Bird‐Migration over the North‐western Part of the Indian Ocean, the Red Sea, and the Mediterranean

Summary. .It is shown on theoretical grounds that so long as nearly all observers travel towards the U.K. in spring and towards India and East Africa in autumn, i.e. , with the general trend of migration, the proportion of the true migration density observed in the Red Sea and south of Guardafui must tend to be less than that observed in the Mediterranean. Many migrants were seen 50–100 miles off‐shore between Guardafui and 3 o S. 9–12. xi. From their course the passerines must have come from S.E. Arabia, with or without a stop on Sokotra. Two Falco spp. and Merops persicus flying east to west may well have come direct from India. Several species were still at sea as far south as the limit of their African ranges, indicating that they were approaching the coast obliquely. Many migrants reach Sokotra, Abd‐el‐Kuri, Pemba, and Zanzibar. Through the last two some travel on passage, cutting across a re‐entrant of the African coast. Only stragglers of a few Palsearctic land species reach the Seychelles or Madagascar, except that many Falco concolor reach the latter from the Near East. Several species have been seen in both spring and autumn in the Arabian Sea 500 miles from land. It is probable that, winds being favourable at both seasons, about six long‐winged land species regularly migrate straight across the Arabian‐Sea, 1600 miles, which would be the longest such crossing in the world. There are a good many records of birds crossing the Gulf of Aden, towards the mouth as well as at the narrow end. Throughout the Red Sea a large variety of birds has been noted; they are least numerous in the extreme north; there is no evidence of concentration on the narrow crossings available at the north and south ends; and apparently many of the birds cross on a slant. In conjunction with the recent work of Bates and Philby the records of birds in the Red Sea and Gulf of Aden leave no doubt that a broad‐front migration crosses the whole of Arabia. Migration across the Mediterranean is surprisingly ill documented. It has been held that migration takes place mainly in three belts and that in good weather birds are not seen far out at sea. Our observations are in direct conflict with these generalizations. We saw ample and varied migration all the way from long. 9 o E. to 28 o E. in perfect weather and between 50 and 150 miles from land. On a general review I conclude that migration between Africa and Eurasia takes place through a great quadrant, one arm of which runs up the Atlantic coast, the other across the Indian Ocean about lat. 10 o N.; and that the movement is of an essentially broad‐front nature throughout.