THE GENUS CTENOMERIA 169 



Hochstetter was not the only author who felt doubtful as to the 

 number of species included in Ctenomeria. The plant in which he 

 failed to find more than 40 stamens, which is not only the basis 

 of C. Kraussiana but also that of C. cor data, differs from Drege 

 8239 in being nearly glabrous instead of distinctly pubescent, and 

 in having crenate or subentire ovate-cordate leaves in place of 

 dentate and often shortly acutely 3-lobate ones. Hochstetter 

 has explained that, from its description, the plant which Thunberg 

 named Tragia capensis in 1794 (Prodr. PL Cap. 14) and in 1823 

 (Fl. Cap. ed. Schult. 37) must be a Ctenomeria not a Tragia, 

 though in the absence of specimens it was unsafe to say whether 

 the plant of Thunberg was the same as that of Drege or as that of 

 Krauss. We find then that of these two forms of Ctenomeria the 

 glabrous one received a new name C. Kraussiana Hochst., while 

 the name C. cordata Harv., which really belonged to the glabrous 

 form, was transferred to the pubescent one. 



Sonder only accounted for one species, which he described as 

 C. Kraussiana. His description was based on specimens distri- 

 buted by Ecklon & Zeyher as " 71," and by Ecklon as " 3845." 

 These specimens agree with each other and with " Drege 8239," 

 so that Sonder in effect reversed Hochstetter's arrangement ; 

 C. Kraussiana Sond. is precisely equivalent to C. cordata Hochst., 

 and as a result the C. cordata, of which Sonder suggests that 

 C. Kraussiana Sond. vix Hochst. may be only a variety, is really 

 G. cordata Harv. It would appear, however, that Sonder had no 

 opportunity of examining a specimen of this glabrous form, his 

 knowledge of which was derived from the description by Harvey ; 

 there is, however, a definite reference to that glabrous form 

 (Linnaea xxiii. 109) as C. capensis Harv., under Tragia rupestris, 

 /3 glabrata Sond. ::: The use of this name may be due to Sonder 

 having received from his friend Harvey an intimation that the 

 pubescent Ctenomeria was congeneric with, yet specifically distinct 

 from, the nearly glabrous G. cordata Harv., and that this pubescent 

 one is identical with Tragia capensis Thunb. Against this sugges- 

 tion, however, is the fact that Sonder, while identifying T. capensis 

 Eckl. & Zeyh. with the pubescent Ctenomeria, did not, any more 

 than Hochstetter, venture to identify either it or the glabrous one 

 with T. capensis Thunb., and the further fact that, in using the 

 name C. capensis Harv., Sonder did so in such a fashion as to 

 restrict it to the glabrous form which Harvey had actually 

 described as C. cordata. The only reasonable explanation of the 

 use of the name C. capensis Harv. by Sonder is that Sonder wrote 

 it by accident in place of C. cordata Harv. 



* This plant was treated by Miiller in 1866 (DC. Prodr. xv. 2, 940) partly 

 as T. rupestris, /3 glabrata Sond., partly as T. rupestris, y minor Miill.-Arg. 

 The latter is a valid species, T. minor Sond. (Linnaea xxiii. 108). So too is the 

 former, and this fact Miiller at one time himself appreciated, for in herb. 

 Stockholm he wrote the species up as T. affinis Miill.-Arg. This name has 

 since been published (Keic Bulletin, 1912, 334). Unfortunately in 1894 a 

 Mexican species, T. affinis Robinson & Greenm. (Proc. Am. Acad. xxix. 393) 

 was described under the same name ; it is therefore necessary to alter the 

 name T. affinis Miill.-Arg. non Robinson & Greenm. to T. Wahlbergiana. 



Journal op Botany.— Vol. 51. [May, 1913.] o 



