Cockayne. — New Species of New Zealand Flowering-plants. 115 



C. angustifolia are found in the mountains near Collingwood, the habitat of 

 Buchanan's species, the name would be C. novae-zealandiae (Buch.), and 

 C. an<i>istifolia be a synonym. 



B. L. Robinson, of Harvard University, in Proc. Am. Acad., xl, 1913, 

 has changed the name Cehnisia to Elcismia and applied this latter to all 

 the species considered valid in Cheeseman's Manual. The reason for this 

 change, according to Robinson, is that Cehnisia Cass, in Diet. Sc. Nat., vii, 

 356 (1817), is distinct from Cehnisia Cass, in Diet. Sc. Nat., xxxvii, 259 

 (1825), and that the New Zealand species fall into the latter, while into the 

 former go Cehnisia tabularis, syn. Arnica tabularis Thunb. and Cel. tomen- 

 tosa, syn. Conyza tomentosa Burm. f. According to the " Index Kewensis," 

 fasc. i, p. 476, Cehnisia Cass. (1817) = Alciope D.C. (1836), which includes 

 the two species referred by Robinson to Cehnisia, while Cehnisia Cass. (1825) 

 is the genus as known in Australasia. The above facts seem to strongly 

 support Robinson's change of name. All the same, it would be a matter 

 for deep regret if a name so universally recognized as Cehnisia had to be 

 abandoned. It seems to me that New Zealand botanists would do well 

 to wait until the next Botanical Congress, so as to see if Cehnisia cannot be 

 placed in the list of nomina conservanda. 



Celmisia Hookeri Cockayne nom. nov. 



This is C. verbascijolia as defined by Cheeseman in the " Manual of the 

 New Zealand Flora," p. 309, that author being in grave doubt as to whether 

 the material he was dealing with really represented C. verbascijolia Hook. f. 

 Nothing needs adding to Cheeseman's admirable description. 



. C. Hookeri, so far as is known, is confined to north-east Otago, where the 

 climate is comparatively dry. On the other hand, C. verbascijolia Hook. f. 

 was based on specimens collected by Lyall near Milford Sound and Preserva- 

 tion Inlet, where the rainfall is excessive. As defined by its author, it is 

 evidently a quite different plant to C. Hookeri, being smaller in all its parts 

 and with heads only 5 cm. diameter, as opposed to those of C. Hookeri, which 

 are more than 8 cm. diameter. 



I think there is little doubt that C. Brownii Chapm., a widespread plant 

 in the south-west of Otago, is identical with C. verbascijolia Hook. f. Kirk 

 was the first *to identify the north-east Otago plant as C. verbascijolia, and 

 it is curious that in his " Students' Flora " the original habitat is not cited. 



X Celmisia Christensenii Cockayne nov. typ. hybrid. (C. spectabilis 

 Hook. f. x C. Traversii Hook, f.) 



Folia vaginam includens + 16 cm. longa, 3 cm. lata, flexibilia ; lamina 

 oblonga, basi subcuneata vel in petiolum brevem sensim attenuata, + 10 cm. 

 longa, supra viridis irregulariter sulcata glabra nisi nervi medii baso, subtus 

 fcomentis pallide fuscis mollibus aliquanto laxis circ. 1 mm. altis vestita. 

 nervo medio breve carinato purpureo sed pilis occulto, margine integerrima 

 pilosa pilis sordide albis sed basin versus fuscis ; petiolus brevis, carnosus, 

 circ. 10 mm. latus, tenue canaliculars, basi in vaginam purpurascentem 

 laxe sericeo-tomentosarn 4-5 cm. longam 18 mm. latam dilatatus. Scapus 

 duplo folia excedens, rigidus, crassus, pilis sericeis albis adpressis vestitus ; 

 bractae circ. 6, lineares, spathulatae vel vix subulatae, supra virides pilosae 

 pilis sericeis albis sed subtus apiceque pallide fusco-tomentosae. Capitula 

 circ. 4 cm. diam. ; involucri bractae lineares, pilis albis pilosae sed apice 

 brunneo-scariosae vel pilis pallido-fuscis vestitae ; ligulae 4-nervae, albae, 

 c re. 2-4 cm. longae. 



