218 Engelmann and Gray, 



same as the Texan plant referred by Spach to G. tripetala, 

 Cav. ; but it does not agree with the figure of Cavanilles, nor 

 exhibit any trimerous flowers. 



62. Eryngium coronatum, Torr. $• Gr. Fl. I. p. 604. 

 Bottom woodlands on the Brazos. August. 



63. Cynosciadium pinnatum, DC. p? pumilum. Differs 

 from the larger and erect form (which is usually a foot or two 

 in height.) in its low and diffuse stems, its umbellets with only 

 four or five rays, and few or no involucral leaves. Prairies, 

 Galveston. April, May. 



64. Leptocaulis echinatus, Nutt. Galveston Island. 



65. Discopleura capillacea, DC. Galveston. May. 



66. Spermacoce glabra, Michx. Near Houston. Sept. 



67. Mitreola petiolata, Torr. &f Gr. Swampy thickets 

 west of Houston. 



68. PoLYPREMUM PROCUMBENS, Liilll. Houston. June. 



69. Hedyotis Bosch, DC. Houston. May and June. 



70. Vernonia angustifolia, var. y Torr. &f Gr. Wet 

 prairies west of the Brazos. July. 



71. Liatris elegans, Willd. Houston to the Brazos. 



72. L. acidota. =. L. mucronata, Torr. &,- Gr. Fl. II. p. 

 70, non DC. Houston to the Brazos, in wet praries. Au- 

 gust, September. In the Flora of North America, this species, 

 which is apparently common in Western Louisiana and Texas, 

 was hesitatingly referred to L. mucronata, DC, from the 

 character of which it differs in some respects, principally in 

 the form of the involucral scales. But among Lindheimer's 

 plants, some specimens of what is no doubt the true L. mu- 

 cronata, DC. occur, (which have been distributed in some 

 sets, probably mixed with L. acidota,) and which render it 

 clear that the present is a different, although very nearly 

 allied species. We have accordingly given a new name. 

 The diagnosis may be expressed as follows ; the habit, foliage, 

 &c. being nearly the same in both ; and the involucral scales 

 more or less ciliate when young. 



L. mucronata : capitulis in spicam strictam arete digestis ; 



