TROPICAL AMERICAN COMPOSITAE. 35 



Santander, Colombia, 16 Jan. 1918. These correspond closely in all 

 essential and most minor features with the original material of E. 

 Squircsii from the delta of the Orinoco and certainly appear to be 

 conspecific ^Yith it. The only differences found during a rather 

 detailed examination were that the leaves of the Colombian plant 

 were slightly firmer in texture and even more shortly petioled or 

 subsessile. These differences are precisely of the kind and degree 

 that occur very frequently between exposed and shade forms of 

 the same plant. The range of the species is thus extended some 

 1500 km. and over the watershed from the Orinoco Valley into 

 that of the Magdalena. However, there is little difference in the 

 latitude, and the habitat, low alluvial thickets, is similar. Renewed 

 examination of the plants of this group brings out what had not 

 been previously noticed, namely, the affinity between this species 

 and E. turbacense Hieron. Of the latter species the writer has not 

 seen the type, Stiibel's no. 51, collected at Turbaco, Dept. Bolivar, 

 in the delta region of the Magdalena. However, Hieronymus when 

 publishing upon the plants of Lehmann (Engl. Bot. Jahrb. xxviii. 573) 

 identifies with his E. turbacense Lehmann's no. 5971, and of this a 

 leaf and a bit of the inflorescence were received at the Gray Herbarium 

 some years ago in an exchange from the Royal Botanical Garden at 

 Berlin. This plant of Lehmann's was collected on the Rio Ortega 

 in the Dept. Cauca, that is to say, some 900 km. to the south of the 

 original station. If Hieronymus has been right in referring it to his 

 E. turbacense, the following differences may be pointed out between 

 that species and the later E. Squiresii. In E. turbacense the leaves 

 are entirely glabrous above, while in E. Squiresii they are puberulent 

 at least on the midnerve and sometimes perceptibly so on the surface 

 as well; in E. turbacense the lowest two pairs of lateral veins leave the 

 midnerve at an angle of about 40° and in length considerably exceed 

 those arising at a greater distance from the base, while in E. Squiresii 

 the lower pairs of veins are no longer, indeed are usually shorter than 

 some of the others, and all of them leave the midnerve at a considerable 

 angle, usually at about 70°. In E. turbacense the involucre is cam- 

 panulate, in E. Squiresii it is somewhat longer and campanulate- 

 subcylindric (in the fresh state) or campanulate-subturbinate (in 

 the dried state). In E. turbacense the pubescence of the pedicels is 

 distinctly longer and more sordid-tawny than in E. Squiresii. The 

 two species are certainly very close. In both the lower leaves have a 

 peculiar form, the lance-oblong blade being narrowed at the base into 

 a more or less elongated portion like a broadly winged petiole. It 



