NOTES ON PENTAS. 129 



Engler's Pjianzeimelt Ost-Afrikas cited by Mr. Elliot a few lines 

 above for P. longituha, where it stands — "P. lanceolata (Forsk.) 

 K. Sch. in Engl.'-Pr. iv. 4, S. 29." Mr. Elliot does not seem to 

 have looked up this reference— had he done so, he would have 

 found there a species omitted from his list, and would further have 

 discovered that P. lanceolata is identical with P. carnea Benth. 

 (which stands tenth on Mr. Elliot's list), Dr. Schumann having, 

 in accordance with the practice at Berlin, restored the oldest 

 specific name. 



Mr. Elliot's treatment of P. caenea next demands attention : 

 but I propose first to tabulate the synonymy of the species, which 

 is somewhat extensive, and in one instance confused. This plant 

 is the type of Brown's Neurocarpcpa as well as of Bentham's Pentas, 

 so that, whichever name be ultimately adopted, the earliest repre- 

 sentative of the genus is identical in each case. I have arranged 

 the synonyms in chronological order. 



Neurocaep^a lanceolata Br. in Salt, Abyss. Appendix iv. p. Ixiv 

 (1814). 

 Ophiorhiza lanceolata Forsk. Fl. ^gypt. Arab, p, 42 (1775). 

 Manettia lanceolata Vahl, Symbolse, p. 12 (1790). 

 ?Mussaenda sgyptiaca Lam. Encycl. iv. 394 (1795-G).* 

 Sipanea carnea Neumann in Kev. Hort. 2 S. ii. 445 (1843). 

 Pentas carnea Benth. ex Hook. Bot. Mag. t. 4086 (1844). 

 Orthostemma roseum Wall, ex Voigt, Hort. Calcutt. 384 (1845). 

 Vignaldia Quartiniana A. Rich. Tent. Fl. Abyss, i. 357 (1847). 

 Mussfenda luteola Hochst. ex A. Rich. 1. c, non Delile. 



Pentanisia sufi'ruticosa ] jr, , i • n v. i^r i • /r. i. \ 



13 ^, • • Klotzsch m Peters, Mossambique (Bot.) 



Pentanisia nervosa v . 9Q7_o /'-|qpo\ 



Pentanisia cymosa J ' \ )• 



Pentas Quartiniana Oliv. in Trans. Linn. Soc. xxix. 82 (1873). 



Pentas Klotzschii Vatke in Linnaea, xxv. 232 (1875). 



Pentas rosea Lanessan in Compte-rendu Assoc. Fran9aise, 



Sess. 5, p. 473 (1877). 

 Virecta carnea Baill. Hist. PI. vii. 331 (1879). 

 Pentas lanceolata K. Sch. in Engl. & Prantl, iv.pt. 4, p. 29 (1891). 



Mr. Elliot divides this species into five varieties, which he intro- 

 duces with a characteristically vague note : " The following varieties 

 might almost be considered specifically distinct ; indeed German 

 botanists seem from MSS. notes to uphold var. c and var. d as distinct 

 species." Seeing that Klotzsch published a description of var. d 

 under three names, and that Vatke (uniting them into one) also 

 printed his description, it is not easy to say why Mr. Elliit 

 should refer to "MSS. notes " in evidence of German opinion. Here, 

 as throughout his "revision," Mr. Elliot gives no synonymy. It 

 may therefore bo convenient to say that, so far as these varieties 

 go (I confess I have formed no high estimate of their value), the 

 synonyms in the above list divide among them as follows : — 



* Mr. Jackson cites this as of " Poir. in Lam. Encyc": but Poiret's work 

 begins with vol. v. of tliat work. 



Journal of Botany. — Vol. 35. [April, 1897.] ic 



