192 Transactions. 



Hah. — Table Hill, Stewart Island ; W. J. Murdoch ! High lands at- 

 Port Pegasus ; T. Kirk ! High lands at head of Paterson Inlet ; H. 

 Guthrie-Smith ! 



Mr. Gruthrie-Smith sent me a live plant of this species some years ago. 

 It has grown well, and has supplied some of the material used in preparing 

 this description. Mr. Kirk's specimens were sent me under the MS. name 

 of G. australis. This name he afterwards used to designate a small state of 

 C. litorosa Bailey, from the coast of Stewart Island. The late Mr. C. B. 

 Clarke referred Mr. Kirk's Port Pegasus plant to C. uncifolia Cheesm., but 

 I am unable to acquiesce in this reference. 



9. Koeleria Cheesemanii (Hackel) Petrie comb. nov. 



A full description of this grass, under the name of Trisetum Cheesemanii 

 Hackel, appears in Trans. N.Z. Inst., vol. 35 (1902), p. 381. In his " Manual 

 of the New Zealand Flora " (p. 882) Mr. Cheeseman does not allude to the 

 obvious likeness of this Trisetum to Koeleria, which Hackel has noted 

 {}oc. cit.). In an elaborate note to his specific character the latter pointed 

 out that " his species is a very distinct once, like indeed in habit and in the 

 character (indole) of the panicle and spikelets to Trisetum suhspicatum, 

 Beauv., yet by its very shortly bidentate flowering-glumes,^ that show a 

 very short mucro or awn from between the teeth or from a little below them, 

 it is so divergent not only from Trisetum suhspicatum, but even from all 

 the genuine Trisetums, that it may preferably be annexed to Koeleria. In 

 truth," he adds, " between this genus and Trisetum no certain limits are 

 found, and it was for this reason that the celebrated Desvaux was led to 

 reduce Koeleria to a section of Trisetum." The passage above cited is given 

 in Latin, which I have translated. 



It is hard to understand how Hackel could have penned this note, 

 seeing that in his well-known work on " The True Grasses " the genera 

 Trisetum and Koeleria are maintained, and are not even placed in imme- 

 diate juxtaposition. 



My opinion that T. Cheesemanii is a true Koeleria is not founded on the 

 statement of Hackel quoted above, but has arisen from the study of the 

 specimens of this grass collected in a number of widely scattered localities 

 in the South Island. In Trisetum the awns on the flowering (fertile) glumes 

 are never terminal, nor are they wanting. In T. Cheesemanii things are 

 very different. On the same panicle one finds spikelets with short sub- 

 dorsal awnlets (mucros) on the flowering-glumes which are shortly bidentate 

 at the tip, or the awnlets may be terminal between the short teeth, or the 

 glumes may have acuminate or acute tips with no trace of awnlets. Indeed, 

 two of these conditions may often be found on a single spikelet. In view 

 of these facts, there can, it seems to me, be no question that Hackel's species 

 must be transferred to Koeleria. The only alternative is the reduction of 

 Koeleria to a section of Trisetum, a course which no modern authority on 

 the classification of grasses has followed. 



