Annual or perennial, glabrous to subglabrous, usually succulent herbs, rarely undershrubs. Leaves opposite or alternate, simple, entire, fleshy; stipules scarious or modified into hairs, rarely absent. Inflorescence usually a terminal cyme or raceme sometimes flowers solitary. Flowers bisexual, actinomorphic, usually hypogynous rarely perigynous, bracteate or ebracteate. Sepals 2, rarely more (often regarded as bracteoles), free or connate at the base, anterior overlapping the posterior, imbricate, caducous. Petals 4-6, (rarely 2 or 3), free or united at the base, imbricate, entire, fugaceous or marcescent. Stamens from 3-many, antipetalous, free, rarely basally connate or epipetalous; filaments filiform, anthers dithecous, introrse, dehiscing longitudinally. Ovary 2-8-carpelled, syncarpous, superior or rarely semi-inferior, free or partially adnate to the calyx, unilocular; placentation basal or free central; ovules 2-many, campylotropous or amphitrop¬ous, often with long and persistent erect funicles; styles 2-8, united, rarely free, stigmas 2-8, elongated. Fruit a membranous or crustaceous, loculicidally or circumscissilly dehiscent, (sometimes opening explosively) capsule, rarely an indehiscent nut. Seeds 2-many, compressed, globose-reniform; endosperm copious and enclosed by more or less coiled embryo.
19 genera and nearly 350 species; widely distributed in warmer and temperate regions throughout the world, but is especially centred in South and Pacific North America. One genus and 4 species are reported from W. Pakistan.
Authors differ on the interpretation of floral parts. Morphologically and anatomically perianth is regarded as uniseriate; sepals are interpreted as bracts, and petals are regarded as perigone composed of tepals. Perianth is here described as biseriate for taxonomic convenience.
Acknowledgements: We are grateful to the United States Department of Agriculture for financing this research under P.L. 480. Thanks are also due to Dr. S.M.H. Jafri (Karachi), Mr. B.L. Burtt, Mr. I.C. Hedge and Miss J. Lamond (Edinburgh) and Dr. F.R. Fosberg (Smithsonian) for their helpful suggestions.
