41 



I have generalized somewhat freely in giving these directions; 

 as a matter of fact, the tests I have actually made are so few (and 

 I am so far from being a chemist) that it is with some trepidation 

 I write about them at all. Nevertheless, so far as they go, they 

 give apparently consistent results. They show four types of 

 reaction, as follows. 



A. Solutions from the petals of white flowers, whether normal- 

 ly so or albino forms, give, as might be expected, almost wholly 

 negative results. They are unchanged with acid and slightly 

 yellowed with alkali. 



B. Solutions from blue, purple, magenta, pink, and red flowers 

 turn pink or red with acid or retain those colors if they had them 

 originally. With ammonia they turn first a greenish blue (or 

 bluish green (it is hard sometimes to tell just what to call this 

 color) changing presently to greenish yellow, pale yellow, or to 

 an almost colorless condition. If too much ammonia is added, 

 the yellowish color appears at once without the intervening 

 greenish or bluish stage. These are, of course, well-known 

 anthocyan reactions. Since anthocyan pigments occur dissolved 

 in the cell-sap and flowers of this series make almost equally 

 good solutions in water and in alcohol (and probably would in 

 any neutral liquid which would mix with water), and since the 

 colors of the flowers are anthocyan colors, it seems safe to assume 

 that we have here cell-sap pigments of that class. The change of 

 the solution from blue to pink with acid is exactly what occurs 

 when the hepatica, for instance, develops a pink form; the change 

 to greenish blue with ammonia parallels that in the withering 

 flowers of Desmodiiim. It would appear, then, that this type 

 of color variation occurs in flowers having anthocyan pigments, 

 that pink forms of normally blue flowers develop in individuals 

 in which the cell-sap is, for some reason, more acid than is 

 usual in the species, and that, as was long ago postulated for 

 boraginaceous flowers which are pink in the bud and blue when 

 mature, an alkaline condition develops in the withering 

 flowers of the Desmodium. 



C. Solutions from some yellow flowers are unchanged with 

 acid (at least at first; I am not sure that they might not turn 

 green if left to stand for several hours) and slightly deepened 

 in color with ammonia. Since these flowers do not give good 

 solutions with water but only with alcohol, their pigments may 



