1889. J BOTANICAL GAZETTE. 1 77 



than up. The effect of the upward curvature of the spur is 

 well illustrated in the behavior of Xisoniades, which invaria- 

 bly turn head downwards. 



It has often occurred to me that Bombylius could suck th< 

 bearded violets more easily than the insects which reverse, 

 and that under certain conditions it might take possession of 

 them, as Anthophora and Bombus have clone in the case of 

 V. pedata. In this connection, it is interesting to refer to 

 Miiller's observations on V. calcarata. He saw Macroglossa 

 stellatarum visit 194 flowers of this violet in 6f minutes. So 

 rapid a visitor might easily take possession of any flower 

 which suits its fancy. V. calcarata appears to have become 

 completely adjusted to a"new set of conditions, for its spur, 

 as shown in Miiller's figures, turns downward, and not up- 

 wards as in V. pedata. 



__ Virginica L.— The proterandry of this flower 

 was first recorded by Bessey (87). He concludes that the 

 adaptations are to favor cross-fertilization and to prevent self- 

 fertilization, and my observations confirm his view. On the 

 other hand, Meehan, in a paper on " The « Sleep of Plants ' 

 as an Agent in Self-fertilization" (485), regards it as com- 

 monly self-fertilized by closing of the petals. This mode ot 

 self-fertilization was discussed by Ch. Fermond 5 in 1859, but 

 C^Virginica is a very poor example of it. Indeed, it is most 

 erroneous to suppose" that it is commonly self-fertilized in this 

 way, for it is one of the most marked cases of proterandry in 



native plants. 



On the first day the stamens stand in the center of the 

 flower and the anthers discharge their pollen, but the lobes 

 of the stigma remain closed. The flower is visited very 

 abundantly bv insects which suck up the honey and which 

 eat or collect the pollen, so that by the time the flower closes 

 the pollen is commonly all removed. On the second day the 

 stamens are bent over, holding the empty anther against the 

 petals. The stigma lobes are now separated, and the flower 

 is in the second or female stage. If self-fertilization by clos- 

 ing of the flower occurs, it is after the anthers have been ex- 

 posed to insects for two days and the stigma for one, but 

 many flowers which I marked exposed their stigmas again 

 on the third day, showing that fertilization of any kind had 

 failed on the clay before. The flowers are therefore male on 

 the first day and female on the second and sometimes on the 

 third. 



6 See Just's Bot. Jahresbericht, IV, 939. 



