laS TETRAXDRIA. DIGTNIA. 



giUa, is clothed with a very short stellate pubescence. 

 The fruit of the Fotltergilla, as was long ago sat^aciously 

 rerrarked by the ctlcbiated \. L Jussicu, is almost ex- 

 actly similar to that of the U a mame lis, hui in the flowers 

 there ex sis a sTang-e disparity; by the intervention of the 

 J'ac/wsayidra, hov.'evei, wliich appears more properly re- 

 ferable to tlie ^hneiitaceiC, tlian the Ei/phorhiaceie, and at 

 the same time allied bodi to HamameUs and Fothergilla^ 

 We have S(*methin<^ like a connected seiies. In these 3 

 genera, the antheis and filaments are united, the anthers 

 also become papyraceous, and are for a considerable 

 Icneih of time persistent, not deciduous as in the Bei-ber- 

 rides. From the Amentaceae, ih'^y differ in possessing a 

 perisperm, and the HamameUs is excluded from the 15th 

 class ot Jussieuby the existence of petals, but this excep- 

 tion is scarcely universal, as tiie order in quest i'>n in- 

 cludes Coniptonia^ described as having 6 petals, at all 

 even;s, an interior and dissimilar calix. Wherever these 

 3 coniiguors genera may be placed, either as a distinct 

 order, (Fo x hergillea) or as a neig' bouring section, 

 th' y cannot but be con.s.dered as ch sely allied to the 

 Amaitucea, notwithstandin.g t:he Singular construction of 

 the anthers in the genus HamameUs, wnich is indeed the 

 Sole character by which it is in any manner allied to th.e 

 Jierberides, tie carreous perisperm being equally com- 

 mon to the E.iphorbiacece. There is son)e reason to sus- 

 pect that tleg'en> s HuiruimeUs'is polygamous, independent 

 of the mistake of L.nnseus m confounding this genus wi'li 

 FotheygiUa. The variety, for instance, which I have ele- 

 signated, y. *pa)'vifoUa, although fiowenn,:,- freely every 

 year, lias never produced se^ii since transplanted into a 

 garden out ot contact with o her individuals Tiiis sub- 

 ject, however, requires more general and accun^te exami- 

 nation. The thr' e genera proposed are thus far exclu- 

 sively confined to North AniCiica. 



161. SANGLISORBA. L. (Great Bwrnrt.) 



Calix 2-]paved, inferior, resembliiij^ brartes. 

 Corolla 4-r!tft, superior. Capsule, between the 

 caiix and corolla, 2 seeded. 



Herbaceous; leaves uneqtially pinnate, stipules growing 

 to tliebast of the petiole; flowers capit^ite terminal. 



SPEcifs. 1. .S'. cavadtusis. 2. media. Probabh a mere 

 var e ly ot No. 1 Of this gun s the:e are but 2 other spe» 

 cies, 1 in Europe and the other in Algiers (Africa.) 



