ROXBURGH'S HORTUS BENGALENSIS. 



417 



the issues raised being of considerable importance. In several 

 cases, new specific combinations seem to be necessary, but as 

 it is not possible here to form definite conclusions on the actual 

 merits of the problems, except in the case of a few species, 

 none are herein proposed. 



Gratiola integrifolia was placed by Hooker 10 under Vandellia 

 erecta Benth., but he appears to have been in doubt as to the 

 identity of Roxburgh's species. Rheede's figure and description 

 both show 2-stamened flowers, whereas in Vandellia, the number 

 of stamens is always 4. 



Andropogon bicolor, Andropogon saccharatus, and Anthistiria 

 cymbaria are treated by Hackel 11 as different from Holcus 

 bicolor Linn., Holcus saccharatus Linn., and Andropogon cym- 

 barius Linn., respectively, but they must be typified by those 

 species. 



Panicum spicatum Roxb. is given by Hooker 12 as a synonym 

 of Pennisetum typhoideum Rich., but the oldest name cited by 

 him is Panicum americanum Linn., still valid for transfer to 

 Pennisetum. 



Ixora fulgens, I. stricta, and /. bmbata are all maintained by 

 Hooker, 13 but as the plates on which they are based are not 

 cited by him, no opinion can be formed as to whether they rep- 

 resent the original types. 



Oldenlandia alata Koenig is also retained by Hooker 14 as a 

 good species, but O. alata Roxb. is based on Hedyotis racemosa 

 Willd., which is Lamarck's species of that name, and this is said 

 by Hooker generally to be cited under 0. paniculata Linn., but 

 probably to belong to a different section of the genus. 



Oldenlandia crystallina Roxb. is also maintained by Hooker, 15 

 with Hedyotis pumila Linn, f . as a synonym, and H. pumila Willd. 

 must be held typified by the Linnaean species. There is no 

 apparent reason for calling the species other than Oldenlandia 

 pumila (Linn, f.) DC. 



Oldenlandia herbacea Roxb. is based on Hedyotis herbacea 

 Willd., which is typified by H. herbacea Linn., although the two 

 are treated as doubtfully identical. 16 If, then, Hedyotis herbacea 



10 Fl. Brit. Ind. 4 (1884) 281. 



"DC. Monog. Phan. 6 (1889) 519, 510, 629, 668. 



13 Fl. Brit. Ind. 7 (1896) 82. 

 "Fl. Brit. Ind. 3 (1880) 145-148. 



14 Fl. Brit. Ind. 3 (1880) 70. 

 18 Fl. Brit. Ind. 3 (1880) 65. 

 16 Fl. Brit. Ind. 3 (1880) 65. 



