00 : 



out the greater part of the life of the flower, lu cultivated 

 specimens in Switzerland. Gothart observed the same order of 

 development: autogamy always ultimately set in. 



Self-pollination appears to take place. In Fig. 35 C 

 the styles are somewhat spreading and the stigmas ripe: 1 — 2 

 antipelalous stamens were still erect and had their anthers 

 close to the stigmas. None of the flowers with well deve- 

 loped stigmas and far-spreading styles (Fig. 35 D\ had .their 

 anthers close to the stigmas. 



Fig. 36. Saxifraga iricuspidata. 



A, Inflorescence; somewhat reduced. B, Small flower (about' i) from West Greenland (Aug. 5, 

 1884); the anthers of three of the stamens are open. C, The same in longitudinal section. 

 D, Petal of the same. E. F. Style and stigmatic papillae of the same. G. A pistillate flower 

 from West Greenland. B, J, Stamen and pollen-grains of the same. K. L, Stamen and 

 poUen-grains of a hermaphrodite flower, magnified as B and J. Drawn by E. W., 1908.^ 



Fruit usually ripens in West Greenland. Parry gathered 

 ripe fruit even at Duke of York Bay. 



Forma micrantha Sternb. Pistillate flowers. A small- 

 flowered form approaching this, was found at Holstensborg 

 (July 13i. Another (Aug. 5» is shown in Fig. 36 B—F. The 

 stamens are unusually short, and the anthers very small. Three 

 of the anthers (marked xi were open and appeared to be normal. 

 The stigmas were normaland one pollen-grain occurred upon them. 



The form shown in Fig. 36 B — F appeared to be on the point 

 of becoming pistillate. This is more decidedly the case with 

 the form shown in Fig. 36 G — J, the anthers of which were 

 unusually small, not open, and filled with angular, and evidently 

 not normal, pollen: for comparison, the anther and pollen of 



XXXTL 15 



