^O IVEOBOr. 



and Georgia, nearer to M. cyananthera than 

 M. angustlfolia, but leaves quite slender uni- 

 nerve smooth, anthers also bluish — thus we have 

 6 sp. of this pretty G. a 7th might be my 3'L 

 pumila nov>' my G. T lerolepta Neog. 1825. 

 EIJPATORIUM and akin 



This extensive G. is now divided into many 

 Celestina, BrickeUia., Mikania^ Kalinia &c, 

 besides my Eutrocliium with scariose terete 

 perianthe, leaves verticillate kc, my Caradesia 

 with perianthe hemispherical multiflore : thus 

 reduced to the sp. with a definite number of 

 florets and segments of perianthe, it is still pro- 

 lific. The American Cacalias are very near, 

 but form 3 peculiar Genera, my Epatitis 205, 

 Mesadenia and Hasteola. 



935. MESADENIA Haf. perianthe 5flore, 

 terete with 5 equal segments, phoranthe with a 

 thick gland in the centre and 5 flowers around 

 it, (this gland is an abortive neutral floret some- 

 times evolved) calicule obsolete or only a small 

 scale, florets with tubular limb deeply 5d, an- 

 thers inclosed, stigmas 2 or 4 hardiy exerted, 

 seeds smooth oblong or clavate, pappus downy 

 capillary. Peremdals^ leaves alternate large 

 thicklsh, Jlowrrs white corymhose or panicu- 

 late. — A north American G. of many sp. quite 

 distinct and natural by the singular structure of 

 phoranthe, noticed by Elliot and Nuttul, the 

 central gland is variable ovate acute, or angu- 

 lar tridentate, or seldom changed into a neutral 

 floret which has then the ovary dentate instead 

 of downy. I have even noticed some flowers 

 where there was only this neutral floret in the 

 perianthe, lacking anthers and style, another 

 anomaly is the quadruple stigmas of some florets, 

 This G. is nearer Kleinia than Cacalia. Akin 



