24i MB. A. W. I}E>'NETT ON PABKASSIA PALLSTRIS, 



/ 



Affinities 



By Alfbed W. BenI^ett, M.A., B.Sc, F.L.S. 



[Read November 19, 1868.] 



The true position of Farnassia lias been a source of much doubt 

 and variety of opinion among botanists, having been placed by 

 authors of acknowledged repute among HypericacesD, Droseracese, 

 Saxifragacese, and constituting an order by itself, Parnassiacese. 

 The chief advocates of its place among the Hypericacese were 

 Don and Lindley. It is singular, however, that of the characters 

 which Lindley gives in his ' Vegetable Kingdom ' as those by 

 which St. John's Worts may be recognized, viz. the axile pla- 

 centatiou, and the polyadelphous stamens, together with the long 

 style, the unequal-sided petals, and the opposite dotted leaves, 

 not one applies to Parnassia^ the affinity being founded entirely 

 on the exalbuminous seed, and on a fancied analogy between the 

 polyadelphous stamens of JSypericicm and the glandular scales 

 which constitute the nectary of Parnassia. With Droseracese, 

 under which order the genus is placed by Babington and most of 

 the older English botanists, the affinities of Parnassia consist 

 mainly in the unilocular ovary, terminating in several stigmata, 

 the parietal placentation, the extrorse anthers, and the marcescent 

 petals. Bentham and Oliver unite Droseraceae with Saxifragacese, 

 an alliance not recognized by the older botanists. Without pre- 

 suming to express an opinion opposed to that held by such high 

 authorities, I may point out the following important diffi^reuces in 

 structure between Saxifraga and Parnassia : — In Saxifraga the 

 capsule is bilocular, the styles never more than 2, the placenta- 

 tion axile, and the anthers introrse ; in Parnassia the placentation 

 is parietal (Dr. Hooker finds no signs of any deviation from this 

 structure in any of the Himalayan species he has examined), the 

 capsule is unilocular, the styles 3 to 5, and the anthers extrorse. 

 In transferring Pflrwa.9^/a to Saxifragacese, I cannot help doubt- 

 ing whether too much force has not beep, given to the peri- 

 gynous character of the stamens, as, if that is insisted on as a 

 material point, Parnassia must be entirely removed from Drosera^ 

 with which genus all botanists seem to agree it has very close 

 relationship, and which has the stamens truly hypogynous, at 

 least in our European species. Indeed a strict carrying out of 

 this test would necessitate the division of Drosera itself into 

 widely separated orders; for St.-Hilaire describes Brazilian 



