standley: north American amaranthaceae 395 



ern South America. Different authors have proposed very different 

 divisions of genera for tlie group, but the characters used often have 

 been vegetative or else they proved unrehable as additional species have 

 been discovered. Martins, who was perhaps the most careful student 

 of the group, proposed a large number of genera, based chiefly upon 

 Brazilian species, but scarcel)^ any of his genera are recognized today. 

 The treatment here proposed follows closely that suggested by Otto 

 Kuntze,^ which was later adopted by Dr. Schinz, 4ind very recently 

 by Dr. Stuchlik.^ It is based almost wholly upon flower structure, 

 and chiefly upon the characters of the gynoecium and androecium. 

 The North American genera are ten in number, as follows. 



1. Cladothrix is closely related to both Gossypianthus and Achy- 

 ranthes and is distinguished principally by the form of the inflores- 

 cence, the flowers being glomerate rather than spicate or capitate. 

 Three species occur in the southwestern United States and in northern 

 Mexico. 



2. Gossypianthus is wholly North American, four species being 

 known. One is confined to Cuba, a second is common to Hispaniola, 

 the southwestern United States, and northeastern Mexico, while the 

 other two are found in Texas and Oklahoma. 



3. Pfaffia consists of a large number of South American species of 

 diverse habit, some of them resembhng the better known species of 

 Gomphrena, while others are tall shrubs or vines. In Mexico and 

 Central America there is a single species which closely resembles some 

 of the species of Iresine. Hebanthe of Martins must be considered a 

 synonym of Pfaffia, although not all the plants described under He- 

 banthe are true Pfaffias. One plant described by Hemsley as a He- 

 banthe is to be referred to an older species of Iresine. The proper 

 place of two other Mexican species described by Hemsley is still un- 

 certain. Dr. Watson in 1883 applied the name Hebanthe palmeri^ to 

 another Mexican plant which is not a Pfaffia, but should be known as 

 Iresine palmeri. 



4. Achyranthes has been discussed by the writer very recently^" 

 and need not be treated further here. 



5. Woehleria is one of the so-called ''monotypic" genera. It is 

 endemic in Cuba and is apparently very rare. In general appearance 



' Rev. Gen. PI. 2: 534-545. 1891. 

 8 Repert. Sp. Nov. Fedde 12: 350-359. 1913. 

 5 Proc. Amer. Acad. 18: 144. 

 "> Journ. Wash. Acad. Sci. 5: 72-76. 1915. 



