178 86. GRAMINA. 



Arctic region. Midarctic — Superarctic zones. 



Descends to 900 yards, less or more, in Hebrides ? 



Ascends to 1350 or 1400 yards, in East Highlands. 



Range of mean annual temperature 37 — 33. 



Native. Kupestral, &c. It is impossible at this time 

 correctly to distinguish the true stations for the present 

 species from those of A. csespitosa when found high on 

 the mountains ; the viviparous states of the latter having 

 been very usually designated " alpina " when found in the 

 arctic zones ; and perhaps the same may be also true, to 

 some considerable extent, of the ordinary or non- vivipa- 

 rous states as well. I confess my inability satisfactorilj'- 

 to separate even the specimens preserved in my own 

 herbarium, and that I must consequently distrust my own 

 notes of the stations and altitudes of the alleged two 

 species, made on the mountains, at different dates, from 

 1832 to 1844. As little, I fear, are the published or ma- 

 nuscript records of other botanists to be relied upon. 

 Such as I find them, so is the above formula of distribu- 

 tion filled in, after rejecting some of the least probable 

 stations and altitudes ; as, for example, Wales (British 

 Flora) and Durham (Flo. N. D.), — the sea side near Dun- 

 dee (G. Don), and a pond in Orkney (Dr. Gillies). By 

 way of accounting for the frequent expressions of doubt 

 or distrust in various parts of this work, I would take the 

 opportunity suggested by a difficult species, to point out 

 a strong difference of position, in this respect, between 

 the writer of a Cj^bele and the writer of a Flora. If the 

 latter can only obtain a few good and well-marked exam- 

 ples to be described, he has no absolute necessity to 

 trouble himself or his readers about the doubtful and dif- 

 ficult specimens. With the writer of a Cybele it is far 

 different ; for he must know (as far as it is possible for 

 him to know) to which species each individual specimen is 



