WASHINGTON, YOU'MK XI, 



1(11 



reaches to the level /' i- another family or -ubfamily, as the 

 characters may warrant, irrespective of whether it originated 

 from main branch /, //, or III . 



JHit sucli a system would be purely mechanical and not ac- 

 cording to nature's own divisions. And while we with (Un- 

 limited powers of observation are forced to adopt to some 

 extent some mechanical arrangement in order to get any classi- 

 fication at all, it is manifestly obligatory, when we can trace 

 the phylogeny and realize the true natural lines, that we utilize 

 such knowledge and not continue our temporary mechanical 

 system, merely because it is easier. 



A 



B 



a bcclefgh tjkl 



m 



7 IT i 



l-'ii'i. ">. Piagriininr.it ir ph.vlouniii- tree. 



Thus in the above illustration it is clearly correct to place 

 the species (or genera) a, b, c, d, e, and / in the same family 

 or genus ./ as their common characters may warrant, but in 

 this family must not be placed g, li, i to o. 



Likewise it may be proper to include s, t, it. r, w, and even .r. 

 in one systematic entity (B) of whatever rank the characters 

 may warrant, but />, </, and r will have to be considered sepa- 

 rately, although they may have developed similar characters. 



And while c, f, and u may look enough alike to tempt putting 

 them all in one group, c and f are first cousins and therefore 

 justly placed together, but f and // have their common ancestors 

 far back in ancient history and have attained similar characters 

 on entirely independent lines. 



