20 



white or yellow, each of the latter colors proceeding directly 

 from white or retrograding to white ; or these three pri- 

 mary colors may, as we have shown, originate immediately 

 from the chlorophyl green, or actually precede it ! 



CHAPTER III, 



The lianumnilaced' are i)laced first in order in our Floras, 

 Ave imagine for the reason that the fiovvcrs of this family 

 are very simple in construction, especially so if we except 

 some peculiar shapes of petals in Aquilegia, Aconitatn and 

 Delphinium, but even in these as well as in all the others, 

 all the parts are separate and distinct. Hermann Miiller 

 calls a flower simple which is without any consolidation of 

 either the essential organs or floral envelopes. 



Grant Allen, in elucidating his color theory, selects from 

 among this family the common yellow buttercup) (Ranun- 

 culus) as a typical example of a simple or primitive flower 

 and maintains that it is prevailingly yellow "because it is 

 an early and simple type of flower." 



But before accepting his conclusions one should be fully 

 assured as to the correctness of his premises ; we therefore 

 would like to be informed in the first place, if it really is 

 "the simplest and least differentiated member of the group," 

 as he puts it, and if there is no member equally as simple 

 and yet having a different color. 



Can one believe that there are no species of that family 

 even more primitive in character? Take for instance Cltm- 

 rtds Virginiana, the Virgin's-Bower ; this flower has not even 

 developed any petals, it is so primitive, and the number of its 

 white sepals is but four, instead of the more common number 

 five. A double Clematis srowinu" in the writer's grounds 



