240 Harper: Some plants from Georgia 



Typical A. arborescens is not definitely known south of the 

 mountains of Georgia,* and the coastal plain form here discussed 

 seems to be one of a small class of plants inhabiting the " meso- 

 phytic " forests of the Eocene region of Southwest Georgia and 

 nearly but not quite identical with Alleghanian species. Other 

 examples of this class are Magnolia pyramidata Pursh, Thaspium 

 barbinode Chapmani C. & R., and Rhododendron Cuthbertii Small, 

 all of which I have found in Randolph County, 



Bartonia lanceolata Small, Fl. S. E. U. S. 932. 1903 



This is probably identical with B. tenella bracluata Wood 

 (Class-Book 586. 1861). If so this will not be the first time 

 that a supposed new species proposed by one of the present gener- 

 ation of systematists has been found to have been already de- 

 scribed as a variety by Wood, who seems to have had an excep- 

 tional faculty of recognizing previously unobserved characters in 

 plants, though he was too conservative to describe many new 

 species. The Vienna rules do not require any change of name in 

 a case of this kind, however so I will not undertake to increase the 

 number of synonyms at present. 



In studying our species of Bartonia, the papers by Robinson 

 (Bot. Gaz. 26: 46-48. 1898) and Williams (Rhodora 2 : 55—57- 

 pl. 15. f. i—y. 1900) will be found useful. 



Amsonia ciliata Walt. 



Two forms of this occur quite frequently in South Georgia, 

 one with leaves two or three times as broad as the other. There 

 are other slight differences, but scarcely enough to entitle both 

 forms to specific rank, though they are distinguished easily 

 enough and should receive some recognition in systematic works. 

 As a rule the broader-leaved form grows in dry pine-barrens and 

 the other on sand-hills. Correspondence a few years ago with 

 Mr. E. G. Baker, of the British Museum, elicited the fact that the 



*It is mentioned in Earle's Flora of the Metamorphic Region of Alabama (Bull. 

 Ala. Agric. Exp. Sta. 119 : 96. 1902), but in the absence of specimens it is impossi- 

 ble to ascertain whether the mountain or the coastal plain form is meant. Wood may 

 have known our plant, for his description of A. arborescens (Class-Book 490. 1861) 

 fits it very well, and one of the localities cited, Macon, Georgia, is on the edge of the 



