Chap, IV. OXALIS. 169 



OxALis (Geraniaoe^). 



In 1863 Mr. Roland Trimen wrote to me from the 

 Cape of Good Hope that he had there found species of 

 Ozalis which presented three fonns ; and of these he 

 enclosed drawings and dried specimens. Of one species 

 he collected 43 flowers from distinct plants, and they 

 consisted of 10 long-styled, 12 mid-styled, and 21 

 short-styled. Of another species he collected 13 flowers, 

 consisting of 8 long-styled, 7 mid-styled, and 3 short- 

 styled. In 1866 Prof. Hildebrand proved* by an ex- 

 amination of the specimens in several herbaria that 20 

 species are certainly heterostyled and trimorphic, and 

 51 others almost certainly so. He also made some in- 

 teresting observations on living plants belonging to 

 one form alone ; for at that time he did not possess 

 the three forms of any living species. During the 

 years 1864 to 1868 I occasionally experimented on 

 Oxalis speciosa, but until now have never found time 

 to publish the results. In 1871 Hildebrand published 

 an admirable paper f in which he shows in the case of 

 two species of Oxalis, that the sexual relations of the 

 three forms are nearly the same as in Lythrvm sali- 

 caria. I will now give an abstract of his observa- 

 tions, and afterwards of my own less complete ones. 

 I may premise that in all the species seen by me, the 

 Stigmas of the five straight pistils of the long-styled 

 form stand on a level with the anthers of the longest 

 stamens in the two other forms. In the mid-styled 



uppermost anthers. The numerous * 'Monatsber. der Akad. der 



stamens are gjaduated in length, Wiss. Berlin,' 1866, pp. 352, 372. 



and the pollen-grains from the He gives drawings of the three 



longest and shortest ones did not forms at p. 42 of his 'Geschlechter- 



prefcent any marked diiferenoe in Vertheilung,' &c., 1867. 



diameter. Therefore these species t ' Bot. Zeitung,' 1871, pp. 416 



do not appear to be heterostyled. and 432. 



