Migration Routes

Evidence of Early Man in North America

By: Edgar B. Howard

Originally Published in 1935

View PDF

We have tried to assemble the latest interpretations of certain geological terms which are intimately connected with the problem of early man, and we have tried to give as complete a picture as possible of certain types of stone artifacts which have become recognized in this country as being the earliest we know about. Equipped thus with definitions which, though they may be open to criticism, we believe will answer satisfactorily the purposes of this paper, we can now continue our search for the answer to the two questions which have prompted so much inquiry: When was America first peopled, and where did these people come from?

It seems to be generally agreed that the first people came from Asia, but how long ago is still an open question. Though there are those who postulate a sunken continent in order to account for man in the New World, it hardly seems necessary to go to such extremes when more readily understandable evidence is present.

There are those who believe that trans-oceanic voyages in “war canoes” were made from Pacific islands by Polynesian peoples. Dr. R. B. Dixon takes up this subject,52 in which he has pointed out that the Polynesians unquestionably demonstrated their abilities as navigators, and that, though it might have been possible for them to have reached the coast of America from Hawaii or from Easter Island, a distance of about two thousand miles, it would have been a very rare occurrence, since that distance would have been at the extreme limits of their range as navigators, as based on tradition, and nearly twice as long as any known to have been made in historic times. Certain cultural traits which are found in the western hemisphere, especially in South America, and on the northwest coast of North America, have led some to the belief that these traits are similar to traits in Polynesia, or Melanesia, or in some cases, to those in southeastern Asia. It is not necessary to become involved in the discussions of the “independent inventionists” and the “diffusionists”, but we believe that it is generally admitted today that, though it is not impossible for an occasional boat to have reached the shores of America from the far Pacific Islands, and that some of the traits found in South America may be explained as having been introduced that way, the chances for migrations were better by way of Bering Strait, and further that, though some traits may have been thus introduced, others may be explained as having developed in the New World. Polynesian diffusion seems to have been recent and there appear to be no American languages connected with Polynesian.

One is apt to speak rather loosely of migrations of peoples and animals, without giving much thought to the causes. People in large numbers do not move great distances unless there is some compelling force—pressure of environment or pressure of enemies. In the case of a sedentary people their attachment to some particular spot of earth is so great that they re-main there under severe adversity. People will return to the foot of an active volcano after it has wreaked destruction on their homes and lands. Farmers will remain with their house half covered with drifting sands to plant a crop the next year or the next. There are many other familiar examples. But with the more primitive peoples who have not settled down to a life of agriculture, their home is where they can secure food by hunting and with the least effort. The chief force to make them move in large numbers to a new land is that their food supply has moved away ahead of them.

Many other causes come to mind, but it would seem logical to assume that most of primitive man’s time was taken up with the search for food, and that the herds of four-footed animals supplied this want with the least effort on his part, and that to get at the cause of his migrations we must find out why the animals moved away. The simplest explanation is that the food supply of these animals became less abundant. If we follow the main thread and are not diverted by the many lesser threads woven into the pattern, we shall arrive logically at a point where we can say that climatic change was the chief cause back of the migrations of the early hunters as it was for the movement of the large herds that they followed.

Brooks53 covers this subject in considerable detail, among other data listing migrations in Asia from 5,200 B.C. to A.D. 50. He points out that these movements are almost always from drier to wetter regions. He does not maintain this as a general principle, however, since he gives examples where, due to increased rainfall, migration will take place towards drier regions. Climatic fluctuations since the last glacial maximum, he believes, have played a much more important role in causing movements of peoples from one place to another than has been generally recognized.

There seems to be no doubt that during historic times ample evidence exists to show that man’s response to unfavorable environmental conditions has had a stimulating effect on him, providing the change was not too harsh. Toynbee54 has made out an excellent case to show the operation of his hypothesis of “Challenge-and-Response”, and since it applies so well during all of historic times, there is no reason to believe that our early hunters from Asia did not likewise respond and react, even more readily to the great climatic changes involved in the restoration of the climatic balance after the dislocation of the last ice-sheet. Huntington’s55 investigations upon this same subject, as applied to Asia and America, marshall a great many facts demonstrating the all-important element of climate back of migrations.

There is necessarily much that is speculative in our present state of knowledge of the subject of man’s arrival in America, but there are, nevertheless, a good many facts which can be assembled and which, when brought to bear on the problem, may help to strengthen some one main hypothesis which is more or less generally accepted, such, for example, as man having come to America by way of Alaska from northeastern Siberia as a hunter.

Let us examine some of the facts which throw light on this theory, and, in so doing, we may find that the geological evidence for or against this assumption will help us to decide when and how such migrations were possible. In other words, the determination of what barriers to man’s migrations existed and when they were removed will help to delimit the time factors involved.

The chief barriers to be considered are ice-sheets and mountain glaciers, so that at once the problem is thrown into the midst of glacial geology. Phillip Smith,56 Chief of the Alaska Division of the United States Geological Survey, presents some facts pertinent to this discussion. In the first place, the distances are given to various islands, such as Diomede Island, off the nearby coast of Alaska. The total distance from Cape Prince of Wales to East Cape is shown to be fifty-six miles. The much greater distances between Siberia and Alaska by way of the Aleutian Islands, and the greater distances between these islands themselves, as compared to that at Bering Strait, is pointed out. The Chief of the Alaska Division goes on to say: “… the writer wishes to emphasize . . . that the land areas of the two continents stand close to each other and have so stood for a long time. In clear weather the two land masses are intervisible, and even under moderately adverse conditions the islands can be plainly seen from the nearby mainland shore.”

He does not overlook the possibility of crossings by way of the heavy sea ice, and, therefore, does not see any great obstacle to man migrating to America by way of Bering Strait. To quote Smith again : “From a geologist’s point of view, the Alaskan and Siberian parts of the continents of Asia and North America are even now continuous, and their apparent discontinuity in the part where they are covered by Bering Sea and the southerly re-entrant of the Arctic Ocean is regarded as not disproving this concept. Examination of a chart on which are recorded the depths of water throughout Bering Sea and the Arctic Ocean shows that in an area whose southern boundary may be taken as marked by a line extending north-westward from Unimak Pass to a point 30 miles south-west of the Pribilof Islands and thence to a point 30 miles south of Cape Navarin on the shore of Siberia, and whose northern boundary lies approximately in latitude 72º, or farther north, the depth of the sea nowhere exceeds 100 fathoms, and through more than two-thirds of that area the depth does not exceed 30 fathoms. This remarkable underwater platform is more than 1,000 miles across from north to south and links the land masses of the two continents together as a unit.”

Pointing out that old shore strands have been found in the vicinity of Nome, the lowest of which so far disclosed is thirty-four feet below the present sea-level, Smith thinks that more work along this line will have to be undertaken before geologists can be sure in assigning a date to these ancient beaches. One further fact that we quote from Smith: “So far as investigations have been carried throughout the central part of Alaska, all the evidence that has been discovered indicates that at no time in the long interval since the Mesozoic era has that region been extensively glaciated.”

S. R. Capps, also of the United States Geological Survey, has contributed additional facts concerning glaciated and unglaciated areas in Alaska.57 With this paper are included maps showing the extent of the Pleistocene glaciation in Alaska. It should be noted that on these maps dotted lines indicate that insufficient data are on hand to establish whether the ice-sheet from the Keewatin centre actually coalesced with the mountain glaciers during the last glacial stage or whether there might not have been an open corridor from central Alaska through the upper reaches of the Yukon into British Columbia and thence into the Great Plains, such a postulated passage way, for example, as Upham58 shows on a map of the glaciated area of North America as running approximately from the Arctic Circle and longitude 137° to the United States-Canadian border and longitude 115°. Capps further states that there is evidence that the glaciers, along the Pacific coast of Alaska, during Wisconsin time pushed out to the open sea in many places. The increase in the glaciers at that time, he observes, would have caused hundreds of them to have coalesced and thus to have formed an ice-barrier comparable to that found along the northern borders of the continent.

Johnston59 gives us a clear picture of the geological stage as set for the migrations of animals and man into the New World. He sustains some of the evidence as already deduced by earlier workers in Alaska and Canada, at the same time not hesitating to point out the vast amount of investigation necessary before we can be sure when that region became sufficiently stabilized to place no obstacles in the path of such early migrations. By way of corroboration Johnston points to the great extension of ice along the Pacific coast which would have made travel by that route anything but easy. However, it must not be forgotten that the later Eskimo found no difficulty in overcoming ice as a barrier.

We must admit that it cannot yet be proved whether the first waves of migration consisted of fishing peoples or of some early ancestors of the Eskimo, or hunters more allied to Plains-living tribes. We can only say that the earliest evidence of man in America shows that the stone artifacts he made were associated with the larger mammals, particularly the bison, so that with this to guide us we may not be far wrong in saying that the first immigrants were hunters who followed herds of bison and other animals into Alaska, though, of course, these animals very likely migrated back and forth between Asia and America long before man began to follow them.

Johnston also observes that glaciation in Alaska and in Yukon territory, referring to the Wisconsin Stage, was confined to the mountains and the adjacent parts of the lowlands. He states: “The great central plain drained by the Yukon, and the lowlands bordering Bering Sea and the Arctic Coast, were unglaciated, probably because the precipitation was too small during the Pleistocene for the formation of ice-fields on these lowlands.” Such facts are rather pertinent to the question of possible routes of migration by early man. Further Johnston states: “If the ice covered only the Eastern part of the Great Plains, there may have been a broad belt ice-free for many thousands of years East of the Rockies and extending
down the Mackenzie Valley.” Though there is no general agreement, due to lack of sufficient information regarding the limits of the ice borders, there does seem to be some ground for believing that the drift sheets south west of the Altamount Moraine, which marks the western limit of the Wisconsin ice-sheet, are pre-Wisconsin in age.60

Therefore, it seems likely that the Wisconsin ice may have stopped short of the mountain ice from the Rockies, and thus have left a belt of two to three hundred miles in width between the plains area and the Moraine. Johnston says: “. . . it is evident that a broad belt East of the Rockies was uncovered by the ice at an early stage; the time . . . may have been 25,000-30,000 years ago.” Coming to the possibility of a land connection at Bering Strait, Johnston assumes that it is necessary to have such a connection in order to account for the migrations, particularly of plants and smaller mammals, conceding that the mammoth, bison and horse could have crossed on the ice. Remains of these animals, as well as those of the musk-ox, have been found in deposits of the coastal plain of Alaska.

In considering the question of a land connection between the two continents, Johnston reaches the same conclusions as other investigators that the general level of the sea was lower, owing to the accumulation of ice on the land, and that this lowering amounted to about one hundred and eighty feet. This would allow a land connection during the height of the last glaciation, but he feels quite certain that there has been no land connection in Post-Glacial times, though there may have been oscillations of sea-level during the waning of the ice.

Daly61 gives approximately the same figure for the lowering of the sea during maximum glaciation. Croll62 gives thirty fathoms at Bering Strait and goes on to show how dry lands resulted there when the water level was lowered due to the shifting of the gravitational centre of the earth. Penck63 even earlier recognized the question of lowering of sea-level as a consequence of withdrawal of water to form ice caps. When the ice melted, the sea-level rose, and so did the land when relieved of this enormous weight, but Johnston points out that the rates were different, the sea-level lagging behind. These all are important factors in connection with our problem, and it is to be hoped that more information of this kind will be forthcoming as geological work in those regions progresses.

In concluding his paper Johnston says: “If the north-eastern ice-sheet began to retreat from the Atlantic coast near New York about 40,000 years ago, as seems possible, retreat also began in the Great Plains region at about the same time or earlier. A wide belt east of the foothills of the Rockies and extending into the Peace and Mackenzie River valleys may have been ice-free for nearly as long a period; but it is not certain that the whole valley of the Mackenzie River was uncovered at such an early stage, for the ice may have occupied parts of it until a fairly late stage. As already pointed out, however, the drainage of the region in late-glacial time appears to have been to the Arctic or to the east, which suggests that the land was uncovered in this early stage, at least along one side of the valley. It is possible, therefore, that a route for migration through the Cordilleras may have been possible more than 10,000 years ago. These western routes probably have been open much longer than any possible route in north-eastern North America either by way of the Labrador coast and Gulf of St. Lawrence or through Hudson Bay.”

Dr. Ernst Antevs has probably contributed more than anyone in recent years towards the geological aspects of the problem of how and when man first came to America, and it will be well to see what he has to say.64 In one of his latest publications65 Antevs says: “According to the best of our belief . . . the first man to arrive in North America was of modern type and probably at the Neolithic stage of culture. He came from Northeastern Asia to Alaska and probably spread along the eastern foot of the Rocky Mountains where an ice-free corridor had formed some 20,000-15,000 years ago. He seems to have reached the Southwest at the age of transition between the pluvial and the post-pluvial epochs, roughly 12,000 years ago. Reported older finds need verification.”

Summarizing the geological factors involved in such migrations, Antevs points out that “the Keewatin ice joined the Cordilleran ice near the Rockies at about the 55th parallel but that an ice-free belt appeared on the western border of the Great Plains and on the lower Mackenzie River soon after the ice retreat had begun.” He further states: “As was pointed out by Johnston, the earliest possible, i.e. late-glacial, land route from unglaciated Alaska to Central North America led eastward to the Mackenzie and thence southward along this river and the Eastern foot of the Rockies. This explains why most of the significant finds of earliest man in the United States have been made on the Great Plains. The trail was probably opened 20,000-15,000 years ago. Possibly this route was also open for a short time between the first and second expansion of the Keewatin ice sheet 45,000, 40,000 or 35,000 years ago.”

Antevs gives the sea levels at Bering Strait as approximately two to four hundred feet lower than today during the maximum extent of the Wisconsin ice, and about one hundred and twenty feet lower some twenty thousand years ago. Antevs believes that during the last glaciation “there was a barrier of high glaciated mountains (in Siberia) extending in a bow from near the mouth of the Lena River to the Sea of Okhotsk and thence following the shores of the Sea northward and eastward”, and that man would have found, therefore, northeastern Siberia a discouraging country through which to travel till the glaciers became considerably reduced.

There is increasing interest in the idea that the Wisconsin stage was not represented by a single ice sheet, but was instead a multiple and complex affair with several ice centers all of which were not contemporaneous. Such a theory would give greater leeway in the matter of corridors, since the Cordilleran ice cap might not have been contemporaneous with the Keewatin during the whole Wisconsin. Keyes,66 for example, has maps to show that the contemporary ice caps of the Patrician centre were the Greenland and Cordilleran ice caps, not the Keewatin. This revelation, according to Keyes, that the Patrician glaciation was altogether distinct from the Labradorean and Keewatin ice masses, explains the inequalities of the commonly recognized till sheets and many other puzzling questions. He says : “Instead of a single gargantuan ice-mass covering half the northern hemisphere and miles thick there was actually a medley of relatively petty ice-caps, widely separated from one another in point of time and space . . . and was a composite to the extent of being a four-stage, five-centered, ten-glacier affair. . . .” Evidently here is a fruitful field for more intensive research.

One thing which should be borne in mind in connection with migration routes from Asia is that the open corridors to and from the Strait probably did not exist at the same time that the land connection existed. As the ice melted and the corridors opened up, the sea filled, and the land connection was broken. The lack of a land connection does not seem to be so important a barrier as the closed routes to warmer regions, as man could have crossed the Strait in winter time over the ice, or in boats when the ice melted.

Antevs67 further points out: “As the Cordilleran ice sheet decreased in extent, cutoffs were opened along the upper Yukon River and its tributaries, the earliest leading out on the plains via the valleys of the Liard and Peace river systems and a later route leading due south via the Fraser. The Pacific Coast route was also opened for travel by land most of the way. . . . About 9000 to 8000 years ago the temperature at some distance from the ice remnants may have reached the present level. It continued rising and from about 7500 to 4000 years before our day was distinctly higher than now in northern Europe and probably also in North America. Some S500 years ago it had dropped to its modern level, at which it has since remained except for a somewhat lower stand between 500 B.C. and 800 A.D.” There is no doubt that conditions for man’s spread through the Arctic regions of Siberia and Alaska were least difficult during this warmer stage a few thousand years ago than immediately following the time when the ice-sheet began to retreat.

Summing up some of these facts, then, we see that it was possible for man to have reached this continent without encountering insuperable obstacles, such as an ice sheet, approximately 40,000 years ago, and again approximately 20,000 years ago, but that 15,000 to 10,000 years ago it would have been increasingly easy so far as the routes to and from Bering Strait are concerned, and easier still about 5,000 years ago during a somewhat warmer time.

Antevs68 has gotten together a very extensive bibliography on the subject of Pleistocene glaciation in Asia. There is also a large bibliography on Siberia gotten up by Jochelson,69 though many of these references are in Russian. Jochelson points out that northern Siberia was favored by a much milder climate during the Neolithic period than at present, and that this belief is corroborated by Sukachev’s70 investigations in finding in the turfs of the Karsky-Tundra remains of firs, pines, larches, birches, and alders, referring this period to the time following the retreat of the last glaciation. A warmer climate is also suggested by the zoologist, Byalynit-sky-Birula.71

In spite of the numerous European references to the glacial geology in Siberia, there are not many facts available in English upon the subject. The result is that one finds rather conflicting opinions about the extent of the ice in Siberia during the last glacial stage. Mather,72 for example, does not conceive of Siberia as being buried under an ice mantle, but states: “At their maximum spread the Siberian ice sheets were trivially small when compared with those of Canada and the Northern States.” Clearly, then, here is something about which more facts are needed; at least, more information should be available in English or other western European languages.

Mather says further that the invasion of America may have occurred just prior to the spread of the last ice sheets, during the maximum of the Wisconsin stage, or after the ice began to retreat.

We are indebted to Mr. Lauriston Ward, of Harvard University, for calling our attention to the intriguing hypothesis of Von Eickstedt,73 who believes that people were locked up in central Siberia, encircled by mountain glaciers and uninhabitable areas for several thousand years, and as the ice barriers melted corridors were opened in Post-Glacial times into Europe, southeastern Asia, and northeastern Siberia, through which people of already mixed origins spread in all directions.

Taylor74 has developed a concept of anthropology on an ecological basis, in which he finds that the ice-sheet was the prime mover in forcing peoples back and forth as the ice advanced and retreated. It is the ice advance that drove man out of central Asia, according to Taylor’s reasoning, the Caspian Sea being a region of stimulus for migrations into Europe, Africa, India, Australia, and northwestern Siberia. He states that in Azilian times and later, approximately 8,000 to 5,000 B.C., Europe was, at times, warmer than today, and northeastern Siberia and Bering Strait may have had a much more attractive climate than today. He says: “This period probably saw vast hordes of earlier Alpines pouring into Alaska and nearby American lands, as they were thrust onward by the expansion of later Alpines in Asia.”

In Siberia there is a vast field still largely untouched archaeologically, and it undoubtedly holds the key to the problem we are here concerned with. Jochelson75 has given us some light on Neolithic and Palaeolithic sites from the Ural Mountains to the Pacific, with a concentration in the neighborhood of Lake Baikal, particularly along the Lena River and its tributaries, and along the Yenisei to the west, a region where so many rivers have their sources. Many discoveries made in Siberia have a striking resemblance to some of those reported in America. For example, there is the discovery made in 1896 near Tomsk, mentioned by Jochelson, of bones of a young mammoth that had been killed and eaten by Palaeolithic hunters. Ashes, charcoal and fragments of stone knives and burnt leg-bones were found, but no trace of man himself. Jochelson stresses the discovery made by Petri76 at Lake Baikal of Neolithic sites, particularly that of Ulan-Khad.77 The cultures discovered by Petri on Verholensk Mountain showed a great mixture (using European terms) of Aurignacian and Mousterian scrapers, of laurel-leaf stone blades typical of the Solutrean, harpoons of the Magdalenian period.

Jochelson refers the upper Palaeolithic remains to the close of the Pleistocene period, the Neolithic being found in layers of recent geological age, the flora and fauna nearly the same as at present. All the Neolithic sites were located on river banks or on the shores of vast water basins. The Neolithic stone industry of Siberia, Jochelson says, may be characterized by the small size of its artifacts, particularly of arrowheads and scrapers. These may be distinguished, he says, by a perfect and finished retouching, the material usually being quartzite, but many of flint, which is not abundant there. Of interest is the statement : “The most widely distributed type of stone arrow-point is of a triangular form with a somewhat grooved base for setting into the shaft. There are also points of lance-leaf and willow-leaf types.”

Of the osteological remains Jochelson states that they show there were probably several narrow-headed varieties in northern Asia in the Neolithic period. Craniometric measurements give evidence of the existence of one physical type, a long-head, not corresponding to the type of the modern population of southern Siberia.

The so-called “Dune Dwellers”, whose stone tools were found by the Central Asiatic Expeditions under Roy Chapman Andrews,78 are described as having been widespread in Mongolia. The type site at Shabarakh Usu showed evidence that Mongolia once had a much more hospitable climate, which became more and more arid, and finally resulted in desert. These “Dune Dwellers”, none of whose skeletal remains were found, left quantities of artifacts resembling those of the Mousterian culture through Upper Palaeolithic to Neolithic types, some of the characteristic forms being small thumb-nail scrapers, flake knives and drills.

Nelson suggests that many of the types were left by a people who may have migrated to Europe as Azilians. One wonders whether they may not have migrated to America, as well, and whether the correspondence of certain types of artifacts from somewhat similar sand dune areas in our Great Plains region is a mere coincidence or whether further investigation is not warranted.


52 Dixon. 1933.

53 Brooks. 1926.

54 Toyubee. 1934.

55 Huntington. 1914 1915.

56 Smith, P. S., 1933.

57 Capps, 1931.

58 Upham, 1896.

59 Johnston, 1933.

60 Leonard. 1916; Sardeson, 1931.

61 Daly, 1915, p. 174.

62 Cron, 1893, p. 395.

63 Penck, 1894.

64 Antes. 1938; 1931; 1932; 1934.

65 Antevs, 1935.

66 Keyes, 1935.

67 Antevs, DM. pp. 507-508.

68 Antes, 1929.

69 Jochelson, 1928.

70 Sukachev, 1922.

71 See Petri, 1948, p. 38.

72 Mather, 1930, p. 190.

73 Von Eiekstedt, 1934.

74 Taylor, 1934.

75 Jochelson, 1928.

76 Petri, 1916b.

77 Petri, 1916a.

78 Andrews, 1932.

Cite This Article

Howard, Edgar B.. "Migration Routes." The Museum Journal XXIV, no. 2-3 (June, 1935): 123-134. Accessed June 18, 2024. https://www.penn.museum/sites/journal/9562/


This digitized article is presented here as a historical reference and may not reflect the current views of the Penn Museum.

Report problems and issues to digitalmedia@pennmuseum.org.