557 



combined ones. The united anthers burst internally, by 

 which means their pollen is communicated to the stigma ; 

 but the anthers of Kuhnia open at the extremity, and re- 

 semble the corolla of an Aristolochia. All the florets are 

 superior, but this holds good likewise in Scabiosa. Hence 

 it appears that no essential character of compound flowers 

 is to be detected, though no order can be more natural 

 than that before us." 



" Tournefort first divided the compound flowers into 

 three sections, according to the shape of their partial co- 

 rollas. These are either ligulate or tubular. Such as 

 consist of ligulate florets only, are called by this writer 

 semiflosculosi; such as are formed only of tubular ones, 

 Jlosculosi ; while those which have ligulate florets in the 

 radius, and tubular ones in the disk, are denominated ra- 

 diati. This division seems natural enough, and yet is 

 not so. For it refers both the discoid and capitate com- 

 pound flowers of Linnseus to the Jlosculosi, which never- 

 theless are too dissimilar to be possibly admitted into the 

 same section. The discoidei of Linnaeus, Ray's aggregati, 

 having aggregate florets, seated on a hemispherical recep- 

 tacle, are, in fact, more allied to the radiati; while the 

 capitati, such as Thistles, are widely different, so as ne- 

 cessarily to constitute a division by themselves." 



" Vaillant attempted a new botanical system ; but it 

 is to be lamented that we are possessed of no more of his 

 labours, than what concerns the compound flowers. In 

 this performance, published in the Memoirs of the Pari- 

 sian Academy for the years 1718, 1719 and 1720; he 

 has displayed an extensive knowledge of species, and has 

 treated the subject admirably. As the Memoirs of the 

 Academy are not within the reach of every body's purse, a 

 German named Von Steinwehr has collected the anato- 

 mical, chemical and botanical papers, into an octavo vo- 

 lume, published in 1754 at Breslaw. In this Vaillant's 

 treatises are preserved entire," (but in the German lan- 

 guage.) 



