302 CLVI. LILIAOE^. (J. D. Hooker.) 



Leaves radical or whorled. Flowers solitary or umbelled. Anthers 



dehiscing extrorsely. 



Leaves radical or subradical ; flowers solitary or umbelled . . 33. CLINTONIA. 



Leaves 3 in a whorl ; flowers solitary 34. TRILLIUM. 



Leaves 4 or more in a whorl . 35. PABIS. 



1. SniXXiAX, Linn. 



Climbing shrubs (rarely erect herbs). Leaves alternate, rarely 

 opposite, persistent, 3-5-nerved and reticulate; petiole usually 2-cir- 

 riferous above the base. Flowers dioecious,s mall, umbelled. Perianth 

 segments 6, free, incurved or recurved. MALE FL. Stamens 6, or more, at the 

 base of the perianth, free ; anthers didymous, oblong, or the cells separated 

 by a forking of the filament. FEM. FL. Staminodes 3 or 6. Ovary 3- 



fonous, 3-celled ; style 0, or short, stigmas 3, stout, recurved ; ovules 1 or 

 in each cell, orthotropous, pendulous. Fruit a globose berry. Seeds 

 solitary, or more often 2 hemispheric, rarely 3 ; albumen horny ; embryo 

 small. Species about 180, temperate and tropical. 



The only authoritative nomenclature for most of the Indian species of Smilax is 

 the great Herbarium of Wallich, in which these are copiously represented, but in several 

 cases almost inextricably mixed. When to this is added the wide differences between 

 the foliage and inflorescence of different parts of the same specimen, and the fre- 

 quent difficulty of matching males with females and flowering with fruiting speci- 

 mens, it is obvious that much confusion must attend any attempt to systematize the 

 species upon any but very complete materials, and such do not exist in any Herba- 

 rium. Unfortunately Kunth, when monographing the genus for his " Enumeratio 

 Plantarum," and M. A. De Candolle for his monograph, had not access to Wallich's 

 Herbarium,and had recourse to duplicates distributed from it to various European Her- 

 baria, which were invariably fragmentary and often wrongly numbered. A study of 

 the Wallichian types has hence obliged me to set aside various determinations 

 of these excellent botanists. Nor can I follow M. De Candolle in his primary and 

 other divisions of the genus. These are founded on the one- or two-ovuled cells of 

 the ovary, and on the inflorescence. That of the ovarian cells is very difficult of 

 observation, is unavailable in the absence of fern, fl., it does not tally with habit or 

 other characters, and I doubt its constancy. That of the inflorescence wants pre- 

 cision, and, as it appears to me, also confirmation by a study of the whole plant, and 

 that it cannot be established on fragments. I am 'far, 'however, from regarding it as 

 valueless because I have been unable to understand and apply it. The arrangement 

 I have adopted is, I think, natural, though far from satisfactory. I have restricted 

 the citations of authors to such as I feel pretty sure of; and, as with so many other 

 Indian genera, I find that any attempt to determine the identity of the species with 

 the insufficient descriptions of the authors of the Malayan Floras is hopeless. 



Sect. I. COILANTHUS, Buds globose. Sepals incurved in flower. 

 Stamens much shorter than the perianth. Ovules (when known) solitary 

 in the ovarian-cells. 



1. S. glabra, Roxb. Fl. Ind. iii. 792 ; leaves 3-6 by H-2i in. elliptic- 

 or ovate-lanceolate acuminate 3-costate, petiole |-f in., sheath - in. long 

 axillary, umbels sub-sessile. Wall. Cat. 5114 ; Bentli. FL HongJc. 369 ; 

 Seem. Bot. Herald. Voy. 420, t. 100; A.DC. Monogr. Smilax, 60. S. 

 Hookeri, Kunth JEnum. v. 162 (excl. syn. Burm. & Lour.}. 



ASSAM, SILHET and the lower KHASIA HILLS, Wallick, &c. TENASSERIM, 

 Gallatly. DISTEIB. China. 



