Pbtrie. — New Salnie. Phanerogams. 57 



produced into a short thin more or less scabrid mucro, as long as the 

 utricles. 



Utricles 4 mm. long, 1 mm. broad, narrow elliptic, shortly stipitate, 

 gradually narrowed above into a rather long slender slightly scabrid beak 

 ending in 2 long slender scarcely diverging scabrid teeth, thin, flatly bi- 

 convex with two prominent lateral nerves and numerous other delicate 

 nerves along either side. 



Nut elliptic, equally narrowed above and below, sharply trigonous, 

 delicately puuctulate. 



Style branches 3 (rarely 2), short. 



Hob. — Denham Bay, Sunday Island, Kermadecs. 



Mr. W. E. B. Olive]' discovered this plant, which he has kindly placed 

 in my hands for description. The. material is imperfect, only the culms 

 being represented. Freshly gathered specimens are needed to allow of a 

 more accurate description of the singular inflorescence, but I hope that 

 my account will not prove very wide of the mark. The lowermost simple 

 spikelet may not always be present, while the compound spikelets, that 

 are usually closely placed, are widely distant in one of the specimens 

 examined. The plant was formerly referred to C. Forsteri Wahl. as sub- 

 species insularis Oliver, but it does not appear to me closely related to 

 that species, or, indeed, to any other of the native forms. 



15. Calamagrostis (Deyeuxia) Youngii (Hook, f.) Cheesm. var. Petriei 



comb. nov. 



Planta a forma typica differt arista a gluma florigera media oriente. 



The present variety is the Deyeuxia Petriei of Cheeseman's " Manual 

 of the New Zealand Flora" (= Calamagrostis Petriei Hackel). Its only 

 important difference from the typical form is in the position of the awn. 

 which springs from the middle of the back of the flowering-glume. Neither 

 Hackel nor Cheeseman had seen Hooker's plant when their descriptions 

 of Calamagrostis (Deyeuxia) Petriei were made out. The figures of 

 Deyeuxia Youngii in Buchanan's " Indigenous Grasses of New Zealand " 

 probably represent a form of Deyeuxia quadriseta B. Br., though the en- 

 larged drawings of the spikelet may represent the real plant. Buchanan 

 considered it a common grass in the hilly parts of southern Otago. If 

 this was formerly the case, the plant must have been largely eaten out by 

 stock, for it is now, I believe, quite rare in these districts. 



16. Poa Colensoi Hook. f. var. brevi-ligulata var. nov. 



Differt a typo ligula valde brevi v. obsoleta crassiore haud vaginante, 

 vaginis angustioribus. 



This variety differs from the typical form in possessing very short or 

 obsolete more coriaceous non-sheathing ligules and narrower sheaths. 

 The leaves are usually more erect and rigid and often more or less pungent- 

 pointed, while the plants frequently form firmly compacted sward-like 

 patches of considerable size. 



In all the extant descriptions of Poa Colensoi the ligule is said to be very 

 long and sheathing ; as Mr. Cheeseman puts it (" Manual of New Zealand 

 Flora," p. 908), ' : Ligules very large and long, sheathing, membranous, 

 hyaline." This condition is, however, very far from constant. My col- 



