1915] St. John, — Elymus arenarius 101 
alterations of the name by different workers from E. mollis to E. 
arenarius and back again. An examination of the various North 
American Floras discloses a similar vacillating policy in the treatment 
of E. arenarius and E. mollis. Many contain such comments as 
"Cette Espéce est trés voisine de l'Espéce Arenarius d'Europe" 
(Provancher), or “ Near E. arenarius” (Gray), or “the confusion that 
seems to exist between this species and E. mollis" (Macoun). 
Specimens of FE. arenarius, var. villosus from the southern part of the 
range, the coast of New England or California, are superficially enough 
like E. arenarius of Europe to be confusing. They have the glumes 
sometimes nearly glabrous and scarcely soft in texture. On the other 
hand, specimens from the far north are strikingly different from Æ. 
arenarius because of the extremely soft texture of the glumes and their 
heavy coat of long villous pubescence, but because of the perfect series 
of transitions these two tendencies seem to be due to climatic factors 
and to be negligible from the point of view of classification. This 
whole series of plants representing KE. arenarius, var. villosus has, 
however, the positive distinguishing characters of culm velvety- 
pubescent at summit, and glumes pliant, pubescent, and prominently 
3-5-nerved. The great mass of material of the European E. arenarius 
has the glumes stiff, hard, with a prominent mid-nerve, with or with- 
out less developed lateral nerves, and with the summit of the culm 
glabrous. If this were an absolute rule the American plant called var. 
villosus by Meyer would seem to be a distinct species, but occasional 
specimens of E. arenarius from Europe show a trace of puberulence at 
the summit of the culm. On the other hand, occasional specimens of 
the plant from the northern American or Asiatic coasts, although 
conspicuously pubescent at the summit of the culm, have the lateral 
nerves of the glumes less developed than in more typical specimens. 
Because of these transitions and the identity of the anthers and grains 
it seems best to treat the American and Asiatic plant, with the culm 
velvety pubescent at summit and the glumes 3—5-nerved, as Elymus 
arenarius L., var. villosus E. Meyer. 
The geographic distribution of E. arenarius, var. villosus is very 
extensive as it grows on the beaches and brackish shores from Alaska 
south along the coast to Santa Cruz, California, on the Arctic Archi- 
pelago, the shores of Hudson Bay, and of Lake Superior, and from 
Labrador south along the coast to Cape Cod, Massachusetts, on Green- 
land, and from Bering Straits south to Japan. This statement of 
