115 
“Lapponia occidentalism Suecia borealis, Norvegia alpes- 
tris,” and queries it as = T. alpinum Don (Scot). 
This suspicion I now think is amply confirmed. T. viola- 
ceum is evidently nearly allied to T. caninum, and differs 
from it principally in being of more compact growth, shorter 
and somewhat narrower leaves, fewer flowered spikelets, with 
conspicuous purple tinge, this latter circumstance suggesting 
the trivial name. The awns not as long as the pales, excepting 
occasionally in case of the two upper spikelets. 
Root fibroso csespitose, not creeping. This fact militates 
against Sir J. D. Hooker’s theory that Don’s plant is but a 
form of T. repens. 
T. biflorum (Mitten), specimens of which I have not seen, 
is probably the same plant. This must not be confounded 
with the biflorum of Brignol, a native of W. Tyrol and Italy: 
but Mr. Charles Bailey, F.L.S., possesses specimens of typical 
T. violaceum from the Tyrol. These are of peculiarly luxu- 
riant growth. 
It is always satisfactory to re-discover a lost species, but 
more especially one of George Don’s, whose memory has of 
late often been the subject of opprobrium amongst certain 
botaidsts who failed to detect species recorded as found 
by Don, and Don only. Let us hope that after all Ranun- 
culus alpestris, Caltha radicans, and other similar lost species 
may, in like manner, reward the future searcher. 
II. Descharnpsia fiexuosa (L) var: c. V oirlichensis (Melvill). 
Differs from var. b. montana (Hudson) by its possessing 
three perfect florets in each spikelet. 
Hitherto the old Aira flexuosa (L) has been placed in the 
subgenus Avenella (Koch), one of the chief characteristics of 
that subgenus being two perfect florets, but no rudimentary 
third one, as opposed to Deschampsia, another subgenus of 
Aira, where the rudiments of a third floret are occasionally 
present 
This curious variety of a well-known species I found early 
