360 NATURAL HISTORY OF PLANTS. 



are to be seen in tlie conformation of the androceum and the recep- 

 tack', and in the number of parts in the gyna^ceum. The anthers 

 are sometimes oval or oldong in Horlxclia, as in H. congesta, cuneata, 

 sometimes ilid\nious, as in FT. tridentata, Gordoni, as well as in 

 //v'.v/V/ saiifoliiioidc.'^. The filaments are flattened, long, and triangular, 

 and nearly petaloid in JL congesta and cuneata, but filiform, as in the 

 true Potentils (a character said to be proper to Ivesia), in H. Gordoni 

 and tndt'ittata} The floral receptacle rises into a higher tube in H. 

 congcsfa than in most Potentils ; but in most of the other species it 

 is more spreading; in others, again, it is contracted below; in short, 

 its form is by no means constant in the species which resemble each 

 other most closely in all other respects. In nearly all the Horkelias 

 the carpels are as numerous as in the true Potentils, while in Ivesia 

 their number is usually as much reduced as in Sibbaldia, most of 

 the species having only four or five ; and some, such as H. Gordoni, 

 may only have two, or even only one. This latter number is constant 

 in tlie very remarkable plant named I. sanfoliiwides,- which has fifteen 

 stamens with didymous anthers, a hemispherical receptacle, and a hairy 

 disk at the insertion of the stamens, while the flowers are arranged 

 in cymes with slender axes, like those of a small Caryophyllad, and the 

 leaves are very peculiar, as we shall see below. We Lave made it the 

 type of a distinct section under the name of Stellariopms. Its style is 

 articulated at the base; it is at first terminal, and then becomes 

 inserted on the internal angle of the ovary below the summit. Its 

 insertion, and the presence of the articulation at its base are also 

 variable in the diflerent Horkelias and Ivesias above enume- 

 rated. All these plants' are herbs, possessing alternate stipulate 

 pinnate leaves, with lobed leaflets. All have cymose flowers borne on a 

 common scape. The foliage sometimes recalls that of Pofeniilla or 

 Geiitn, sometimes that of Spircsa or Sangnisorba. The leaves of , 

 7. HanloUnoides appear at first sight to differ widely from these forms, i 

 Thoy are little silky cylinders, which at a distance appear of a I 



/ 



J 



• Accordingly Asa Gray, who considered B. and Potentilla, a Californian species, witli sill^y 



Gordoni sis a doiiblful species of tliis genus, be- leaflets pressed together to form a cylindrical 



cause of the form of its hlumenls, had placed this blade, 

 plant in the genus Ivisia {loe. cit., 53U). 3 ^ dozen species have been as yet described, 



« It is probably to this plant that Uentham all natives of North America, e.-pecialiy the 



& HOOKKK allude in their Genoa wlien they Western regions — California, the .Rocky Moun- 



cite as intermedjate between Jlorkelia, Ivesia, tains, &c. 



