124 Kleinere Mitteilungen. 



quality; these factors, these "qualities" are not indeed all activated to the 

 same extent. It is very difficult to say how many properties are involved 

 in such a curve, and I personally believe that it will never be recorded 

 with precision, even by the help of the most intimate researches in here 

 dity, and this by the fact that an intense affinity of factors, inseparable 

 from each other, may be involved. How many properties are involved 

 one cannot say. But if we subject the curves given by us to observation 

 we shall indeed perceive remarkable manifestations. In our first treatise 

 concerning Curve IV e. g. the different measurements which are effected 

 in the course of the development of the tree, and which are consequently 

 to be considered as independant properties, are those from 9, 15, 16, 18, 

 19, 21, c. m. ; in curve I of the first treatise those of 11, xzVi, 15, 16, 17, 

 18, 20, 22, 24 c. m. tc. tc. From these numbers we thus see that of the 

 total number of measurements or classes of measurements, an exception- 

 ally large number manifests itself as individual properties without our 

 thereby knowing anything of eventual other properties. This is the 

 reason why we have generally so to interpret the frequency curve that 

 theoretically just so many properties compose it as we have set off mea- 

 surements of classes on the abcis. Theoretically we must regard each of 

 such class or such measurement as the representative of a factor. As a 

 working hypothesis this comprehension of the substance of a frequency- 

 curve may possibly have its value, and I will at the same time connect 

 it with the further course of my researches respecting the coffee plant, 

 there, where indeed we must dwell a moment on the signification of the 

 half curves which I have principally found in my study of Canepkora. 



§4. 



It is known that half-curves are somewhat general. How then do 

 half-curves originate? In the material of my researches I came across 

 them in Caneplwra; and indeed I came across them as well in the frequency 

 curves that gave the number of ca]iitella per leaf-axil. The coffee flowers 

 are grouped in capitella of 4 — 5. In the leaf axils there are joined more 

 of such capitella: 4 or 5 or 6. With all the other varieties of coffee the 

 frequency curves of the number of capitella per leaf-axil gave the normal 

 curve except with Canephora which in this respect gave us half-curves. 

 Two types of these-curves are represented here in curve V. 



In the interpretation of how such half-curves occur we may make 

 use of the demonstration which I have set forth in the former paragraph 

 in which theoretically each class or measurement or number taken from 

 the abcis represents an independant "property". Herewith we must also 

 observe the results that have become apparent in these two treatises 

 regarding the individual behaviour of the properties which concern strong 

 individual activity as well as individual latency. Curve V should then be 



