by a border of snow, marked by the feet of the old birds, and raised at least two inches above the general level (vol. ii., p. 213). This circumstance somewhat upsets one's notions regarding the cause of migration, which we have been accustomed to consider as dependent chiefly on change of temperature and abundance or scarcity of food. In this instance there seemed to be no lack of food, for there were plenty of fish in the pools between the floes, and the old birds were seen carrying them in their bills. The numerous small crustaceans also, which are common on the shores of the Arctic Sea, would probably furnish sufficient sustenance to the parent birds should fish be unaccessible by reason of the ice.
At page 210 of the Appendix a coloured plate is given of two eggs of the Sanderling, Calidris arenaria, which were found by Capt. Feilden in lat. 82° 33' N., on June 24th, 1876. This nest, from which he killed the male bird, was placed on a gravel ridge at an altitude of several hundred feet above the sea; and the eggs were deposited in a slight depression in the centre of a recumbent plant of willow, the lining of the nest consisting of a few withered leaves and some of the last year's catkins.
The eggs of the Knot, Tringa canutus, he was not so fortunate as to obtain, though it breeds in some numbers along the shores of Smith Sound and the north coast of Grinnell Land. It must be common throughout the Parry Islands during summer, for Sabine found it nesting in great numbers on Melville Island. Captain Feilden and his companions frequently sought for the nest, but without success. However, on July 30th, 1876, the day before they broke out of their winter-quarters, where the) had been frozen-in eleven months, an old Knot with three nestlings were found on the border of a small lake, not far from the ship. A description of the plumage of these nestlings is given on page 212.
On June 21st, the first nest with eggs of the Brent Goose was discovered in lat. 82° 33' N., and subsequently many were found (vol. ii., p. 216).
Ten species of fishes were collected between lat. 78° and 83° N., and were submitted for determination to Dr. Gunther. Of these, two proved to belong to species hitherto undescribed, and have been named by Dr. Günther Salmo arcturus and Salmo Naresii. The former, a species of Chair, is the northernmost sal monoid at present known. It comes nearest to the Charr of Killin (Inverness- shire), but differs from it in having a more slender body, rather
