Aster. COMPOSITiE. 119 



what hairy) petioles, entire or sparingly serrate, the veins loosely reticulated 

 beneath ; those of the short pubescent branches small and sessile ; heads 

 numerous and rather crowded; involucre campanulate, shorter than the 

 disk ; the scales closely imbricated, lanceolate-linear, rather obtuse ; ache- 

 nia glabrous. — Hook.! {Sf Lindl. !) fl. Bor.-Am. 2. j). 9 (note); Ridclell! 

 synops. I. c. 



Cliffs and banks of streams, throughout Kentucky, Dr. Short ! S^x. and 

 Ohio, Dr. Riddell! Mr. Sullivant'f Dr. Paddock! S^x. Mountains of 

 Georgia, Mr. Buckley ! Arliausas, Nuttcdl ! Sept.-Oct. — A beautiful spe- 

 cies (deservedly dedicated to the well-known botanist who has so greatly 

 contributed to our knowledge of the plants of the Western States), 2 to 4 feet 

 high, remarkable for its lanceolate-cordate, petioled, often slightly falcate, 

 rather membranaceous leaves ; which vary from 3 to 5 inches in length, 

 smooth and somewhat shining above, pale and puberulent (but scarcely, if 

 at all scabrous) beneath ; the lower ones often serrate or toothed tOAvards the 

 base; those of the branches oval or oblong, of the ultimate branchlets or 

 peduncles minute and subulate. Heads showy (usually larger than in A. 

 undulatus), racemose at the summit of the stem or on the short branches, 

 often forming a thyrsus. Scales of the involucre appressed, minutely 

 pubescent, whitish, with oblong green tips. Rays violet-blue, lanceolate : 

 disk yellow, changing to purple. Pappus brownish or tawny. 



29. A. undulatus (Linn., Ait.) : pale with a close and cinereous often 

 scabrous pubescence ; stem paniculate or racemose-thyrsoid at the summit ; 

 leaves ovate or ovate-lanceolate, somewhat tomentose-pubescent beneath and 

 scabrous above, acute, with the margins often imdulate or slightly crenate- 

 serrate ; the radical and lowest cauline cordate, on slender slightly margined 

 petioles, which are usually dilated and clasping at the base ; the others ab- 

 ruptly contracted into a short broadly winged clasping petiole ; the upper- 

 most cordate-clasping ; those of the branchlets lanceolate or subulate ; invo- 

 lucre obovoid, nearly the length of the disk ; the scales linear, mostly acute, 

 pubescent, closely imbricated ; achenia slightly pubescent, or at length gla- 

 brous.— Linn, hort. Cliff, p. 408, S^- spec. {ed. 1) 2. p. 875? (not of ed. 2.) ; 

 Ait. KeiD. {ed. 1) 3. p. 206; ''Hoffm. j^hytogr. hi. p. 77. t. C. f. 1;" 

 Parsh, fl. 2. p. 551 ; Nees, Ast. p. 57 ,- Darlingt. ! fi. Ccst. p. 464. A. 

 diversifolius, Michx.! fl. 2. p. 113; Ell.! sk. 2. p. 361; Bigcl.! fl. Bost. 

 ed. 2. p. 312; DC! prodr. 5. p). 234. A. paniculatus, Nutt. ! gen. 2. p. 

 156, not o[ Ait. A. sagittifolius & A. scaber. Ell.! I. c. A. patens, Bart, 

 compend. fl. Phil. 2. p. 113. A. heterophyllus y. Nees, Ast. p. 55? A. 

 autumnalis & A. heteromallus, Wender. 1 ex Nees. (The leaves vary in 

 size, from 1 to 3 or 4 inches ; in shape from broadly ovate to nearly lanceo- 

 late, the radical often obtuse, the primordial roundish ; and in pubescence, 

 being sometimes minutely and softly tomentose on both surfaces, and fre- 

 quently scabrous beneath as well as above.) 



/3. stem strict ; panicle simple ; the heads rather larger ; cauline leaves 

 elongated oblong or lanceolate. 



Dry woodlands, nearly throughout the United States ! (3. Kentucky, Dr. 

 Short ! Sept.-Oct. — Stem 1-3 feet high, often purplish ; the flower-branch- 

 es spreading, rather rigid, with very small (often subulate) leaves. Radical 

 and lower leaves remotely crenate or serrate, or, like the upjK'r, merely un- 

 dulate. Heads middle-sized, loosely disposed on the branches in a some- 

 what racemose manner, often unilateral, all pedicellate ; or rarely somewhat 

 glomerate. Scales of the involucre nearly membranaceous, with oblong or 

 somewhat lanceolate green tij^s, acute or acuminate, ciliate. Rays pale vio- 

 let-blue : disk yellow turning purple. Pappus becoming tawny or brown- 

 ish. — That the var. /3. is merely a peculiar state of this species, is evident 

 from connecting specimens. The original A. undulatus {Linn. hort. Cliff. /) 

 is not A. patens, but we believe belongs to this species, which must in any 



