54 



colors under certain circumstances might precede this green 

 and also that more primitive than all is the fundamental 

 sh-uctural or chlorotic white. 



CONCLUSIONS. 



We tind that among the Ranunculacea', yellow is not by 

 any means the prevailing color. In Clematis ydJoir is 

 comparatively a rare color. 



Very many existing yellows are known to have been pre- 

 ceded by entirely different colors. Brown-purples and sim- 

 ilar hues formed of green and red are earlier than yellows. 



Simple or primary yellow does not regularly change 

 through white to led or purple ; an exceedingly common 

 sequence being yellow to orange, vermilion or scarlet, or to 

 brown ; or through green to l)lue, or directly to crimson. 

 Even in the Buttercup, there is a change of yellow^ to scarlet ! 

 Not infrequently green and [)urple, green and brown, green 

 and red, green and white, green and violet, etc., ai)pear 

 before any yellow color develops. One w^ith a trained eye 

 for color can scarcely look at a flower without detecting a 

 difference in the development of its hues ; the anthers per- 

 haps diverging into one series of color, the petals into 

 another, the pollen often different in hue from the anther 

 containing it, and possibly the sepals different from either, 

 as in Lobelia Cardinalis ; it frequently occurs that the very 

 hairs ((re of a different color from the stem or leaf from 

 which they grow. Size, shape, or construction, as a rule, 

 has no effect on color, neither simplicity nor complexity. 



Nature is highly democratic in her feelings; she bestows 

 the royal purple or cerulean l)lue as freely upon Clematis, 

 Hcjiatica and Anemone of the simplest family as u})on 

 Aster, Hyacinth, and Gentian of the highest families, and, 

 conversely, she distributes the golden yellow on the wdiolc 

 in as great a pro])ortion among the higher as among the 

 simpler species. 



