A SUMMARY OF THE REVISED EDITION. 



colors is requisite if one would avoid confusion. Such a 

 scientific arrangement of names will be found in the 

 Introduction, and the adaptation in the Color Key, 

 Early in the nineteenth century De Candolle arranged 

 these flower-colors in two comprehensive divisions which 

 he named: (1) Xanthic, Yellow, and (2) Cyanic, Blue. 

 This pioneer and somewhat crude attempt to reduce a 

 multiplicity of color types to a system and establish a 

 certain relationship between its two great divisions, is 

 precisely the best means of enabling us to understand 

 the extreme limitations of our North American flower- 

 colors. Retaining yellow and those hues of De Candolle's 

 Cyanic division which grade through crimson and purple 

 to ultramarine, we have exactly the colors which belong 

 to the flora of our range except the exceedingly small 

 percentage of orange and red belonging to the Xanthic 

 division ; these last are contributed by plants which for 

 one reason or another survive beyond their proper home 

 in the subtropical region. We have scarcely a true red 

 flower within our range ; the same may be said of a blue 

 flower; only the Family Boraginacess shows any approach 

 to true blue. A careful study of my Color Key will 

 disclose the fact that about seventy-five per cent of our 

 flower-colors is equally divided between yellow, white, 

 and magenta-purple ; the remaining twenty-five per cent 

 is scattered between pink, orange, and a negligible quan- 

 tity of red and so-called blue. It is well to note also 

 that many of our white flowers are albinos and the rest 

 are in a large measure showy agglomerations of fussy 

 little blossoms, the very opposite of our wonderfully 

 developed White Water-lily. It is very evident, there- 

 fore, that conditions of light and heat are responsible for 

 the modification and limitation of all flower-colors in 

 the North, and that these in their turn are the direct 

 evidences of an arrested development. With this under- 

 standing of the very limited range of color involved, 

 and with the aid of the Color Key, it ought not to be dif- 

 ficult to trace a given specimen if one prefers this method 

 of procedure. The chances are that scarcely a true blue, 

 orange, scarlet, or red flower will be encountered in 

 the field. 



