174 OBSERVATIONS ON THE GENUS PHYLLACHNE. 



question was iDrovisionally referred by Sir Joseph Hooker to 

 Stylidium (as S. siihiilatim , 'Handbook,' i., 168), that celebrated 

 investigator of New Zealand plants having only seen specimens 

 very imperfectly in flower. I had long since exx)ressed to some 

 con-espondents in New Zealand my belief, that this rare and re- 

 markable plant would be found referable to Forstera {i.e., Phijllachne), 

 the shortness and straightness of the column was indicative of 

 of that genus, and this anticipation can now be verified from 

 perfect flowering specimens, which, on my solicitation, were very 

 obligingly procured by Professor W. Petrie, M.A., of Otago. 

 The diagnosis is subjoined: — 



Phyllachne (Forstera J subulata. — Acaulis foliis elongatis subu- 

 lato-hnearibus pedunculum longe superantibus confertis, 

 calyce ebracteato glandulari puberulo 2-3-lobo, lobis 2 ro- 

 tundatis vel duobus oblongatis et uno rotundato, coroUs 

 lobis 4-5 tubo glandulari-puberulo duplo longioribus, co- 

 lumnam conspicue excedentibus, fauce inappendiculata, 

 capsula ovata ad apicem tardissime dehiscente. 



From the sea-level to the height of 2090 feet at Dunedin, 

 Buff Hill and Invercargill, in tufts of Oreobolus ; Petrie. 



Longest leaves measiu'ing nearly 1-^ inch. Peduncles 

 half an inch long or shorter, almost glabrous. Lobes of 

 the calyx about 1 line long (reminding of those of several 

 Stylidiums), when round occasionally bilobulate. Lobes of 

 the corolla oblong, not exceeding 2 lines. Column very 

 similar to that of P. clavigera, immovable. Stigmas smooth. 

 Fruit about one-thkd of an inch long. Seeds measuring 

 in length half a line. 

 Speaking thus of a member of the Stylidece, I may be allowed 

 incidentally to remark that I see no reason for changing my views 

 in reference to the true systematic position of Donatia, as expressed 

 in the 'Fragm.' viii. 41, although its ripe fruit remains still un- 

 known. I gave on that occasion several instances of Orders which 

 contain genera with free and connate petals ; I may here add that 

 Donatia holds in this respect the same relation to other StylidecB 

 that Dialypetalum does to Lobelia, Lysinema to Epacris, and even 

 some species of Marianthiis to others. 



I adhere also still to the view then enunciated, that Candollea 

 could not possibly be kept generically apart from Hibbertia ; and 

 this opinion was strengthened by my recent journey in West 

 Australia. Furthermore, there seems no reason why the great 

 name of DeCandolle should not be given again to the large and 

 lovely Stylidium of Swartz — science could not offer a more glorious 

 tribute; while Loureho's Stylidium should supersede i^/arZ^a ; and 

 the dedication by Gaertner, in Forstera (' De Fructibus,' i. t. 28), 

 be renewed, if the plant typical for this genus (Athecia) among 

 CalyciflorcB, which has undeservedly sunk into obhvion,- could be 

 traced anew in Forster's or Gaertner's collections. 



Melbourne, January 21, 1878. 

 • It is merely mentioned by Roemer and Schultes (' Syst. Veg.' v. p. xxi. <fe 269). 



