16 



lum huxlfolium of Elliott, (from the mountains of Grenville 

 district, South Carolina,) must be referred to L. serpylii- 

 folium, DC. 



13. Amelanchier Canadensis. — On our way, we cut down 

 a Service-tree, (as the Amelanchier Canadensis is here called,) 

 and feasted upon the ripe fruit, which throughout this region 

 is highly, and indeed justly prized, being sweet with a very 

 agreeable flavour ; while in the Northern States, so far as our 

 experience goes, this fruit, even if it may be said to be edible, 

 is not worth eating. As ' Sarvices ' are here greedily sought 

 after, and are generally procured by cutting down the trees, 

 the latter are becoming scarce in the vicinity of the ' planta- 

 tions,' as the mountain settlements are universally called. 



14. Astilbe decandra. — Hitherto we searched in vain for 

 the Astilbe decandra ; but we first met with this very inte- 

 resting plant in the rich and moist mountain woods between 

 Elk Creek and Cranberry Forge, and subsequently in similar 

 situations, particularly along the steep banks of streams, quite 

 to the base of the Itoa?i. Mr. Curtis found it abundantly 

 near the sources of the Linville River, and at the North Cove, 

 where it could not have escaped the notice of Michaux ; and 

 it is doubtless the Spircea Aruncus var. hermaphrodita of that 

 author. It indeed greatly resembles Spircea Aruncus, and at 

 a distance of a few yards is not easily distinguished from that 

 plant, but on a closer approach the resemblance is much less 

 striking. Michaux appears to have been the original dis- 

 coverer of this plant, and from him the specimens cultivated 

 in the Malmaison Garden, and described by Ventenat under 

 the name of Tiarella hiternata, were probably derived. It 

 was afterwards collected by Lyon, and described by Pursh 

 from a specimen cultivated in Mr. Lambert's garden at 

 Boynton. We noticed a peculiarity in this plant, which ex- 

 plains the discrepancy between Ventenat and Pursh (the 

 former having figured it with linear-spatulate petals, while the 

 latter found it apetalous,) and perhaps throws some additional 

 light upon the genus. The flowers are dioecio-polygainous, 

 the two forms difl'ering from each other in aspect much as the 

 staminate and pistillate plants of Spircea Aruncus. In one 

 form, the filaments are exserted to twice or thrice the length 

 of the calyx, and the spatulate-linear petals, inconspicuous 

 only on account of their narrowness, are nearly as long as the 

 stamens : the ovaries are well-formed and filled with ovules. 



