334 Schively — Contributions to the Life History of 



dent ; legumes were ripe at the end of two months. Not 

 the slightest indication of purple flowers was seen upon 

 any plants during the winter and spring. 



Summary of Views on Floral Variation. 



To recapitulate, Amphicarpcza presents the case of a plant 

 bearing at least three, probably four distinct types of flower ; 

 and these not occasionally, nor spasmodically, but simply 

 following regular laws of development. 



In summer, purple flowers may be expected to appear upon 

 the upper main stem, and the upper axillary branches. A 

 certain intensity of the sun's rays seems necessary, for 

 these are not found upon plants growing entirely in the 

 shade. Neither do they develop extensively upon plants 

 constantly exposed to the sun. Here the water supply may 

 be insufficient ; plants grown in such a situation are not apt 

 to reach a fair height. 



But later the same plant, having, we will suppose, as is 

 often the case, produced an abundance of purple flowers, 

 now proceeds for a month or more to bear aboveground a 

 cleistogamous form. 



Frequently, too, plants which have not developed a single 

 colored flower, may, late in the season, produce this last 

 type. 



Similar cases of this method of development are not un- 

 known. Among the Violaceae it is quite common, but all 

 of the cleistogamic forms are produced close to the ground, 

 hiding among the leaves, and only occasionally upon spe- 

 cial short branches. 



Specularia perfoliata bears inconspicuous flowers late in 

 the spring ; in June the bright purple ones are found. 



Henslow 13 mentions the finding of flowers of Linaria 

 and Potentilla in the autumn, in which he discovered pol- 

 len tubes passing to the stigma. These specimens, though 

 imperfect, however, still possessed color. 



Amphicarpcea certainly bears its cleistogamic flowers in 

 great abundance, some of them in a very conspicuous posi- 



