20. TILIACE^. [i. Okewu. 



rarely 0, free, usually imbricate. St. (sometimes few in 

 Triumfetta aud Corchorus) free or sometimes 5-adelphoas.but 

 not united into a tube, often on a gonophore ; anthers 2-celled. 

 Ovary 2-10-celled. Ovules anatropous. Frt. various, often 

 drupaceous or deeply lobed. Seeds 1-many, exarillate, usually 

 albuminous. Embryo straight or slightly curved, 



A. Aethers globose or oblong, opening by slits. 



Trees or shrubs. Petals usually with a glandular 



base. Fruit drupaceous, smooth . . . . 1. Grewia. 

 Shrubs or herbs. Fls. in dense cymes. Pet. eglan- 



dular Frt. dry echinate 2. Trxu,njetta. 



Annuals. Peduncles 1-3-flowered. Frt. capsular . . 3. Corchente. 



B. Anthers linear, opening by pores. Fr** drupaceous. 



Treea 4. Elwocarpvu. 



1. Grewia, L. 



Trees, shrabs or rarely undershrubs with stellate pubescence, 

 simple 3- 7 -basal-nerved serrate or serrulate leaves aud yellow, 

 rarely white, flowers in axillary (not panicled in C. N. species) 

 sessile or stalked umbels. Petals shorter than the sepals, the 

 base usually occupied by a large gland with a pubescent rim.* 

 St. numerous on a short gonophore (but see- Note), Ovary 

 2-4-celled. Style 1 with 2-4 spreading stigmas or multifid 

 peltate stigma, Ovules 2-several in each cell. Fruit often 

 lobed of 1-4 pyrenes enclosed in a succulent or ultimately 

 fibrous mesocarp. Pyrenes 1-2-seeded, Seed albuminous 

 with large flat thinly-fleshy cotyledons. 



The number of species is greater according to some authors than 

 those here retained, esp. in the asiatica series. The extreme forms of 

 these variable groups can no doubt be easily distinguished, but the way in 

 which others have been repeatedly changed about from one cover to 

 another in herbaria by those who maintain their distin "tness a3 species 

 shews how many intermediate and connecting forms exist. Whether the 



*Note.— The glandular area at the base of the petals may be absent/ a* 

 was nated by Sir D. Brandis, who founded the species leptopctala on this 

 character. The absence of the gland is, I find, always correlated with 

 the reduction of the gonophore and I hold the character to, be variable in 

 certain species. 



193 



