NOTES ON TRILLIUM 331 



states, that the ovary is obviously unilocular, the three placentas 

 being sessile on the ovary-wall, and not, as usual in the genus, 

 projecting more or less nearly to the centre of the chamber. The 

 artist in the ParacUsiis has not drawn the placentas, though they 

 are indicated in Salisbury's original sketch. 



Mr. T. Smith, of Newry, who has kindly sent me living speci- 

 mens of nearly all the cultivated species, tells me that " there are 

 two forms of T. fjrandijJorum, one which grows in bogs and one on 

 dry soil," and says that " we generally have to lose them in order 

 to find out which is which." I have not had the opportunity of 

 comparing these two forms. 



A specimen in Herb- Kew. from Goat Island, Nicaragua, has 

 leaves with a stalk 1 cm. long ; and there is also a monstrous form 

 from Syracuse, New York (from herb. Gray), with leaf-stalks as 

 much as 3 cm. long. 



11. T. ovATUM Pursh, Fl. Amer. Sept. i. 245 (1814) ; Watson, /. c. 

 T. calif oniiciim Kellog in Proc. Calif. Acad. ii. (1863), 50, fig. 2. 

 T. grandiflorum Hook. Flor. Bor. Am. ii. 180, in part, i.e. as 



regards the Columbia locality. 



T. obovatum Hook. /. c. non Pursh. 



The western representative of T. (jrandiflorum, from which it is 

 distinguished by its generally narrower and lanceolate petals. The 

 flowers are on an average smaller, but considerable variations in 

 size occur. Watson says (Proc. Amer. Acad. xiv. 274), "stigmas 

 somewhat stouter and more recurved," but in my limited experience 

 I do not find this a general character. 



It is often difficult to distinguish T. ovatumivomT. fjraudifiorum 

 in the herbarium. Mr. Smith, however, says that in the garden 

 this difiiculty does not exist, as T. ovatum " opens its flowers first, 

 and immediately after emerging from the ground, and before the 

 leaves have developed, the stem afterwards lengthening until the 

 flower dies ofl' purple, while T. grandijiomm grows to its full height 

 before the flower shows at all." 



Note.— In Erythea, vii. 104 (1899), Mr. C. V. Piper describes a 

 new western species, T. cmssifoliinu, allied to T. ovatum, but 

 differing " in its erect rhizomes, shorter petals, and thickish 

 differently shaped leaves." The petals are 2-2-5 cm. long, white 

 or pinkish in colour, broadly to rather narrowly lanceolate, acumi- 

 nate or acute, and scarcely as long as the narrowly or broadly 

 lanceolate acute sepals. The locality is "on damp hillsides, foot- 

 hills near Wenatchee." 



12. T. Rugelii, sp. nov. Herba robusta glabra, foliis late 

 rhomboideis subsessilibus breviter acuminatis ; flore pedunculate 

 e foliorum medio cernuo ; sepalis patentibus lanceolatis obtusis ; 

 petalis patentibus rotunde-ovatis, sepalis sub^quilongis ; antheris 

 linearibus quam filamenta plus triplo longioribus, et stigmata crassa 

 brevia et recurvata valde excedentibus, ovario sph^rico sex-sulcato. 



A fine plant springing from an apparently horizontal root- 

 stock. Stem pale brown, very wrinkled and somewhat straw-like 

 when dry, 40 cm. long by 5-6 mm. thick above the base, becoming 

 much slenderer in the upper third. Leaves 12-13 cm. long and 



