1900-1902. No. 2.] VASCULAR PLANTS OF ELLESMERELAND. 173 



Aira caespitosa, L. 

 var. arctica, (Trin.) m. 



A. arctica, Trinius, Gram. gen. et spec, 1831; Rothrock, FI. Alaska; Deschampsia 

 brevifolia, R. Brown, Chlor. Melv., 1823; Greely, Rep.; Hooker, F1. Bor. 

 Amer.; Seemann, F1. W. Esk. land.; Ledebour, F1. Ross.; Aira brevifolia, 

 Lange, Consp. Fl. Groenl.; Hartz, Fan. og Karkr.; Simmons, Bot. Arb. ; A. 

 caespitosa var. brevifolia, Gelert, in Ostenfeld, Fl. Arct.; Kjellman, in 

 Vegaexp. ; Trautvetter, Consp. Fl. Nov. Seml. ; A. caespitosa var. borealis, 

 Tbautvetteb, 1. c. ; Andersson & Hesselman, Spelsb. kSrlv. (?); Aira caespi- 

 tosa var., Trinids, Spec, gram.; A. caesp. f. alpina, Kruuse, List E. Greenl. 

 (ex p. ?) ; Deschampsia caespitosa, Habt, Bot. Br. Pol. Exp. ; non Aira brevi- 

 folia, Pursh, Fl. Amer. sept.; nee A. caespitosa 'brevifolia, M. v. Bieber- 

 STEIN, Fl. Taur. Cauc. ; nee A. caesp. var. brevifolia, Hartman, Skand. FL, 

 Ed. 2; nee Nathofst, N. W. GrSnl. 



Fig. Thinius, Spec. Gram., 3, T. 256; Tab. nostra 9, fig. 7. 



At the first glance, the plant which Rob. Brown, 1. c, p. 33— 34, 

 described as Deschampsia brevifolia seems so very unlike Aira 

 caespitosa, that one does not even think that they belong to the same 

 genus. But on closer examination it appears, that only relative charac- 

 ters separate them: the arctic plant is small and short, coarsely built, 

 with short, often somewhat involute leaves and coarse more or less inflated 

 sheaths. The culm leaves, especially the uppermost one, are very short, 

 sometimes reduced nearly to the sheath alone. The panicle is generally 

 densely contracted, almost spike-like. Brown, who had only seen a 

 few specimens collected in Melville Island, could hardly realize that a 

 plant of so different a habit could belong to A. caespitosa, and still it 

 would be tempting enough to uphold it as a species, did not intermedi- 

 ate forms connect it with the type of the species in America as well 

 as in Asia. Hooker, 1. c, records a /? major, which "seems almost to 

 unite Z). brevifolia and D. caespitosa", and Trinius, who doubtless 

 had had a larger material of intermediate forms at his disposal, does 

 not hesitate to reduce the species of R. Brown to a form of the latter 

 (Spec, gram., 3, T. 256). The same is done by later authors, who have 

 studied those connecting forms (cf. works of Trautvetter, Kjellman, 

 and others quoted above). 



Authentic specimens of the grass, which Trinius, Gram. gen. et 

 spec, p. 56, calls A. arctica, I have not seen, but on the authority of 

 Trinius himself (figure quoted above) and Ledebour, 1. c, 4, p. 422, I 

 must assume it to be the same form as Brown described. Then the 

 name of Trinius must be used, as that of Brown must be cancelled 

 on several accounts, viz., Pursh had already, 1814 (1. c), described an 

 Aira brevifolia from "the plains of Missouri" with the addition "this 



