60 THE SCOTTISH NATURALIST 
S 
Sannox Wood and the district immediately round the 
mouth of the North Sannox Burn, 26th to 28th May. 
Ca = Glen Catacol, 29th to 30th May. 
Cn = Correin Lochan, 30th May. 
P = Pirnmill, rst June. 
M = Machrie Water, about three miles from the coast, 2nd June. 
Cg = Bogs and cliffs by the shore below Corriegills, 2nd to 
4th June. 
W = Catacol district (Waterston, Sept. 1920.) 
T = Bonawe, near Taynuilt, Loch Etive (Waterston, Aug. 
I9IQ). 
E = Head of Loch Etive (Waterston, Aug. 1919.) 
V = Recorded by Verrall in the paper quoted. 
| 
As we had been led to expect, midges were exceedingly 
abundant and troublesome. At the camp site near Pirnmill 
they continued their attacks all night long, and shortly after 
dawn made such a furious onslaught that they compelled us 
to evacuate the spot and seek refuge in the breeziest place 
we could find on the open beach a mile or two away. This, 
however, was on an exceptionally sultry night, and the 
experience was not repeated. As had been anticipated, the 
worst enemy was Culicozdes arcuatus, which far outnumbered 
all the other species of the genus; this seems to be the 
case in all hilly districts. Of C. nebeculosus, which was 
troublesome at Dreghorn, Ayrshire, only a single specimen 
was met with in Arran. 
The new species diagnosed below as C. heliophilus had 
the unusual habit of flying (and biting) chiefly in the hot 
afternoon sun. By 6 P.M.(summer time) it had disappeared, 
and was being replaced by the hosts of C. arcuatus. 
Subsequent experience on Dartmoor seems to indicate 
that C. heliophilus may chiefly attack sheep, but no observa- 
tions on this point were made in Arran. 
C. pulicaris was common, and exhibited great variation 
in the wing-markings; on this character four fairly well- 
marked varieties could be distinguished, which were at first 
believed to represent distinct species, but no_ structural 
differences were discoverable between them. 
Some remarks on the Simuliidze of Arran have recently 
