Not only do the distances of the migration paths of different species vary considerably, from a trip of a few miles to a voyage from the Arctic to the Antarctic, but the individuals of one and the same species do not all travel to the same degree.

The familiar swallow, Hirundo rustica, though subject to certain geographical variations, is found throughout the PalÆarctic and Nearctic regions, nesting throughout Europe to between 63° and 70° north and in Africa north of the Sahara, where, however, Canon Tristram found it also wintering in the oases. South of the Sahara to the Cape it is a winter visitor. In Asia it breeds, according to Seebohm, in Asia Minor, Persia, Afghanistan, and western Siberia, and winters in Scinde and western India. One form breeds in and north of the Himalayas, eastward to China and Japan, and winters in India and Burma, and another ranges from eastern Siberia across Behring Strait throughout North America, so far south as Mexico. This form winters in Burma, in Central America and Brazil, but the Mexican birds are more or less stationary at all seasons.

Our swallow and its congeners have an almost cosmopolitan range, summering in the Northern and wintering in the Southern Hemisphere or comparatively near to the Equator in the Northern. Towards the centre of its range its migrations are either short or the bird is non-migratory.

Mr W. L. Sclater, addressing the South African Ornithologists' Union (42), stated that the swallow arrives at Cape Town at the end of October, and is common from November to March; practically all have left by the middle of April. Swallows begin to arrive from the south in Africa north of the Sahara in the latter half of February; early in March they reach southern Europe, later in the same month they are in Central Europe and by the middle of April large numbers arrive in England. Thus swallows leave South Africa actually after they have arrived in England; the South African birds cannot be the same which are in North Africa a month earlier! The swallow supports Seebohm's thesis that the individuals which go farthest to the south in winter, breed farthest north. A day-migrant and by no means a rapid one, the swallow may be timed from place to place, and it is not presumption to suggest that the birds which reach Britain to nest came from lands little south of the Sahara and well north of the Equator, and that those which pass through England and along our shores in May and even June are on their way from Southern Africa to the northernmost limits of their range. Mr Sclater points out one very interesting fact; when the swallow reaches South Africa it is in ragged worn plumage, before it begins its northward journey it passes through its one annual moult.

Waders and shore birds which reach South Africa in autumn—the spring of the Cape—are moulting into winter dress; before they leave they have often assumed or partially assumed the breeding dress. When they arrive the native South African birds are breeding, but though Mr Sclater thinks that some nest a second time in the south, no satisfactory evidence has ever been brought forward to support the suggestion. These long-distance travellers not only move from a zone of moderate temperature to a warmer one, but many of them pass through the hotter zone to a country having a similar temperature to the one in which they bred, thus enjoying summer but not torrid heat all the year round.

There are birds in which the northern and southern forms are distinct. The wheatear, Saxicola oenanthe oenanthe, reaches us early, sometimes during the second week in March, and speedily settles down to nest. Towards the middle or end of April a brighter larger bird appears, the Greenland wheatear, Saxicola oenanthe leucorrhoa, which was recognised in Greenland, Iceland and eastern North America before it was seen that both forms occurred in Britain. This larger bird loiters through Britain, for its northern home is not ready for it until the Arctic spring. We know it breeds farther north than our wheatear, but its winter range is not fully worked out. The smaller bird is found in north and north-western Africa, and the larger form farther east, even south of the equator on the eastern seaboard, and probably, when we know more about the range of the two we shall find that the form breeding farther north, winters farther south.

The folly of laying down the law on the strength of the knowledge of the habits of a few species is shown by the study of the movements of American birds. Mr Cooke shows that as a rule "the migration is a synchronous southward movement of the whole species" in autumn, "the different groups of individuals or colonies retaining in general their relative position." The black and white creeper Mnistitta varia breeds from South Carolina to New Brunswick, nesting in the south in April and reaching the northern limits in the middle of May. In the middle of July old and young birds have been seen at Key West, 500 miles south of the breeding range, and towards the end of August they have reached the north coast of South America. The New Brunswick birds cannot be ready to leave before the middle of July, and Mr Cooke allows them fifty days for the trip, bringing them to the Gulf States in September; he argues that this is proof that the earlier migrants must have been birds from the southern part of the range. Black-throated blue warblers, Dendroica coerulescens, reach Cuba at about the time that others of the same species are arriving in North Carolina; the first, he concludes, are birds from the southern Alleghanies and the others from northern New England or beyond (20). Other species illustrate the same order which he calls "normal," but show that it is not an invariable rule.

Southern-bred Maryland yellow-throats, Geothlypis trichas, reside throughout the year in Florida; those in the middle districts of the range migrate for a short distance only, whilst the Newfoundland birds pass over the winter home of their southern relatives to the West Indies. The palm-warblers of the interior of Canada travel 3000 miles to Cuba, passing through the Gulf States early in October; those from north-eastern Canada travel later and slowly and settle in the Gulf States, after a journey of only half the distance. He sums up wisely—"No invariable rule, law, or custom exists in regard to the direction or distance of migration.... Each species presents a separate problem, to be solved for the most part only by patient, pains-taking observation and by the recognition of sub-species."

The order in spring is yet unproved. "With many birds ... the first individuals to appear in spring at a given locality are supposed to be old birds that nested there the previous year." These are followed by those which nested a little farther north, followed later by those whose homes are in the most northerly part of the range. "If, then, for any species, the southern nesting birds lead the van in both fall and spring migration, and the near guard in each case is composed of northern breeding birds, it follows that some time between October and April a transposal of their relative positions occurs; and that the more southern birds pass over the more northern ones, which delay their migration, knowing that winter still holds sway in their summer dominions." It is not known where this transposal takes place, nor whether the northern birds remain in winter quarters till the southern birds have passed, or start a slow migration, during which the southern birds pass over them. Later another transposal occurs; the northern birds cross the southern part of the range, passing birds which are already nesting. "Spring migration seems to be therefore for some species a game of leapfrog—the southern birds first passing the northern, and the northern passing them in turn" (20).

The custom, now fortunately becoming widespread, of marking birds by affixing a numbered metal ring to one leg, may help to elucidate this and many other problems, but until a large number of results are collected it is unwise to draw conclusions. Almost every month the recovery of some of these marked birds is noted in the scientific journals, but so far, beyond indicating the minimum distance travelled by individuals, little can be proved. It is, however, plain that birds do not invariably act as they ought to do if they obeyed all the laws which have been invented for them. A few records or results may be quoted, but any suggestions from these must be treated as suggestions only; many more must be forth-coming before we can say, proved.

The white stork, Ciconia alba, has been systematically ringed in Rossitten in East Prussia, in Denmark and in Hungary for some years, and Mr A. L. Thomson gives a brief summary of the interesting results up to date, in "British Birds" for May, 1911. Ten young birds, taken during their first autumn journey, show a general south-easterly trend through Europe. Three east Prussian storks were obtained in Syria, one in the April after it was marked, the other two in April and July of the second year; another was taken in Palestine, and one Hungarian stork in Syria. In the May of the year following that in which it was marked in Prussia one was obtained at Alexandria. In their first autumn Prussian storks have been recorded from near Lake Chad in October, from Rosaires on the Blue Nile in the same month, and from the Victoria Nyanza at the end of November. A ringed bird is reported from German East Africa, but full details are wanting, but one shot at Fort Jameson in north-east Rhodesia in December 1907 had only been hatched in Pomerania a few months before; it left the nest on August 19th and began its journey south on or about the 26th. In its first winter a Prussian bird was shot in the Kalahari Desert.

Seven Prussian and about a dozen Hungarian birds have been obtained in winter quarters in the Transvaal, Natal, and other parts of south Africa, and one in German south-west Africa; one, recovered in the July following the summer in which it was marked, was possibly a weakling bird which had failed to make the return journey. Storks which had returned are recorded in their first, second and third summer; most of them having been found within a few miles of their birthplace. One bird, marked as a nestling near Brunswick in 1906 was reported in June 1908 from Sorquitten in East Prussia, over 430 miles away. Mr Thomson, from his reference to this being roughly in the same latitude but reached by a different "line of flight," suggests that it is an exception; this will be shown when some hundreds more cases have been collected. It may, however, be found that some birds deliberately launch forth in search of new homes, though at present it looks more like a bird which on approaching the breeding area had banded itself with the wrong local body of travellers.

A bird marked at Cassel in western Germany has been recorded from Barcelona in Spain; this so far single record may indicate that storks get lost, that there is no regular direction, or that there is more than one, a south-westerly as well as a south-easterly route. That too, we hope, will be shown in the future.

That much will be learnt from storks is evident, but the lessons will be only general; each migratory species must be treated separately, and to some extent this is being done. In the opening chapter I mentioned the uncertainty about the behaviour of any individual song thrush, merely as an example. So far, the few records of marked song thrushes add to rather than solve this problem. Years ago the song thrush was looked upon as a permanent resident so far as Britain was concerned. Then it was found to be migratory even in Britain, and it was suggested that each song thrush performed a short migration, southern British birds leaving the country, northern British retiring to winter in the south of England, and northern European birds replacing these as autumn immigrants. The study of local races quickly altered this opinion; it was guessed that in Britain there was a sedentary insular race and a migratory passage race; others, however, saw that some of our home-bred birds left in autumn. What do we find? A song thrush, marked as a nestling in July in Northumberland, is found in November in Durham; another one, marked in Berkshire travels to Norwich and is recorded in November, but a third, born in Aberdeen takes an autumnal flight of at least 1500 miles and is found in Portugal. Evidently we cannot yet frame any rule for our British-bred birds.

It is said that home-bred lapwings are somewhat sedentary, and that the large winter flocks are composed of Continental immigrants. The frequent westward migration of lapwings during exceptionally severe winter weather has led to the supposition that these birds fly for refuge, under these circumstances, to Ireland. This is true, so far as it goes, but a lapwing marked as a nestling near Stirling has been found in the south of France, and two others in Portugal, whilst five have been recovered in Ireland.

The results of marking sea-birds are interesting, showing that the young birds often wander northward in search of food before there is any marked autumnal southward migration. Terns and black-headed gulls have been found a month or more after they have left the nest to the north of their breeding colonies in Cumberland and mid-Wales. A bird from Ravenglass was taken in its first January in Brittany. Rossitten black-heads have been shot in the Isle of Wight and in Breydon in Norfolk.

This may only mean that the young blackhead is a confirmed wanderer in search of food, but the few results with woodcocks, marked as British-bred nestlings, are puzzling. They have been known to linger in the neighbourhood of their home until November, and have been found in Portugal only a month later. Birds marked at Tyrone have been found so far apart as Cornwall, Harrow and Inverness; what route for the Irish birds can be guessed at?

Birds marked as adults present further problems, but also provide interesting evidence. Hooded crows, captured on migration in spring at Rossitten and then released, have been recovered in autumn actually in the same place and in other localities in Germany, and one marked in October was taken two years later, in spring, in Finland. The sum of these records of crows proves one thing conclusively—the fallacy of GÄtke's due east to west and west to east flight, and supports a coastwise migration for this species.

Adult teal, captured in decoys, ringed and released in South Denmark in September and October, were taken in November and December in Hampshire, Suffolk and the Moray Firth, whilst others from the same place were recorded from other parts of England and Ireland, from western France, Holland, the south of Spain and the north of Italy. Fly-lines, if followed, are divergent and complicated. Four young herons were marked in one nest in Denmark; one was recorded in Holstein in June, and another in Mecklenburg in July; the third was killed near Salisbury in Wiltshire in October, and in the following February the last was obtained in the north-west of France. Two from another nest were recovered in Denmark, one in July and the other in February, twelve months after birth. Another heron reached Andalusia by August. In each case where there was indication of a direction it was south-westerly. Many more records might be mentioned, but these are sufficient to show the value of the method and the present insufficiency of results.

Many of these records show that the speed of the migrating birds, even in spring, is not great. Mr Cooke proves that most species in North America travel slowly through the districts where food is plentiful and during the earlier part of the journey northwards only a few miles are covered per day; they travel with the slowly advancing vernal wave, but, as we shall show in the next chapter, many species actually outstrip it, and travel from warmer to colder climates.

Map showing the range of the American Golden Plover, with its known migration route.

(From The National Geographic Magazine.)