standley: north American amaranthaceae 393 



as separate tribes, Lagrezia must be therefore placed in the latter. In 

 Dr. Schinz's key to the tribes there is no means of telling to which of 

 them a plant with a 1-seeded utricle should be referred. In 1895 Dr. 

 J. N. Rose described a new plant from Manzanillo, Mexico, which 

 he called Celosia f monosperma.^ This should be referred to the genus 

 here discussed, and maj^ be known as Lagrezia monosperma. The 

 few other species of the genus are natives of Madagascar and southern 

 Africa. 



2. Chamissoa is represented in tropical North America by two 

 species, the widely distributed Chamissoa altissima, and C. maximiliani, 

 known within our limits only from Costa Rica. Chamissoa macrocarpa 

 H. B. K. has been reported frequently from the West Indies and Cen- 

 tral America, but the specimens so determined are C. altissima. 



3. Amaranthus seems to have its center of distribution in the south- 

 western United States and northern Mexico. A large number of species 

 are found in the somewhat similar region of Argentina, a number that 

 doubtless will be increased when that country is better explored botani- 

 cally. About 40 species are known from, North America. Most of 

 these are common weeds of cultivated land, but several species are 

 loiown only from the southwestern mesas and foothills. 



Several segregates from Amaranthus have been proposed by different 

 authors, notably Mengea, Euxolus, and Scleropus. With our present 

 knowledge of the group it seems impossible to maintain any of these 

 genera, for the characters depended upon to separate them will not 

 hold when all the species of the genus are taken into consideration. 



4. Acnida is a wholly North American genus. Five species are 

 found in salt marches along the eastern and southern coasts of the 

 United States, on the southwest coast of Mexico, and in the West 

 Indies. Three others occur in the central and southwestern United 

 States. The genus is very closely related to Amaranthus, differing 

 only in the absence of a perianth in the pistillate flowers. Acnida 

 tuberculata is so closely allied to Amaranthus torretji that it is practically 

 impossible to distinguish staminate plants of the two species, whose 

 ranges largely overlap. 



5. Acanthochiton consists of a single species, a native of the sand- 

 hills of western Texas, New Mexico, Arizona, and northeastern Mexico. 

 It is distinguished fi'om Acnida only by a vegetative character — the 

 large size and peculiar form of the bracts ; but it has always been accepted 

 as a valid genus. 



* Contr. U. S. Nat. Herb. 1: 352. 



