Kleinere Mitteilungen. 



123 



I cannot refrain from mentioning a still more remarkable case, also 

 among the Canephora. This is represented in Curve IV also regarding 

 leaf-length. 



After the 14''' branch-pair we see that the measurements 5 — 11 V^ c. m. 

 disappear for good! And this indeed only after they have manifested 

 themselves very strongly. They do not occur any more in the further 

 development of the tree : they have become latent for good. The becoming 

 latent of properties during the development of one and the same tree is 

 again an indication of the manner in which we must interpret in my case 

 of Coff'ea, but perhaps also in ordinary frequency-curves. Just as the indi- 

 viduality of the "properties" which makes together a curve of frequency has 

 been continually indicated throughout these treatises, so does this local 



becoming latent indicate once more a certain independance of the properties. 

 Opposed to this becoming latent is the fact that suddenly a local definite 

 measurement may acquire great activity. If we but dwell on our last curve 

 we see after the 7"^ branch-pair a great development of the measurement 

 14 — 15 c. m. while during the same period of time the measurement 

 21 — 22 c. m. is scarcely developed at all. Moreover examples of this 

 phenomenon can be taken from all the curves given in these treatises. 

 Also this individual local increase of one or more measurements points to 

 a certain independance of these measurements, of these "properties". 



So we have to regard the whole curve of frequency as being princi- 

 pally a complexity of pro|;erties; the ordinary one-topped curve is also then 

 the function of a more or less extensive number of factors of a definite 



