Page:The Botany of the Antarctic Voyage.djvu/115

Campbell's Islands] American plant again, which is the Rolens redolem, Vahl, (not of Forster), is considered by A. Brongniart as a variety (redolens) of the Hierochloe antarctiea, Br. (Voy. de la Coquille, Bot. Phan. p. 144. t. 23.), and he quotes Raspail (Ann. Sc. Obs. vol. ii. p. 83.) in confirmation of his views ; it certainly differs but slightly from the plant of the Old World, and chiefly in the longer cilise at the back of the lower florets, in having the glumes three-nerved, the panicle more contracted, and rather larger glumes. The leaves are generally quite smooth and glossy, the under surface only obscurely scabrid.

Besides the above mentioned species there are four more very distinct ones in the Southern Hemisphere, (as well as three natives of the Cape of Good Hope); one, the H. ifr««o«i,s, mihi (vid. infra), from Lord Auckland's group ; another, the H. utriculata, Kunth, a native of Chili ; and two hitherto undescribed species. The first of these is from Australia, H. rariflora, mini, and was discovered by Mr. Baxter in King George's Sound ; it may readily be distinguished by its slender culms, branching upwards, and its narrow scabrid leaves, loose few-flowered panicle and small locustae. The other, //. Fraseri, MS., will be described in the Flora of Tasmania ; it is the smallest of any of the southern species, and has a more coarctate panicle of small locustae, resembling that of H. australis, E. and S. ; the florets themselves are marked with purple as in H. paucifiora, Br., the glumes are large, broad and three-nerved, the lower palea silky, acute, with a short awn, the leaves are mnch shorter than the culm.

Mr. Brown's remarks upon the structure of the florets of this genus, in the ' Plantae rariores Javanicre' are very important (vid. fasc. 1. p. 8. sub genere Ataxia), especially regarding the nature of the upper palea of the middle flower. I have described that organ as sometimes two-nerved in H. redolens, which is a strong argument in favour of that author's assumption that "the median nerve is here formed of two confluent cords," and hence that the real nature of the upper palea is the same as in other genera of grasses. Where I observed two nerves to exist, both were rather faint, but distinctly scabrid on the back. The frequently emarginate or bifid apex of this palea in the other species of the genus, where I have not seen a double nerve, is also worthy of remark.

The Hierochloe redolens is a scarce plant in Campbell's Island and of stunted growth; it was not observed in Lord Auckland's group.

2. Hierochloe Brunonis, Hook, fil.; pellicula ovata subeffusa nutante, glumarum valvis sequalibus flosculis longioribus lanceolatis acuminatis basi trinerviis nervis laterabbus brevibus, flosculis masculis 5-nerviis pubescentibus basi pilosis superne bifidis inter segnienta acuta aristatis dorso scabridis margiuibus breviter ciliatis, herniaphrodito apicem versus integrum breviter aristatum piloso, foliis involutis glaberrimis, ligulis late ovatis scariosis. (Tab. LII.)

Hab. Lord Auckland's group and Campbell's Island ; in the former islands growing towards the tops of the mountains, alt. 1000-1-400 ft. ; abundant near the sea in the latter.

Gramen pulchrum, suaviter odoruin, 1-ly pedale. Radix fibrosa, fibris tenuibus, elongatis, fibrillosis. Culmi parce caespitosi, basi inelinati, lignosi, crassitie pennse corvina?, reliquiis pallidis fibrosis foliomm vetustorum obsiti, ad collum ssepe bis terve divisi et surculos nodosos ad nodos radicantes emittentes, superne erecti, gracdes, curvati, foliosi, striati, una cum vaginis foliorum unc. diametr., remote nodosi, nodis constrictis, flavo-fuscis. Folia subdistiche inserta, suberecta, stricta v. paido eurvata, |—1 ped. longa, longe lineari-subvdata, gradatim attenuata, imo apice acuta, involuta, subcoriacea, extus laevia v. obscure striata, glaberrima, nitida, straminea, intus profrmde striata, opaca, sub lente scabenda, pallide viiidia. Vagina erects, subcompressse, culmo appressa latiuscvdee, usque ad