A NEW MIDLAND VIOLET 89 



I do not wish even to hint at any general conclusions regarding 

 the phenomenon as supposed to occur in violets generally. 



It has been asserted on the one hand that the so called cleisto- 

 gamous or apetalous flowers of violets though they may have 

 rudimentary stamens, are entirely devoid of pollen and that there- 

 fore no fertilization whatever takes place, that is, in other words 

 the ovules in these aestival flowers develop parthenogenetically. 

 Others have affirmed that fertilization occurs, therefore pre- 

 suming that pollen or its equivalent is formed and that close ferti- 

 lization takes place without the opening of the flowers at all. 



Whatever examination has been made of other apetalous 

 flowers we need not discuss here. In examining these flowers of 

 the new plant for diagnosis we found that two anther bearing 

 stamens are always present. The anther sacs are two to four in 

 number, in the latter case the two nearest the pistil larger. None 

 of them are, however, never more than one-fifth as large as the 

 anthers of the petaliferous flowers. The pollen grains though com- 

 paratively scanty are more than sufficient to account for the 

 fertilization of the numerous ovules. Pollen need not be as copious 

 as in the petal-bearing spring flowers for the proximity of the 

 two stamens pressed against the stigma obviates waste of pollen 

 which in the closed flower can not be readily lost. 



I have found a considerable number of such flowers just in 

 the stages in which the enlarging pistil apparently by friction of 

 of growth upwards together with the compression of the sepals 

 on the outside, causes the rupture of these apparently abortive 

 but only more minute anther sacs of the two closely oppressed 

 stamens. The pollen was found in these instances in various stages 

 of germination of pollen tubes, nearly all of which led to the 

 opening in the stigma of the curved or hooked style (See Fig. 3) 

 the progress of the pollen tubes farther in was not made. A more 

 exact histological study would be necessary. Even after tracing 

 the tubes into the ovary, cytological study of the behavior of the 

 chromosomes in the oospores alone would serve as final evidence 

 when fertilization had actually taken place or not. It would 

 seem that the problem would be an interesting and not perhaps 

 a difficult one for general and more exact study. 



Developing pollen tubes were found in apetalous flowers 

 which were as a rule not over 4 mm. long and not less than 2.5 

 mm. In larger flowers the stamens have dried up completely and 



