COMMON CKOSSIULL. 
113 
invariably arrive votli a north or north-east wind, hut seem- 
ingly only to rest, altliongli temptations to remain are often 
too strong to he resisted ; for, as far as I can ascertain, they 
nearly always appear first in Unst, and, as a rule, they occur 
in the other islands in only very limited mimhers. During 
the dark nights of autumn, considerahle numl)ers kill them- 
selves against the lantern of the Flugga lighthouse, which is 
placed upon an outlying rock at the extreme north of Unst, 
— a very significant fact with regard to their line of llight. 
I also remember three which met a similar fate at the 
Skerries light on the east coast, hut none were seen in any 
other part of the islands for some weeks afterwards. Solitary 
individuals now and then appear in the gardens, but never 
vdthout others being observed in some neighbouring localities. 
Early in July some very large fiocks alighted at Skaw, in the 
north of Unst, remaining there for many weeks, although the 
spot is exceedingly bleak and bare. A great many were 
killed, and, I am sorry to say, sent to me by the fishermen, 
who, never having seen anything like them before, resolved to 
make the most of the opportunity. As many as twenty-one 
^vxre brought to me by a man who had killed them all at two 
shots. Although the remnant of the flocks remained there so 
long, only very few visited any other part of the island. In 
the stomachs of those which I examined were numerous small 
seeds, mixed with sand and small sharp fragments of stone. 
It is probably from the defect of food that they do not always 
confine themselves to trees. I have seen them about cabbage 
gardens, and even upon the bare hills. At Halligarth I have 
often observed them climbing about the corn-stacks ; on being 
disturbed from whicli, they would merely fly as far as the out- 
house roofs, sitting there in rows until the danger was over. 
The crops of those shot upon such occasions contained small 
seeds and grain ; the latter seems always to be broken before 
it is swallowed. In the garden, rowan-berries were largely 
fed upon ; and by concealing myself I have been able closely 
to observe the mode of feeding. Holding on to a branch, the 
ir 
