Petrie. — Descriptions of New Native Flowering-plants. 19 



it was founded on a single specimen, and refers it to U. compacta R. Br. 

 Mr. C. B. Clarke ranks it as a variety of the latter. The U. nervosa 

 in Cheeseman's Manual seems to me in no way closely related to the 

 Kelly's Hill plant. The figure of Boott's species in Flora Tasmanica is a 

 poor one, and affords little help' in identifying it. 



Uncinia caespitosa Col. var. collina var. nov. 



Rhizoma late stolonifera culmos plurifoliosos plerumque diffusos edens. 

 Culmi foliaque quam in typo breviores. Folia culmis breviora vel eos 

 aequantia, 10-20 cm. alta 4-6 mm. lata, plana suberecta coriacea, in 

 apices incrassatos longe productos abeuntia. Spiculae 4—8 cm. longae 

 + 8 mm.' latae, subclavatae, supra densiflorae, infra floribus laxioribus ; 

 glumae femineae utriculis nonnihil longiores ; nux elliptico-oblonga. 



Hab. — Mount Hikurangi (East Coast), at 4,800 ft. ; Ruahine Mountains, 

 at 3,500 ft. : B. C. Aston ! Tararua Mountains (Mount Holdsworth), 3,200 ft. 



This is a very distinct-looking form, and has considerable claim to 

 specific rank. It forms • large open somewhat sward-like patches, and 

 never grow^s in distinct tufts, as the typical form usually does. 



Carex secta Bootb var. tenuiculmis var. nov. 



Var. C. virgatae Sol. habitu subsimilis ; a planta t5^pica differt culmis 

 gracillimis baud raro filiformibus, foliis tenuibus perangustis flaccidis 

 complanatis vel apicem versus concavis (costa media parum conspicua) 

 40-60 cm. longis culmos aequantibus vel excedentibus ; inflorescentia 

 6-15 cm. longa simplici + pendula spiculis plerumque remotis parvis pauci; 

 floris sessilibus praedita vel a parte inferiore breviter ramosa (ramis paucis 

 brevibus indivisis) ; rhachide pergracili vel filiform! ; utriculis breviter 

 stipitatis vel paene astipitatis. • 



Hab. — Damp localities in eastern and southern Otago : D. P. Damp 

 localities in the Hanmer and Castle Hill districts. North Canterbury. : Arnold 

 Wall ! Damp localities in Chatham Islands : L. Cockayne ! W. R. B. 

 Oliver ! 



I am indebted to Professor Wall for drawing my attention to this 

 interesting plant, which I had unwittingly placed in the same species-wrajDper 

 as C. virgata. From this it differs entirely in the structure of the utricles. 

 A*hand-and-eye examination will hardly disclose its true position. The 

 Chatham Island specimens show only immature utricles, but I think their 

 identity with the mainland plant cannot be doubted. They also show 

 remarkably long leaves that greatly exceed the culms. 



Poa novae-zelandiae Hackel var. Wallii var. nov. 



A forma typica differt foliis coriaceis ± complanatis et a marginibus 

 mvolutis incrassatisque muticis ; panicula breviore folia baud vel vix 

 excedente, a parte inferiore foliis ± abscondita ; spiculis majoribus hand 

 compressis ; glumis florigeris latioribus trinerviis acutis hand incurvatis 

 subcoriaceis ; palea breviore ac latiore. 



Hab. — Mount Miromiro (Amuri County) : A. Wall ! Top of Mount 

 Kyeburn (Maniototo County) : H. J. Matthews ! Mount Pisa (Vincent 

 County). 



The present plant appears to be confined to wet shingly stations, and 

 ranges in altitude from about 3,500 ft. to 5,000 ft. or 6,000 ft. At the 

 higher altitudes the plants are very dwarf. 



