430 PROCEEDINGS OF THE AMERICAN ACADEMY. 



blunt or bristle-tipped, as well as hairs truly capillary or setiform or 

 awn-like, simple, barbellate, or plumose, rarely clavate or even gland- 

 tipped, all occur within the tribe. 



With rare exceptions it has been found possible to distinguish pretty 

 readily between the truly capillary pappus of such genera as Eupa- 

 fon'iini, Mikania, or Brickcllia and the also terete but stiffer bristle- 

 formed pappus characteristic of Aqriavtkus and of certain species to be 

 segregated from Agcratum. On the other hand it is believed that this 

 bristle-formed pappus can in practically all cases be distinguished from 

 the scale-pappus of the true Ageratums and that it forms a useful 

 basis of separation for some elements long classified with Af/rrafiDu. 

 More difficult is the sharp distinction of plumose pappus from forms 

 in which the setae are merely barbellate, and while this traditional 

 character seems almost necessary at times it has been employed as 

 sparingly as possible. 



A distinction which is more important and which seems never to 

 have been adequately studied even by close students of the Compositae 

 is the difference between a very short cup-shaped, saucer-shaped, or 

 coroniform pappus, such as occurs in Agcrafvm § Coclcstina, and the 

 similar low and often slightly angulate or toothed annidus which often 

 crowns the mature achene in genera like Alovria, which lack all true 

 pappus. When seen on fully ripe achenes from which the corolla has 

 fallen away these structures often appear considerably alike, yet 

 even in this late stage differences are usually to be noted. The 

 true pappus, even when very rudimentary, inclines to be cup-shaped 

 and possesses a thin edge. The annulus is a mere low usually thickish 

 cartilagineous or fleshy ring with a blunt edge. If examined in a 

 younger stage, while the corolla is still in place, it will be seen that the 

 true pappus, however rudimentary, is distinctly exterior to the base 

 of the corolla, while the annulus is merely the ringlike base from which 

 the mature corolla disarticulates. This base sometimes enlarges 

 slightly after the corolla falls and has in such species as Agcratum 

 cchioides or more properly Alomia echioides been taken for a true and 

 much reduced pappus. In the peculiar genus Jaliscoa the annulus 

 takes on a saucer shape and has frayed quasi ciliolate edge, yet when 

 carefully examined in a young state it can be seen to be a sub-corollar 

 rather than an extra-corollar structure. 



To add to the complexity an intra-corollar disk is often present in 

 ^'arious rudimentary forms, as for instance in minute fleshy or glandu- 

 lar papillae or in a fleshy ovoid, depressed-globose, or napiform en- 

 largement at the base of the style, capable of some persistence in the 

 mature achene. 



