233 



IV. Staminate flowers occur rarely (S. oppositifolia), 

 but pistillate flowers appear to be common in a great 

 many species, e. g., in S. aizoides, cennia, groenlandica, Hir- 

 culus^ nivalis, opjjositifolia, rivularis, stellaris, and tricuspidata. 

 1 also observed them in cultivated specimens (Hortus Ilaf- 

 niensis) of S. groenlandica, S. Cijmbalaria, S. Bocheliana, S. 

 moschata var. glandiilosa and, to all appearance, in S. cotyledon. 

 The pistillate flowers are smaller than the hermaphrodite ones; 

 stamens are always present, but are smaller than usual, the 

 anthers especially being small; pollen-grains are sometimes 

 developed, but smaller than usual, and imperfect. In some 

 cases the anthers dehisce, in others they do not. It appears 

 especially to be the terminal flower in the inflorescence which 

 thus develops. 



In some species, deformed flowers with small petals have 

 been observed — in systematic works named ^'cnjptopetala' 

 — e. g. in S. cernua, S. groe)Uandica, and S. stellaris. In some 

 cases the petals at the same time that they were becoming 

 small, were in the act of developing anthers at their apices. 



Other numbers in the flowers than the normal five with 

 two carpels have been observed; for instance, 6 and 7-merous 

 flowers in S. cernua. It is especially the terminal llower in 

 the inflorescence which shows a tendency to increase in 

 the number of the carpels (probably because that flower is 

 better nourished and therefore becomes larger and has room 

 for a greater number of carpels). Thus, trimerons pistils 

 have been found in S. aizoides, groenlandica, hieraciifolia, 

 stellaris, and tricuspidata, 4- and 5-merous in <S. oppositifolia. 

 The terminal flower in the inflorescence may also be seen to 

 differ in another point from the other flowers, viz. in the fact 

 of its being far in advance of the others in regard to develop- 

 ment, e. g. in S. Aizoon, and among cultivated species, in 

 S. geranioides and rotundifolia. In S. Geum I found the pistil 

 of the terminal flower to be formed somewhat differently from 



