126 C. H. Ostenfeld. 
is necessary in order to come to a decision upon that point. I 
should like to suggest to British botanists an examination of the 
material in the large British herbaria (anatomical examination 
would also be required) in order that the distribution may be better 
known. Provisionally 1 should think that var. austriacus is confined 
to the moors (mostly Eriopliorum- moors) of north England and 
Scotland, while var . germanicus is the common form of the British 
Isles, at least in the southern and western parts. 
In Denmark we have both varieties, but var. austriacus is rather 
rare and occurs almost only in deep Sphagnum- moors, while the 
common form, var. germanicus, often grows in drier localities 
(heaths, etc.) 
As to their general distribution I have seen specimens of var. 
austriacus from North America, Greenland, Iceland, Scotland and 
North England, Scandinavia, Denmark, Russia, East-Prussia, 
Silesia, Bavaria, and the other mountainous central-European 
countries, and of var. germanicus from the German plain, the 
French plain, Denmark, West-Norway (?), England and Ireland, 
the Hebrides and Faeroes (where also specimens occur which 
approach the other variety). Thus the latter has a much more 
restricted geographical area than the former. 
I have omitted to mention the anatomical distinctions as l have 
not had time to examine my material in that direction, and may 
refer to Palla’s paper quoted above. 
Gramine/E. 
Arundo phragmites L. 
During our visit to the Norfolk Broads the giant form of 
Arundo phragmites growing especially in Surlingham Broad 
attracted our attention ; the leaves of the flowering stems were 
about 4 cm. broad. As far as I am able to find out by examination 
of my material this form comes very near to the southern race of 
Arundo : A. isiaca Del. ( Phragmites gigantea Gay, P. altissima 
Mabille) which is common in the Mediterranean region; but the 
latter has mostly brownish, not dark-violet, spikelets. Ascherson 
and Graebner (Synopsis II, 1, p. 330) mention a var. pseudodonax 
Rabenh., of which I have not seen specimens, but which, according 
to the description, answers fairly well to the Norfolk form. It 
would be interesting to learn, if we really have a special (southern ?) 
form here, or only giant specimens of the common plant, caused by 
the favourable conditions of growth. 
Poa trivialis L. var. suhalpina Beck (FI. N. Oesterreich, 1890, 
p. 86). 
Along old watercourses on Crossfell (Cumberland) and, as far 
as I remember, also on Ben Lawers (Perthshire) we found an 
interesting small form of P. trivialis growing among mosses. The 
panicle is tinged with violet and rather small; the stolons are not, 
as ordinarily, borne above the surface, but are buried in the mosses. 
I think it may be identified with the above quoted var. sub- 
alpina Beck, which has the following description: blade of the 
uppermost stem-leaves very short, not reaching the panicle ; spike- 
lets more or less violet-tinged; grows in springy (“ quelligen ”) 
places (1200 m. alt.) 
