Popular Science Monthly/Volume 53/May 1898/The West Indian Bridge Between North and South America

1393724Popular Science Monthly Volume 53 May 1898 — The West Indian Bridge Between North and South America1898Joseph William Winthrop Spencer



MUTABILITY OF OCEANS AND CONTINENTS.—A glance at a map of the American continent, inclosing the West Indian seas within its mass, suggests that these basins are sunken plains, submerged to only a moderate extent, but the soundings show depths reaching to more than three miles. "It is not too much to say that every spot which is now dry land has been sea at some former period, and every part of space now covered by the deepest oceans has been land."[1] This enunciation still held place among the latest writings of the great geological teacher—Sir Charles Lyell. As the earlier geologists had not the means of measuring the amount of terrestrial movements, the doctrine of mutability of continents and seas, as taught by Lyell, was doubted by many who later substituted the hypothesis of their permanency from the most remote times, although subjected to ceaseless changes of form. The hypothesis of permanency of continents and seas was largely based upon the littoral character of sedimentary formations, although the evidence of the abysmal or oceanic origin of the widespread chalk deposits could not be easily disposed of. Again, the development and distribution of animal and plant life have been skillfully used as evidence against certain great changes in insular and continental connections, beyond limited proportions. The amount of the concession has varied greatly among the different advocates, so that even under the general hypothesis of permanency, the configuration of the West Indian region has undergone great changes, yet not sufficient to bridge over the seas between the two Americas.

The biological evidence alone, in favor of the permanency of continents and oceans, sometimes suggesting the most startling evidence to the contrary, is insufficient to base theories upon, unless supported by physical indorsement. Indeed, some of the most interesting questions concerning the distribution of biological forms in the West Indian seas can only be explained by the recent discoveries of the physical changes of the region. Some of the smaller oscillations of land and sea have often been measured, but the determination of many of those of stupendous proportions, while occasionally hinted at, was not accomplished until the standard value of the great yardstick was found in the West Indian seas. The discovery of the Antillean bridge between North and South America, which is the theme of this communication, is also important in establishing methods for determining similar great changes of land and ocean in other regions in late geological times.

Growth of the Science of Geomorphy.—The methods of inquiry belong to geomorphy, or the study of land forms. Geomorphy is the outgrowth of topography, which was made a science fifty or sixty years ago by Prof. J. P. Lesley and his coworkers. Its birth is graphically described by the author himself:[2] “That the European Jura … had to wait for its elucidation until the American Appalachians had been mapped may seem strange, but it admits of easy explanation. In Pennsylvania, paleontology and detailed local stratigraphy were impossible; we were untrained in both, … The country was no ground for mineralogy; no rare and curious minerals exist in it; it is a waste of sand, mud, iron, and limestone strata of various textures and color in endless repetition; to know one was to know all; to know it here was to know it everywhere. Nothing remained to study but dynamic forms; and these so numerous, so grand, so variously grouped that they excited our perpetual enthusiasm, and led to infinite research; they supplied the place of fossil forms, of forms of crystals, and of variations of mineral elements; they were a world of the exhibition of the natural forces by itself, and as such we took possession of it and settled in it as our fathers did in the valleys themselves, and thus became not mineralogists, not miners, not learned in fossils, not geologists in the full sense of the term, but topographers; and topography became a science, and was returned to Europe and presented to geology there as an American invention. The passion with which we all studied it is inconceivable, the details into which it led us were infinite. … They trained us to that fertility of fancy which precedes the ripening of the constructive or geometric faculty, made to act upon the more difficult problems of geology; while its generalizations were so vast … that customary European local research … seemed tedious. … The moment, therefore, that one of us beheld the ranges of the Jura, with their combs and offsets, their vast escarpments, and far-glittering white gaps, he was at home among friends, where geologists, born among them, felt that they were strangers. For the valleys of the Jura were filled with later formations full of fossils, which the Appalachian valleys never are. There much is hidden, here all is told. Fossils themselves in the Jura distracted study from the topography.”

Geomorphy, like its antecedent, the science of topography, is an American innovation, and for the same reasons. We have here better opportunities for studying both the newer and older topographic forms, developed over broader geographical domains, and surmounting geological formations commonly of wider extent, so that the evolution of geomorphic features from the simple to the complex form is more easily understood. The explorations of the submerged topography are also along more favorable lines than in the older world. The many repetitions of the middle geological formations in western Europe, and the favorable opportunities they afford for studying certain fossils and minerals, have also distracted from the development of geomorphy there, while in America the investigations of the physical features are now occupying the most prominent place among geologists.

Former Changes of Level in the West Indies.—The Lesser Antilles, or Windward Islands, have often been regarded as a chain of sunken mountains, and the Greater Antilles as having been connected with the mainland. But the dissimilarity of their features, compared with those of our northern continent, and the almost general absence of the higher types of animal life in the islands, caused many to question such connections in late geological times. However, more analytical methods have brought about different conclusions.

The doctrine of the permanency of the ocean basins in the West Indian region received its coup de grâce by the discovery of radiolarian earths upon the highest land of Barbadoes, which was elucidated in the classic treatise of Messrs. A. J. Jukes-Browne and J. B. Harrison.[3] Radiolaria are minute organisms with beautiful silicious shells, which are now forming geological accumulations at only great oceanic depths. Similar deposits also occur in Trinidad, Cuba, Hayti, Jamaica, and probably other islands. They furnish testimony of the upheaval of the floor of the ocean from depths of two miles or more. The age of the radiolarian earths in the West Indies appears to have been about the early Miocene period, which was considerably prior to the building and completion of the West Indian continent. These deposits are most interesting, as they show, upon biological evidence, the uplifting of the bed of the ocean, and later on we shall show that an equally great terrestrial movement occurred in the opposite direction, upon the sinking of the Antillean continent in Pleistocene times.

Modern Changes of Level Measured.—Such great earth movements are slow pulsations extending over large continental areas, but they do not belong to the phenomena of thrusts which give rise to mountain structures, nor are they the effects of volcanic forces. The land rises or sinks so gently that usually these movements do not disturb the courses of the drainage of the rivers, although barriers sometimes do appear across valleys. The modern changes of level illustrate these gentle oscillations. Islands and farms, which were located upon the low coast of New Jersey in the early settlement of the country, are now reduced in size, and are partly converted into salt marshes. The rate of sinking in this locality has been estimated by Mr. Mitchell[4] at only two feet in a century. The depression of land about the mouth of the Mississippi has lately been measured by Mr. E. L. Corthell, who finds that the sinking there is at the rate of five feet in a century.[5] For these low lands this subsidence promises to become a serious economic question in the not distant future. On the other hand, certain northern regions are rising. Thus, in the district of Niagara Falls, the rise is a foot and a quarter in a century, or perhaps a little more. In the St. Lawrence Valley, upon the northwestern flanks of the Adirondack Mountains, the upward movement, for the last fifteen hundred years at least, has been from four to five feet per century. Such gentle changes, accelerated or retarded, and continuing sufficiently long, are capable of transposing the ocean floors and mountain heights.

Recent Elevation of the Eastern Coast of America.—Numerous soundings, chiefly for the use of navigators, but occasionally taken for scientific purposes, have been made off the American coast and in the West Indian seas. In order to insure mariners against the occurrence of sunken rocks, the surveys have often been carried from the coast all the way to oceanic depths. Upon the submarine coastal plains the extensions of some of the great rivers have long been known. From data thus collected, Lindenkohl traced the Hudson River across the submerged banks off the New York and New Jersey coasts to a depth of nearly three thousand feet[6] Later, the writer[7] gathered evidence from the drowned St. Lawrence, Delaware, Susquehanna, and other rivers as far as the Mississippi, and it became apparent that the whole of the eastern coast of America in recent times stood three thousand feet higher than now, with suggestions of a still greater elevation which he then hesitated to follow up on account of their startling character. The existence of the deeper submarine valleys was thus still left unexplained, as if they were either unimportant phenomena or as if they had been formed by causes not in operation at the present time.

Character of Land Valleys.—Ordinarily speaking, valleys are produced by the chemical and physical action of the rains, rills, and rivers denuding-the surfaces of the land, while the character of the rocks gives rise to modified Fig. 1.—Cross-section representing a broad valley, a b r b a; with base level lowered, the valley (b c r c b) was produced. Another rise of the district causes the stream to make a new cañon (d) too recently for the valley to become mature. The terraces at b b and c c show the positions of former base levels. features. Streams flowing through old valleys are usually insignificant, if compared with the size of the valleys. When a stream commences its work, it gradually deepens its channel until it is reduced to the base level of erosion—that is to say, when the slope of the floor of the valley rises so gently from the sea level, or from some other barrier, that it does not permit any further deepening of the river channel. In the first stage, the stream only excavates the narrow gorge or cañon. Upon the base level being reached, the rains and rills widen the valley, while the work of the river itself is chiefly that of carrying away the mud washed off the neighboring lands by the rains, or of undermining an occasional bank. The process continuing, the valley may become dozens of miles in width. With the subsequent rise in the land, a lower base level is formed, and accordingly the process is repeated from the canon-making stage to that of mature valleys, as illustrated in Fig. 1.

With the land rising at intermittent epochs, the valley becomes a series of steps, as illustrated in Fig. 2. Each of the steps is being cut slowly backward, but, owing to the character of the rocks, the

Fig. 2.—Longitudinal section of a valley dissecting a plateau (represented by broken shading along a line a g x), showing three base levels, with the front of the platforms (b c and d e) characterized by waterfalls.

lower platforms may be worn away faster than the upper, so that in time the steps will disappear as the valley becomes mature, when its whole floor is as low as it is possible for the stream to deepen it.

If upon the completion of a portion of a valley the country subsides, the lower reaches may become submerged, as shown in Plate I, when the newly formed bay gradually becomes filled with mud or the washes of the land.

Valleys even in mountain regions are almost independent of the undulations of the strata, in that the courses of the streams frequently

Plate 1.—Valley at Phillipsburg, St. Martin. Lower portion occupied by shallow salt ponds, beneath which more than thirty feet of mud have already been deposited. Beach in front of ponds formed by wave action.

occupy the crests of folds, while the geological troughs are represented by features in bold relief, as is illustrated in Fig. 3. It is easily shown that valleys, even among mountains, whether parallel to the ridges or breaking through them, are the result of the denudation of the rock surfaces of the country, and accordingly, when similar forms occur beneath the sea, they suggest the same origin.

On the Magnitude of Land Valleys.—Concerning the magnitude of land valleys, they vary from the smallest ravines to valleys such as the Mississippi, whose flood plains are from thirty to eighty miles wide for the lower five hundred miles of its length. Several comparatively short rivers, crossing the coastal plains of the southeastern Atlantic States, have valleys from two to four miles wide, even at a hundred miles from the sea, and upon nearing the coast

Fig. 3.—Cross-section of a complex valley in the southern Appalachians, showing it to be independent of the geological structure, which is represented by the shading. Dotted lines mark what was the upper limit of strata which Lave been denuded away. The straight lines (F) illustrate some of the faults affecting the region.

they may be from five to ten miles wide, bounded by only broken hills. All the rivers crossing the coastal plains are flowing over deeply buried channels. That of the Savannah River is buried beneath two hundred and fifty feet or more of superficial deposits, and the old Mississippi Valley is now known to reach one thousand feet below sea level at New Orleans. These buried channels prove that in recent times the continent has sunken to a great extent. The valley of the St. Lawrence River differs from that of the Mississippi in being drowned but not subsequently filled with the mud brought down by the streams. In its lower reaches it is seventy miles wide. These examples of continental valleys are greater than any of the drowned ones to be considered in this paper. The examples cited are those of valleys crossing extensive plains at no great elevation above the sea. with their lower reaches depressed even below high tide.

The Colorado River of the West flows from table-lands eight thousand to ten thousand feet above the sea. Its cañon section is about two hundred and twenty miles long. This is not a simple gorge, but a broad valley from five to twelve miles wide, bounded by walls rising two thousand feet above its floor. This floor was an old base level of erosion formed at no considerable altitude, so that the streams meandered sluggishly over it, and the rains and rills widened it to broad proportions; but, owing to subsequent elevation of the region, the river has cut down its channel to a still lower base level, and in doing so it has been deepened thirty-five hundred or four thousand feet more, with the production of a narrow gorge having precipitous walls, not yet widened into a mature valley. Channels like that of the Colorado also occur crossing submarine plateaus.

Plate II.—Valley of the Rosseau, Dominica. A short, deep valley, heading in an amphitheater and dissecting a denuded plateau.

The short tributaries of these high plateau valleys may be likened to gigantic washouts. Such short, deep valleys commonly form at their heads amphitheaters, which may be a few miles in length, as is illustrated in Plate II. These are sometimes so precipitous that they can not be scaled. The depths vary from a few hundred feet to-a couple of thousand feet, or even more. The same features are found beneath the sea.

On the Declivity of Land Valleys.—The declivity of the greater land valleys crossing regions of continental extent is usually very gentle—a foot per mile, more or less. Smaller valleys may have a declivity of five or ten feet per mile. In the short amphitheaters, descending from the high plateaus, the gradients may be two hundred or even five hundred feet or more per mile. The valleys dissecting recently elevated table-lands, such as those of Mexico, have not uniform slopes, but the descent is characterized by a great series of steps, the surface of each often appearing nearly level to the eye, but with abrupt margins. The vertical heights of such steps vary from five feet to even five hundred feet. The character of the slopes is illustrated in Fig. 4. If stretches of several miles be taken, so

Fig. 4.—Longitudinal section of a Mexican valley (above Atoyac), showing the descent in steps.

many gradation plains occur that the mean declivity of the valley may be from seventy-five to one hundred and fifty feet per mile. Each of these represents pauses in the elevation of the region, while the streams were flowing at levels so low that they could not further deepen their channels, but were widening them out into broad valleys or plains—that is to say, the gradation plains represent base levels of erosion. The valleys on the surface of the table-lands have usually low gradients, which may be as small as those of rivers crossing continental plains. Similar gradation steps are found in valleys beneath the sea.

Submarine Plateaus.—The low coastal plains of the Southeastern States do not terminate at the seashore, but pass beyond, forming shoals and banks, and eventually submarine plateaus, overlooking the edge of the continent, which is fifteen miles eastward of Cape Hatteras, but three hundred miles distant from the coast of Florida. (See map, Plate III.) They extend to and include the Bahamas and other islands. These submarine plateaus have various depths. An elevation of one hundred to three hundred feet would greatly enlarge the Southeastern States, and raise the Bahama banks into broad plains (in reality a continuation of the coastal plains of the Southern States), separated from Cuba and Florida by only narrow channels. A lower broad plateau occurs in this same region, at

Plate III—The soundings are given in feet.

depths from twenty-five hundred to thirty-five hundred feet. From its margin the edge of the continent drops abruptly to depths of twelve thousand feet or more. On the western side of Florida the drowned plains gradually slope down to about three hundred feet, beyond which there is an abrupt descent to the abyss of the Gulf of Mexico. The Yucatan plains extend as broad submerged banks to about three hundred feet beneath the sea, and the sea floor then falls rapidly to that of the Gulf of Mexico, at twelve thousand feet. Such slightly submerged plains of great breadth occur on the banks between Honduras and Jamaica. The Windward Islands are fragments of a plateau which an emergence of less than three thousand feet would unite into one body of land. Both to the east and west of Jamaica there are plateaus depressed to between three thousand and four thousand feet. Still lower plateaus are indicated in the Caribbean Sea, which reaches to a depth of fifteen thousand feet.

From the existence of these submerged plateaus it becomes apparent that a change of elevation of from two thousand to four thousand feet would not merely unite Cuba and Florida, but would greatly enlarge the West Indian lands, and almost connect North and South America. Such a change of elevation would nearly barricade the Antillean water into three basins—the Caribbean Sea, the Sea of Honduras, and the Gulf of Mexico. The Sea of Honduras, between Cuba and Jamaica, reaches to the phenomenal depth of twenty thousand feet, in the form of a long, narrow channel. The structure of these sea basins, at various depths, has been mapped by many writers, but the character of the channels which dissect the drowned plateaus had been almost entirely passed over until the appearance of the writer's previous papers.[8]

The Drowned Valleys or Fiords.—The greater number of the submerged valleys discovered are found to be continuations of buried channels of modern rivers. In some cases the valleys have been so completely submerged that even the divides between those trending in opposite directions have also been drowned—as, for example, the straits of Florida. These drowned valleys have been traced across the submarine plateaus and terminate in enlarged embayments indenting the margin of the continental mass. A few great submarine valleys are found parallel with the mountain ranges. The continuation of many land valleys beneath the sea, and others completely submerged, is shown in the accompanying map (Plate III), and attention may now be called to a few notable examples.

The Savannah Valley, deeply buried on reaching the present sea-shore, crosses the submarine coastal plains at a depth of sixteen hundred and fifty feet lower than the floor of the plateau itself, which is already submerged at that point to nineteen hundred and fifty feet beneath the surface of the sea. The canon becomes still deeper upon nearing the edge of the continental shelf. The Altamaha becomes a canon with a depth of fifty-three hundred feet at the point where the continental shelf is submerged twenty-five hundred feet. Its length is about three hundred miles. It terminates in an embayment thirteen thousand five hundred feet below the surface of the sea. Among the Bahama Islands, and between them and Cuba, there are several similar fiords or drowned valleys, reaching to depths of from two thousand to twelve thousand feet or more. The straits of Florida, and the fiords extending from them, have afforded special opportunities for studying the submerged valleys. The shallowest parts of the straits of Florida are two thousand and sixty-four feet beneath the surface of the sea. From this col, or divide, the Floridian channel extends for a distance of three hundred and fifty miles to a point where it has a depth of ten thousand three hundred and fourteen feet, upon nearing the floor of the Gulf of Mexico. From the same divide the deep Bahaman valley trends in the opposite direction, skirting the northern side of the Bahaman group, and becomes a fiord reaching to a depth of about twelve thousand feet near the edge of the continental shelf. The Abacan channel, crossing the Bahama banks, may be followed to a similar depth. (See map, and Figs. 5 and 6, page 21.)

The valley of the Mississippi (buried to a depth of one thousand feet before reaching the present mouth of the river) is well marked across the submerged plateau to near the floor of the Gulf of Mexico. The drowned valleys of many other Southern rivers are similarly traceable to the floor of the Gulf of Mexico. The same is true of the submarine channels dissecting the drowned plateaus of the Honduras and Caribbean Seas, as shown on the map.

Many of the submerged valleys have tributaries converging from all possible directions, as, for example, those of the Floridian channel. There are also numerous short fiords, tributary to those of greater proportions, but these come from the abrupt margins of the continental shelves or islands, like the amphitheaters indenting the high table-lands.

Of the numerous submarine valleys discovered, a considerable number is shown upon the map, but only a few are cited in the text; by far the larger proportion lie along directions transverse to the trend of the coast lines and mountain ranges, and consequently they

These sections, Fit's. 5 and 6. illustrate gradation plains and steps between the col of the straits of Florida and the floors of the Mexican and Atlantic basins The cañonlike character is shown by the dissection of the continental shelf, whose floor is seen in broken shading. The gradient of the Colorado Canon is also shown in Fig. 5. The depth of submergence is given in feet; and the mean slopes in feet for distances of a given number of miles.

were not depressed by the terrestrial movements which elevated mountain zones or depressed ocean basins. A few of the submarine valleys are parallel to the mountain ridges, and may have been deepened by earth movements, although such is not apparent in the many tributaries. It has been found that the greatest amount of upward warping is in the mountain regions, and the greatest depression is supposed to be along the margins of oceanic abysses. Such movements would have a tendency to somewhat increase terrestrial and submarine declivities; but if transverse upward warping became exaggerated, the valleys would become barricaded so as to form basins, like that of Lake Ontario, or even greater sea basins. Such, however, is not the case with the valleys crossing the submerged coastal plains.

Comparison of Land and Drowned Valleys.—Since the submarine valleys, wherever traceable to the shores, are found to be continuations of the Fig. 7.—Longitudinal section of the Cazonan channel (south of Cuba), showing a similar but shorter fiord dissecting the land mass, the submerged floor of which is shown by the broken shading. Fig. 8.—Longitudinal sections of the Atozac Valley, descending from the Mexican plateau (eight thousand feet above the sea), on the same scale us that of the drowned valleys, but which is here too small to show the numerous steps, except the mean declivity for comparison with that of the abrupt slopes of the submarine steps. Fig. 9.—Longitudinal section of a similar valley south of Monterey in Mexico. buried valleys crossing the coastal plains, they have considerable magnitude where first detected in the soundings. Parenthetically it may be stated that during the late minor oscillations of the continent the old valleys have become filled with sand, etc., for some distance seaward; but beyond this fringe they are always found where the soundings have been taken sufficiently near together. Submerged valleys gradually increase in size until they enter the oceanic embayments indenting the margins of the continental mass. These embayments have been found to vary from perhaps ten to forty miles in width, or within the limit of size shown in the modern Mississippi and St. Lawrence Valleys. Even the upper cañon of the Colorado River has a width of twelve miles. Consequently, the breadth of the drowned valleys crossing the submerged coastal plains is no greater than that illustrated by those of the land.

The resemblance between the terrestrial and submarine valleys is even more striking in the case of their tributaries, which come from various directions. Where there are short tributaries descending from submerged plateaus, they have also the form of amphitheaters like those dissecting the margins of table-lands.

The slopes of the submarine channels consist of a series of steps, like the gradation plains descending from high table-lands. These steps represent long pauses in the elevation of the plateaus. For comparison of the declivity of the drowned valleys with those of the land, sections are given on pages 21 and 22 in Figs. 5 to 9, all drawn to the same scale. The declivities of some of these gradation plains do not exceed a foot per mile—that is, these slopes are as low as those of continental valleys which are reduced to the base level of erosion. While much is yet to be learned of the detailed features between the submarine steps, we already know that their mean declivity is less abrupt than that of land valleys descending from the Mexican plateaus. The submerged steps appear to have the same origin as those on the border of table-lands—that is to say, they were formed during the pauses in the terrestrial oscillations when the now drowned continental plateaus formed table-lands.

The deep channels crossing the submerged plateaus for a distance of two hundred or three hundred miles, with a depth of from two thousand to six thousand feet—and among the Bahama banks to even greater proportions—show a close resemblance to the Colorado Valley and Cañon.

The Geological Yardstick and West Indian Bridge.—From the apparently complete analogy between the characteristics of the land and submarine valleys or channels—namely, (1) the submarine valleys being continuations of those of the continent or islands; (2) both having tributaries converging from every possible direction; (3) both classes of valleys having their magnitude of corresponding proportions, with (4) similar great canons and amphitheaterlike tributaries; (5) both terrestrial and submarine channels with similar gradation plains and steps characterizing their slopes; (6) without obstructing barriers—the conclusion is reached that the depths of the submarine channels may be used as yardsticks to show that the land lately stood nearly as high as the valleys are deep.

Applied to the West Indian region and adjacent parts of the continent, it would thus appear that these regions stood from ten thousand to twelve thousand feet, or in some localities fourteen thousand feet, higher than now. The West Indian bridge reached a height of from two to more than two and a half miles, while the doors of the Gulf of Mexico and the Caribbean Sea, to which depth the drowned valleys can be traced, were continental plains, like those of the Mississippi or Amazon of the present day, with perhaps some shallow lakes or small seas. The backbone of this West Indian bridge was near the Atlantic side, and remains of it are seen in the ridge dissected during the epoch of high elevation by the rapidly descending streams, which now constitute the chain of the Windward Islands.

The plains now forming the floors of the Gulf of Mexico and the Honduras and Caribbean Seas were apparently drained into the Pacific Ocean across the Tehuantepec Isthmus, Honduras, Nicaragua, Panama (see map, page 000), and other low depressions farther south. The writer′s recent explorations in the Tehuantepec Isthmus confirm this hypothesis. Plateaus of Mexico and Central America, rising to from six thousand to ten thousand feet, are there reduced in height for a distance of more than sixty miles, so that we find half a dozen passes as low as eight hundred feet above the sea. During the earlier period of the elevation of the Antillean region the Tehuantepec Isthmus was a strait in which a deep-water fauna was living. Later the strait was transformed into land troughs, fragments of the old base-level floor of which are still extant, and through them narrow geological canals were formed when the low Mexican plains had again sunk beneath the gulf waters. This question of the elevation of the Mexican and Central American barriers would carry us beyond the limits of the present paper, so that all that can now be said of them is that they were elevated at a very recent date, corresponding to the subsidence of the West Indian region, to heights reaching from six thousand to even more than ten thousand feet. This elevation was a sort of compensation in the terrestrial balance.

The Date of the Continental Bridge.—Over the West Indian region there are many widely distributed geological formations which were acumulated[9] in early Miocene times. During the following late Miocene and Pliocene periods these formations were lifted above the sea and became lands of great extent. This elevated condition continued so long that the country became enormously denuded, and was reduced to valleys and low base-level plains, some of which now appear to constitute the gradation plateaus beneath the sea. On the sinking of the land, after the Mio-Pliocene period of elevation, new deposits (the Lafayette) were accumulated in the valleys, the age of which is provisionally placed at the end of the Pliocene, although some might regard it as early Pleistocene, for there is no sharp line of demarcation characterizing the limits of these formations. But these deposits immediately underlie those of the Glacial period of the north. Subsequently the land of the West Indian region rose, and this time to its greatest height, so as to allow the excavation of the very deep valleys and cañons through the then existing table-lands which now form the submarine plateaus. This high elevation was characterized by such deep sculpturing as to give rise to some of the boldest physical features of the lands with which we are familiar. With the subsequent depression of the West Indian bridge, the fragmentary islands became much smaller than even now, and upper portions of the drowned valleys were again partly filled with a still newer formation (known as the Columbia of the Southern States). This last has since been elevated a few hundred feet, with some minor changes of level continuing to the present day. The Columbia formation belongs to a mid-Pleistocene epoch. Consequently, the time of greatest elevation and development of the West Indian continent was during the early Pleistocene period—more popularly called the Ice age. These great and recurring changes of level of land and sea, in later geological times, have produced remarkable and startling revolutions in the physical geography of the region, which might seem incredible but for the great array of evidence which is now being accumulated on every hand.

Relationship between the Distribution of Life and the West Indian Continent.—The elevation of the Antillean bridge is a question of dynamical geology, but the consequent changes of the physical geography naturally affected the distribution of life; accordingly, the biological aspect should be called upon as evidence of the physical changes.

The commingling of the littoral fauna of the Pacific Ocean with that of the Antillean waters confirms the recent separation of the two seas by the elevation of the Central American region.

The character of the deep-sea fauna is very important. The elevation of the West Indies to less than three thousand feet would exclude the Atlantic waters from the Antillean seas, except through two or three shallow straits, and one channel of enormous depth, between the Virgin Islands and St. Croix. According to the late Dr. Brown Goode, the deep-sea fishes belong to modern forms, apparently overflowing through the Caribbean Sea into the Gulf of Mexico. There is no relationship whatever between the deep-sea fishes of these basins and those of the adjacent waters of the Pacific Ocean. The seemingly recent introduction of much the larger proportion of deep-sea fishes from the Atlantic naturally indicates the existence of barriers between the different basins—in other words, an elevation of the region by a few thousand feet. Furthermore, the general modern character of the fish fauna, with the exclusion of the Pacific forms, suggests the inundation of the floors (including possibly a few small sea basins) of the Caribbean and Honduras Seas and the Gulf of Mexico in times so recent as to correspond with the great sinking of the Antillean continent.

The land shells of the Bahamas, Puerto Rico, Haiti, Cuba, and Jamaica are generally related to each other, and are more or less connected with those of Yucatan and Florida, as was long ago pointed out by Mr. Thomas. Bland, and more fully studied by Mr. Charles T. Simpson, who has shown that the land shells are poorly represented in the Windward Islands, but that they are closely allied to those of South America. The larger islands have been lately submerged so as to leave only their higher districts above the sea, but from these small remnants the land snails could be developed with local variations upon the re-elevation and enlargement of the islands. On this basis the connection of the islands is affirmed, and the remaining question is one of time, as we do not know in how short a period specific variations may be effected.

The occurrence of mammals is of more importance than that of other animals, as they are less likely to be distributed by adventitious circumstances than other groups. In the West Indies there is a remarkable poverty of mammalian life. In Cuba and Haiti, six living species of Capromys or Hutia occur (three in each island). These are small rodents, similar to a Pleistocene type found in Brazilian caves. One species of insectivore, Selenodon, occurs in Cuba, and another in Haiti; these are of a Madagascar type, as well as an iguana, a reptile found in several of the islands. The little agouti occurs in the Island of Trinidad. These few forms represent almost wholly the indigenous mammals of the West Indies. The monkeys of St. Kitts belong to the Old World, and appear to have been artificially introduced. The scarcity of higher life has given rise to the supposition that the islands have been separated from the mainland continuously, since a period before the appearance of modern mammals. But this generalization was made without considering the physical history of the islands.

In the Southern States, including Florida, there was an assemblage of many animals, such as the rhinoceros, mastodon, etc., which lived in the late Miocene or early Pliocene period (according to Professors Cope and Scott). This fauna became entirely exterminated without leaving any immediate successors. Nor have any been found in the West Indies. This, however, is not strange, when we consider how few localities are known on the continent where such remains have been discovered; and, further, how the broad lands of the Mio-Pliocene times were subsequently inundated by the sea at the close of the Pliocene period, thus reducing the region to a number of very small islands.

Belonging to the Pleistocene period, in the Southern States, the remains of an extensive mammalian fauna have been found, including animals of the tapir, horse, elephant, bison, deer, and other families. To this period also belong the giant sloths of the Megalonyx type. In the West Indies, the North American sloth (Megalonyx) and an extinct Hutia have been found in Cuba. Three species of rodents as large as the Virginia deer were obtained in Anguilla by Mr. Wager Ray, and determined by Professor Cope. The writer lately saw in Guadeloupe, in the possession of Mr. L. Guesde, the tooth of a small elephant, which had been discovered in the island. A tooth of the late Florida elephant has also been found in the Bahamas (Lucas). While these animals named are few in number, yet they show land connection between what are now the islands and the continent in the Pleistocene period. The subsidence in mid-Pleistocene days submerged not merely the continental bridge between North and South America, but caused all the lower land of the islands to be drowned, so that there was an almost complete extinction of mammalian life. Indeed, the tombs of this Pleistocene fauna are now largely beneath the sea.

Further biological testimony of the continental bridge is found among the extensive remains of mammals discovered at Port Kennedy, near Philadelphia, upon which Prof. E. D. Cope was engaged at the time of his recent death. These fossils belong to the old Pleistocene fauna, separated by submergence (the Columbia) from the more modern remains. Among these old Pleistocene species there occur South American types, most notably abundant remains of bears, that are not found among the fossils of the Western or Central States, but which appear to have migrated by way of the West Indian bridge.

Although a few of the higher animals might survive the general extinction caused by the submergence of the Antillean continent by escaping into the higher lands of the few remaining islands, yet these migrations—as, for example, from savannas to mountain forests—would tend to complete their extermination. Besides the restriction of areas, the changes from elevated temperate to low tropical climates would further bring about the destruction of many animals.

Finally, it may be said that the distribution of animal life requires the Pleistocene connection between North and South America; and consequently the biological and physical evidence, so far as known, coincides in bearing joint testimony of the late West Indian bridge.

Conclusion.—The late West Indian continent is a geological feature which belongs to the period which just precedes the modern, and it was the breaking down and sinking of the bridge which brought the physical geography of that region into its present form. While the presence of man upon the West Indian and eastern American plateaus is not even suggested by remains of him, yet at that time there roamed over the savannas and through the forests several animals identical with and many closely related to the existing species; and this is as near as we may hope to connect the sunken lands with human associations.

The application of the deep-sea channels to the measurements of the subsidence of the land is a recent innovation. The elevation of mountain regions to great heights has long since passed beyond the bounds of doubt, but their forms have only recently been used for interpreting their history. The reason for this lies in the fact that many of the necessary topographical and geological explorations have not been made until recent years, and because geologists were chiefly engaged upon problems of greater antiquity than modern land features, especially on fossil remains which told something of the history of the sea basins; so that their interest in these questions caused them to largely overlook the history of the land.

If we look at the Alpine plateau surmounted by the sublime Matterhorn, we may be told by geologists that, somewhere in the Alps, the Tertiary formations have been found at heights of ten or twelve thousand feet; in the Himalaya Mountains similar marine strata occur to an elevation of twenty thousand feet. We are still left in ignorance of the geographical history of the region, in spite of the geological youthfulness of the formations. Since the comparatively recent Tertiary period much of the molding of the physical features has been effected. If we observe the Matterhorn, as shown in Plate IV, any one may see a gently sloping plain dissected by a valley thousands of feet deep, which has almost passed the cañon stage, but is yet very immature. Now the geomorphist or scientific geographer will almost at a glance interpret the story recorded in these forms. The plateau is the remains of a base level of erosion, and bears testimony of its formation at a level of nine thousand or ten thousand feet lower, or near sea level, and the marine fossils found in the district will tell since what geological epoch the low plains were formed which have since become elevated into the high plateau. The Matterhorn itself, in spite of the faults, is largely a remnant of a higher mountain mass, which was worn down to a base-level floor by the insidious action of the rains and rills. The valley bears record of an elevation so recent, in spite of the hardness of the rocks and the slowness of degradation, that it has not yet entirely passed its youthful stage. The actually slow excavation of such a great valley, since the last geological period preceding the modern, certainly impresses us with the enormous duration of geological time. It is such magnificent land forms as these plateaus and great valleys, found submerged beneath the sea, which have furnished us with the means for measuring the terrestrial

Plate IV.—View of the Matterhorn, Plateau and Valley. From a painting by Mr. Charles J. Way.

movements in the opposite direction which have reduced the West Indian region to a group of islands.

As to the causes of these terrestrial movements, at present little can be said. Changes of level of land and sea, similar to those found in the West Indian region, have commonly occurred over perhaps most parts of the earth's surface, only the evidence has not been so fully collected. Such oscillations have greatly affected the migrations of animal and plant life, and have produced changes of climate. These physical changes are in themselves sufficient in a great measure to account for the glacial phenomena of the Pleistocene period.

  1. The last edition of Lyell's Principles of Geology.
  2. Manual of Coal and its Topography. By J. P. Lesley. 12mo, pp. 1-224. Published by J. P. Lippincott, Philadelphia, 1856.
  3. The Geology of Barbadoes. Quarterly Journal of the Geological Society of London, vol. xlviii, pp. 170-226.
  4. Of the United States Coast Survey.
  5. Geographical Development of the Lower Mississippi. Read before the Toronto meeting of the British Association.
  6. Appendix XIII, Report of the United States Coast Survey for 1887 (1889), pp. 270-273.
  7. High Continental Elevation preceding the Pleistocene Period. Bulletin of the Geological Society of America, vol. i, 1889, p. 65.
  8. Terrestrial Submergence Southeast of the American Continent. Bulletin of the Geological Society of America, vol. v, pp. 19-22, 1893. Reconstruction of the Antillean Continent. Ibid., vol. vi, pp. 103-140, 1894.
  9. accumulated?