233 
Linnzeus, relying perhaps too much on the habit of some 
small specimens, which resemble Exacum filiforme, and still 
more Andrewsia paniculata, referred it to his Rotate ; Genti- 
«nee of Juss. Jussieu himself included it in the Caryophyl- 
lee in his genera Plantarum: but in his Mémoire sur les 
Hypericées, in the 20th vol. of the Mém. du Mus. he says that 
Sarothra may perhaps be referred to this order; and he 
alludes to the relation which Richard and Michaux have 
considered it to bear with Hypericum, from which, continues 
M. de Jussieu, * it nevertheless differs by its stamens in a 
definite number (5-10), and by the attachment of the seeds 
to the margins of the valves, as in the Gentianee, to which, 
if it were monopetalous, it would be farther allied by its habit 
and by the perisperm which Geertner has attributed to it.” 
De Candolle has not included it in any of the orders of the 
. Thalamiflore : from which it may be inferred that that emi- 
nent Botanist intends to place it among the Gentianez. Nut- 
tall, in his * Genera of North American Plants," where he has 
paid much attention to their natural affinities, observes, that 
this plant is closely allied to Hypericum, and scarcely in any 
respect like a Gentian, its affinity to that genus being entirely 
fanciful.” In the 2d vol. of the Mém. du Mus. M. St. Hilaire 
is disposed to transfer Sarothra to Violariee, on account of 
the definite stamens, the nature of the perisperm, the position 
of its embryo, and its seeds fixed to the inner wall of the 
capsule: notwithstanding that in Viola the seeds are at- 
tached to the middle of the valves, whilst here the inser- 
tion is at the margin. Our author in a subsequent 
treatise in the 3d vol. of the Mém. du Mus.,—taking this latter 
character into account,—unites Sarothra with Frankenia and 
Sauvagesia, in order to constitute the groupe of Frankeniacee ; 
and in a note he remarks, ** En traitant de ce groupe, J'avais 
éloigné toute idée de faire du Sarothra un Hypericum ; cepen- 
dant il existe entre ces genres un trait de ressemblance qui 
aura decidé le Botaniste célébre dont la plume savante a, dit- 
on, été empruntée souvent par l'auteur du Flora Americe 
Borealis. Dans une suite d’Hypericum indigénes de la 
France, J’ai observé toutes les dégradations possibles entre 
