114 



of members according to one and the same view. When judged quite apart 

 from any partial treatment, even the best system at present known as such, 



while pretending to follow tha mutual relations of A. B. C. D , in 



reality, merely denotes the relation of A -B according to one view ; that of 

 B C according to another view; and that of C D according to still another 

 view ; and that of some remote members, such as C and K, very faintly, 

 if at all. The construction of such a system is something like sewing a 

 fox's skin to a lion's. Figuratively speaking, such a system is comparable 

 to a marquetry picture of a mountain, one piece of wliich is taken from a 

 picture of the mountain viewed from the south, a sacoud piece as viewed from 

 the north, a third as viewed from the east, and so on ; while the real natural 

 system is to be likened metaphorically to a model of the mountain itself. 

 Such a model, if seen from different sides, presents different phases, as a real 

 system denoting real natural relations should ; but, on the contrary, the 

 marquetry picture shows but one appearance like the system at present known 

 to us. The phase of the former is dynamic, while that of the latter is 

 static. If we may be allowed to call such a static system "natural", as we 

 usually do, then the systems of LINNAEUS, ANTOINE JUSSIEU, DE CANDOLLE, 

 ENDLICHER, BRONGNIART, BENTHAM- HOOKER, VAN TIEGHEM, ENGLER, HALLIER 

 or even what TREUB has proposed, should be regarded as natural. On this 

 occasion, however, it can not well be said that one system is natural while 

 the others are artificial. All systems should be true and natural regarding 

 one part according to a certain view, but regarding another part by another 

 view. But, if by natural system we mean, as we ought, that one in whicii 

 we can see all the natural relations of all the members of the system (according 

 to any view and between any two members), then it must be quite different 

 from the systems of the great authorities above mentioned. It follows that 

 the above - named systems are not natural but artificial, and have been establi- 

 shed merely for the sake of convenience. We can not, therefore, go so far as 

 to say that this system is more natural than that ; but rather that the former 

 is more convenient than the latter. In this sense, even ENGLER'S system, 

 which is at present regarded as the most natural, is nothing more than a very 

 convenient system. In the following pages I shah 1 try to give a full account 



