396 standley: north American amaranthaceae 



the plant is similar to some species of Achyranthes, but the androecium 

 consists of only a single stamen, and the stigma is bilobate rather than 

 capitate. 



6. Gomphrena is represented in North America by some fifteen spe- 

 cies. Thirteen of these are low plants, with usually large, sessile, and 

 globose heads, closely related to the cultivated globe amaranth, Gom- 

 phrena globosa. The other two are tall plants with narrow, long- 

 pedunculate, cylindric heads, closely simulating certain forms of Achy- 

 ranthes. A large number of species of Gomphrena have been described 

 from South America, no less than 66 being reported from Brazil 40 

 years ago. 



7. Iresine is one of the most interesting genera of the Amaranthaceae, 

 chiefly because the plants are of an attractive rather than a ''weedy" 

 appearance. About 30 species are known within our limits. The 

 segregates Trommsdorffia and Rosea were proposed bj^ Martins, but 

 it seems impracticable to maintain them as distinct genera. 



8. Dicraurus is distinguished from all other genera of the Gom- 

 phreneae by the alternate leaves; otherwise it is too closely related to 

 Iresine, and, indeed, it seems probable that ultimately it may be 

 united with that genus. Two species are known, D. leptodadus and 

 D. alter nijolius. The first, a plant of western Texas and northeastern 

 Mexico, has all its leaves alternate; but in the second, a native of 

 Lower California, while most of the leaves are alternate, the lower 

 ones frequently are opposite. 



9. Lithophila was based by Swartz upon a plant which is common 

 on the seashores of the West Indies. Some authors have referred the 

 genus to Iresine, but it may be maintained because of the strongly 

 compressed perianth and 2 rather than 5 stamens. Besides the type 

 species, three others, of somewhat diverse habit, are known, all inhabi- 

 tants of the Galapagos Islands: Lithophila radicata (Alternanthera 

 radicata Hook, f., 1847), L. rigida {Alternanthera rigida Rob. & Greenm., 

 1895), and L. subscaposa {Alternanthera subscaposa Hook, f., 1847). 



10. Philoxerus was proposed by Robert Brown in 1840. The plants 

 of this group have usually been referred to Iresine and Lithophila. 

 Philoxerus seems, however, a valid genus, distinguished from Iresine 

 not only by habit but by the compressed perianth, and from Lithophila 

 by the different structure of the androecium and by the stipitate flowers. 

 Many species of the genus have been proposed, but how many of them 

 are valid is an unsettled question. In North America only a single 

 one is known. 



