178 Kraemer: jMorpiiology of the Gexus Viola 



Kcrner upon f^ sepiiicola. He found that this plant produced 

 clcistogamic flowers only when growing in shaded woods and 

 that the same plants when exposed to the sunshine produced 



r 



beautiful blue and scented blossoms. This is possibly due to the 

 fact, as he suggests, that when exposed to the sun bees will visit 

 the flowers which they would not do if the plants were growing 

 in dark woods. 



4. In some plants, as in V. Jiasfata^ the rhizomes are frequently 

 light-colored, approaching white. It would seem, further, accord- 

 ing to Kerner, that such rhizomes were produced in these plants 

 only when growing in dark recesses and well covered with soil, 

 for if the plants are exposed to greater light these rhizomes become 

 violet-colored. This would indicate that color of rhizome, like 

 the previous characters, means simply a characteristic that is de- 

 pendent upon situation or habitat. 



5. What value shall we place upon colors in flowers? Kerner 

 has shown that when V, calcarata grows in the meadows of the 

 western Central Alps that the flowers are of a blue color, whereas 

 when growing in the Tyrolese Vintschigan they are yellow and 

 when found on plants in the limestone mountains of Hungary 

 they are violet-colored. It would seem from all observations that 

 a number of factors are at work in producing colors in plants and 

 that they are dependent upon conditions upon which we know 

 very little. 



6. The presence of hairs upon stigma, anther cells, side and 

 spurred petals have possibly some significance when taken in con- 

 nection with the shape of style (whether straight or with a knee- 

 like bend) and length of spur of anther in assisting us to ascertain 

 certain facts In the developmental history of this genus. Kerner 

 considers that a style that is bent or deflexed is a device for imped- 

 ing the progress of the insect into the flower so that all of the 

 stamens are moved by the insect touching the 2-spurred ones and 

 that the proboscis of the insect is thus covered with pollen. We 

 would, therefore, conclude that in flowers with a straight style we 

 haV'C plants that are less differentiated and not so old in point of 

 development. In this connection it is necessary, however, to con- 

 sider also the development of hairs upon petals, stamens and 

 stigma. Were these developed to impede, likewise, the progress 



