Gentiana,] GENTIANE^, 



55 



* Corolla clavala apice cojimvente. 



1. G, Saponaria {L,); caule ascendente, foUis ovato-lanceolatis obovatisque margine sca- 

 bris, floribus aggregatis subsessilibus bracteatis, caiycis integri lobis tubum subasquantlbus, 

 corollae cyaneae lobis ovatis obtiisis plica (fissa) duplo longioribus, antheris connatis, semi- 

 nibus anguste alatis. Z. Sp. PL 1. p. 330, {excbuo citat Plukn.) Lam. EncycL 2. p. 63T, 

 Ait Hort Kew. 2. p. 111. Sims BoL Mag. t 1039 (excl. syn. Froel et Mich.) Pursh, Ft. 

 1. p. 185, (excl. cit. Andr,)~~G. Catesbaei. WalLp. 109. Nutt \. p. 172.— G* fimbriata. 

 Fa/fly Symb, 3. p. 46. — /3. linearis; foliis lineari-lanceolatis, aiUheris demum Uberis. — G. 

 Catesbffii. EIL I. 339.— G, linearis. FroeL Mon. p. 37. Pursh, 1. />. 186.~G. Pneumon- 

 anthe- Mich. 1. p. 17b'. — G. pscudopneumonanthc. IL 5. et p. 146.— G. pubcriila. 

 MicL? \.p. 176. 



Hae. — /3. Has been collected in Caiiaila, {Mrs. Percivah) in the nci-j^liboiirliood of Quebec, and at Lake 

 St. Charles. Mr.Sheppard. I have seen specimens of the broad-leaved var. from as far northward as Mary- 

 land ; but it undoubtedly occurs with its variety /3. at the above-mentioned places, as they by no means 

 specifically differ from each other. I am certain that this is the true G, Saponaria of Linneus ; for his 

 diagnosis, and the drawings which he quotes, viz. those of Catesby (Carol. I. t. 70) and of Morison (vol. 

 3. sect. 12. t. 5) leave no doubt of it. The quotatiou of Plukeuet is, however, erroneous. The drawing of 

 Catesby also proves that G. Catesbcei (Walt.) is identical with our species ; and it is natural that Froelicii 

 found fault with that drawings, from the absence of the plicje, which may be easily overlooked in this species, 

 but not in the following. G. Saponaria is much rarer than the following, which has been often taken 

 for it as well by American as European botanists. The variety (i. may be still commoner than a. 

 which has been quite overlooked by some of the American botanists ; as, for instance, Michaux, Elliot, and 

 lately, Beck. This species is always of lower growth, but scarcely smaUcr-flowcrcd, than the following ; 

 the leaves are somewhat coriaceous ; the flowers almost all terminal and crowded : nevertheless, these 

 characters, though valuable in general, are not constant in all cases ; but I have not found that the varia- 

 tions which take place in the proportions of the lobi ^ud plica corollce, ever exceed certain bounds. There 

 are, besides, most valuable characteristic differences afforded bv the seeds of these species: G. Saponaria L. 

 has oblong-winged seeds, the wing of which is narrow un both sides, and long and attenuated towards theends; 

 the seeds of G. Andrewsii are orbicular, and very broadly and equally winged ; while those of G. ochrohuca 

 have no wing at all. The connexion of the anthers, and the length of the /o^/ca/^c?5, are of much less conse- 

 quence. By the investigation of Sir W. Hooker, (in the Companion to the Bot. Mag. p. 171.) it has been 

 proved that G. Pneumonantke of the American authors is merely a variety of this species ; and however 

 distant they may seem to stand when studied only in their common forms, nevertheless the variety 

 (i. of G. Saponaria is so similar in some cases to G. Pneumonanthe, that it is very difficult to distinguish 

 them. For in those forms in which the plica is quite entire, (as, for instance, in that of Dr. Boott, in the 

 Hookerian herbarium, quoted by himself,) it is impossible to find out any decided difference in the struc- 

 ture of the corolla. I examined authentic specimens from Silesia, the lobi corollm of which were blunt and 

 somewhat converging. On the other hand, G. Saponaria has sometimes an open corolla, as G. Pneumun- 

 antJie commonly has. But there is a minute character, which seems to be most constant, and cer- 

 tainly has an important connexion with the whole of the species — viz. a distinct roughness on the njargins 

 of the leaves, so that they appear serrated under a magnifying glas«, while G. /'rtCMwoHaw^/ie is always en- 

 tirely smooth : I think that that roujrhness is the remains of those scabrous forms, which exist almost in 

 alt American species of GentiaUte. 



2. G. Andreicsii (Giiseb.); caule ascendciitc, foliis ovato-lanceolalis acuniinatis margine 



