352 



at which the number is reckoned. This will account for the dis- 

 crepancy which occurs in the statements of different observers. The 

 large roots are spongy or often conspicuously nodose for their 

 whole length. With three of the genera we have no concern, as 

 two of them, PJiilodice and Toiiina, both together numbering five 

 species, are restricted to tropical South America, and the other, 

 Mcsanthemum, numbering three species, is endemic in tropical 

 Africa. 



The North American genera may be briefly distinguished as 

 follows : 



Segments of the perianth four or six. 

 Stamens separate. Anthers two-celled. Stamens as many as 



the perianth segments. i. Eriocaitlon. 



Stamens one-half as many as the perianth segments. 



2. Diipatya. 

 Perianth of three segments. Stamens three, monadelphous below. 



Anthers one- celled. 3. LacJinocaulon. 



Eriocaitlon is the most extensively diffused genus, being found 

 in tropical and subtropical regions throughout the world. As 

 classified by Kornicke in his monograph the species are divided 

 into fourteen sections. All the species occurring within the bor- 

 ders of the United States, so far as known, are acaulescent or 

 nearly so, the heads single on erect peduncles or scapes, the peri- 

 anth with one exception four-parted and the stamens four ; ihe 

 Mexican species are the same except in having six-parted flowers 

 and six stamens. The perianth segments, at least the upper ones, 

 are usually spotted with a minute black gland near the centre or 

 the apex. The heads are generally quite villose and grayish in 

 appearance, the parts of the perianth being strongly bearded. The 

 flowers are each subtended by a bract quite similar in markings 

 and general appearance to the perianth segments. Seeds oval, 

 brown when mature and, under the lens, covered with blunt or 

 spiny protuberances. 



As the perianth segments are in two series and often separated 

 at a considerable distance, there is much variation in the language 

 applied to them by botanists. Kornicke calls the floral envelopes 

 a double perigonium, the exterior calyculate and interior sub- 

 coroUine. Kunth speaks of them as a double calyx, while others 



