V. PLANTS WRIGHTIAN^. 41 



are similar, closely approximate, and papilliform. The flowers are as if intermedi- 

 ate between those of P. ovatifolia and P. ovalifolia. P. monticola, H. B. K, has 

 smaller flowers, narrower wings, &c.* 



KRAMERIACE^. 



t Krameria parvifolia (Benth. Bot. Voy. Sulph. p. 6. t. 1.) : fruticosa, pube ap- 

 pressa canescens ; caule erecto 1 - 2-pedali ramosissimo ; ramis divaricatis ; foliis 

 linearibus sessilibus junioribus mucronatis ; pedunculis sparsis ssepe bis bibractea- 



* Polygala flabellata, Shuttleworth, PI. Rtigel, exsicc. No. 37, from Key West, appears like a smoother 

 and very narrow-leaved, small-flowered variety of P. grandiflora (the var. /3. Torr. 4" Gray, Fl. 1. p. 

 671), the opposite extreme from his variety canescens. No. 39. 



P. leptostachys, Shuttleworth, from Aspalaga, Florida, is a good species, which I have from Dr. Torrey 

 (E. Florida, Leavenworth) under the name of P. tenuis, n. sp. ; and a less slender form is also sent by 

 Dr. Chapman, who distinguished it as an undescribed species. It is well distinguished from P. verticillata 

 and P. ambigua, by its glabrous, more slender and lageniform, somewhat curved seed. The calycine 

 wings are oval and subsessile, and the slender spikes are long-peduncled, as in P. ambigua, but the nar- 

 rowly linear leaves are nearly all verticillate. Mr. Shuttleworth's name, under which the plant has been 

 distributed, should be preferred. 



P. Rugelii, Shuttleworth, No. 26 of Rugel's Southern Florida collection, is only a form of P. lutea the 

 flowers of which have turned green, as in P. nana, probably in drying. Besides the characters given in 

 the Flora of North America, P. nana differs from P. lutea in having a prolonged cylindraceous rostellum 

 at the hilar end of the seed : in P. lutea the rostellum is very small and incurved. 



The corrections to the synonymy of Polygala cymosa and its allies in the Flora of North America, as 

 made in the Supplement, p. 670, are stated in a manner that is liable to mislead, and several of the names 

 are wrongly referred. It is true that to the P. cymosa of Walter must be referred the P. corymbosa of 

 Michaux, the P. graminifolia of Poiret, and the P. acutifolia of Torr. ^ Gray, Fl. 1. p. 128 : P. ra- 

 mosa of Elliott, however, belongs not to this species, but to the P. corymbosa, Torr. 4" Gray I. c, as 

 is manifest from Elliott's whole description, and especially from the hairy seeds which he attributes to 

 it ; and this name of Elliott's must be adopted for the species, since the P. attenuata of Nuttall surely 

 belongs to the other species, as also does the P. corymbosa of Elliott. The character and the synonymy 

 should stand as follows : — 



Polygala cymosa (Walt.) : caule simplici elato superne attenuate subnudo ; foliis radicalibus lanceo- 

 latis linearibusve acutis ssepius elongatis gramineis, caulinis sensim abbreviatis subulatis ; cyma corymbosa 

 simpliciuscula (e racemis simplicibus nunc parce ramosis) ; rhachi post lapsum florum squarrosa ; alls caly- 

 cinis ellipticis mucronatis ; seminibus subglobosis glaberrimis ecarunculatis. — P. cymosa, Walt. Car. p. 

 179 ; Torr. ^ Gray, Fl. 1. p. 128, Sr P- 670 (excl. syn. P. ramosa, Ell). P. corymbosa, Michx. Fl. 2. 

 p. 54 (in partem) ; Ell. SL 2. p. 187, non Nutt., nee Torr. ^ Gray, Fl. P. graminifolia, Poir. Diet. 5. 

 p. 500; DC. Prodr. 1. p. 329. P. attenuata, Nutt. Gen. 2. p. 90. P. acutifolia, Torr. ^ Gray, Fl. 1. p. 

 128 (P. cymosa /3. graminifolia, Torr. ^- Gray, I. c.p. 670) ; forma cyma composita. 



Polygala ramosa (Ell.) : caulibus e basi plurimis (8- 10-uncialibus) usque ad apicem foliosis ; foliis 

 obtusis, radicalibus spathulato-obovatis, caulinis oblongo-linearibus ; cyma decomposita fastigiata e racemis 

 brevibus confertis ; alls calycinis oblongis acuminato-cuspidatis ; seminibus ovoideis hispidulis carunculo du- 

 ple longioribus. (Flores et semina iis P. cymosae subdimidio minores.) — P. ramosa, Ell. Sk. 2. p. 186. 

 P. cymosa, Poir. Diet. 5. p. 500, non Walt. P. corymbosa, Nutt. Gen. 2. p. 89 ; Torr. 4* Gray, Fl. 

 l.p. 128, non Michx., nee Ell. 



Walter's name must be kept for the first of these species, to which it certainly belongs, as the specimen 

 in his herbarium, as well as the character, shows. Michaux's P. corymbosa consists principally of P. cy- 

 mosa, Walt., with which in his herbarium some P. Baldwinii, I believe, is intermixed. The two species 

 are distinguished, and for the first time well characterized, by Elliott, whose name for the last should be 

 retained. I am much indebted to Mr. Carey for the elucidation of these species. 



