysed 

 curve are just so many modifications. Under the influence of other circum- 

 stances other properties, other factors, come to predominate and confer 

 another character on that part of the tree, precisely so as with the 

 "modifications" of whole plants. Concerning this concept relating to the 

 nature of modification I refer back to both of my former publications 

 respecting the statistics of coffea. 



? 2. 



The problem of the non-genetic factors. 



It speaks for itself that the results before communicated and contem- 

 plated anew have set in a shrill light the antithesis between nonheriditary 



