BIRDS AND MEN ON ST. KILDA
influence, the main cause must be a biological one, some as yet un-
explained factor favouring fulmar survival and spread throughout the
range of the species.
The gannets, meanwhile, had not multiplied to the same extent.
The young gannets were the St. Kildans' greatest delicacy and their
capture the islanders' greatest sport. As the birds lay only one egg,
and their total numbers were probably rather less than to-day, their
numbers were held severely in check. But since about 1890 the
gannet too has embarked on a period of increase—not so striking as
that of the fulmar, but none the less definite. Two quite new colonies
have been established in the Shetlands, and there has been a marked
increase in the numbers of birds in the Irish and Welsh colonies.
For the last quarter-century, the increase is in the neighbourhood
of 15 per cent. In this case, too, there seems to have been a wave
of biological increase affecting the species as a whole, in addition to
any local effects caused by the St. Kildans' changed habits.
One of the chief aims of our party was to estimate the number of
gannets on St. Kilda and two other rarely-visited breeding colonies.
This was part of the scheme organized by James Fisher and Gwynne
Vcvers for the enumeration of the world population of gannets—the
first occasion on which a complete census has been taken of any
wide-ranging wild species.
The layman may well ask how gannets are counted. The first
sight of a big colony is bewildering, and a census would seem im-
possible. However, it is eminently possible, as repeated counts by
separate observers have shown. The simplest and best method of
counting gannets is just to count them. Each observer takes a section
of cliff, and goes over it with his glasses, ledge by ledge, counting the
number of breeding pairs. One bird is always on the nest: when
both arc present they will be close together, so that a pair can be
distinguished from the separate sitters.
Direct counting, however, is difficult or impossible from the sea,
unless in a dead calm. You then have to count birds on some
especially favourable section, and estimate the proportion which this
bears to the total area occupied by nesting birds. In some cases a
photographic method is the best—telephoto photographs are taken,
and the birds counted on enlargements from them.
Experience shows that direct counts in favourable circumstances
are accurate to 2 or 3 per cent.; and it can be taken that the world
figure (which provisionally may be put at 166,000 birds) will be
accurate certainly to within 10 and probably to about 5 per cent.
The other two gannetries which we visited were Sule Stack and