Popular Science Monthly/Volume 44/December 1893/How Old Is the Earth


WITHIN the memory of men now living, and especially during the last thirty years, the processes of the creation of the earth and its inhabitants, of the solar system, and of the starry heavens, have come to be understood in a very different way from that in which they were thought of by our fathers and forefathers. Instead of the former belief that divine fiats at successive times suddenly spoke into existence the forms of animal and plant life now occupying the earth, the earlier faunas and floras found fossil in the rocks, and at still earlier dates the earth itself, the sun, and the entire astronomic universe, it is now recognized and confidently accepted on all sides that all animals and plants, the globe which we inhabit, and the sun and stars, have been created through slow processes of development, which are well denominated evolution—that is, an unrolling or unfolding. These changes have been in progress during unnumbered and inconceivably long ages; they are still going forward; and they will probably continue as far into the unfathomable future as they have come to us through the dimly and in part somewhat clearly discerned past. To us who are borne upon its bosom this current seems like the lower Amazon, too broad for us to see its banks, coming from the high Andes and the lower plains, and going to its rest in the ocean.

According to the well-approved nebular hypothesis of Kant and Laplace, the material of our earth and moon became separated from the condensing mass of the sun after the outer planets had been similarly produced, but before the birth of Venus and Mercury. At early stages in the condensation of the revolving nebula, it had thrown off successively the portions of matter which were afterward gathered, by their independent condensation and revolution, to form Neptune, Uranus, Saturn, Jupiter, and Mars; while another portion, which was probably never united like the other planets, made the many small asteroids. The matter which has been changed into our splendidly luminous sun was at one time very attenuated and occupied the whole space inclosed by the orbit of the outermost planet, which was developed from a comparatively very small cloud or ring of this matter, centrifugally detached from the revolving exceedingly tenuous mass. In like manner the material of each of the planets, including the earth, was shed from the whirling nebula at times during its decrease in volume when its circumference was approximately coincident with their orbits.

Again, in their turn the planetary masses have undergone the same evolutionary process, taking a rotary motion and throwing off, as they condensed, the material which now circles about them in the shape of moons and rings. In the case of our own planet and its single large satellite, probably the far greater part of the original cloud or ring whence they were produced had assumed a somewhat globular or discoid form and taken a movement of revolution which still continues as the earth's daily rotation, before the moon's mass was separated from that of the earth. It seems to me, however, very improbable that the present contour of our globe should preserve, as suggested by Fisher, the scars of this loss in the depressions of the deep ocean basins.

Many relatively small portions of the ring of matter producing the earth and moon may have become early separated from the chief condensing mass, and after its division in our globe and its satellite have been drawn by gravitation into them, marring the face of the moon, as Gilbert supposes, with its multitudes of both small and very large crateriform scars. On the earth, too, if this hypothesis be true, such falling asteroid-like bodies must also have made similar small and huge blots by their violent impact; but they evidently were effaced by the slow processes of atmospheric and stream erosion, or in basin areas were deeply covered by sediments, before the formation of the oldest of our fossiliferous strata. The absence of either air or water on the moon has allowed the steep and jagged mountain rims of the deeply depressed lunar craters to remain nearly unchanged from the almost inconceivably remote time, according to this view, when the asteroid bombardment of the moon was completed.

Geology, or the science of the earth's changes and development, deals with the rocks forming the crust of our globe. From their stratigraphic sequence and the successive fossil faunas and floras found in them, the geologist gathers the history of the sedimentation or volcanic eruption of the rocks and the concurrent changes of animal and plant life. Moreover, by reasoning from the physical condition and structure of the rock formations, from volcanic action, earthquakes, and the upheaval of continents and mountain ranges, he conjectures what may be the condition of the deep interior of the earth, through its observed influence upon the crust. Both these phases of the science have yielded estimates of the age of the earth, the former based on the geologic processes of erosion and deposition, the latter on the earth's loss of internal heat, the magnitude and the effects of the oceanic tides, and other conditions whose investigation belongs more specially to the physicist and astronomer. Each takes up the question for the later part of the earth's planetary existence, after it was so far condensed and cooled that it had already attained approximately its present size and was enveloped by a crust which, through many stages of diverse changes, has continued to the present day.

The estimates derived from these two directions of inquiry, however, differ considerably among themselves, and especially it is noticeable that in general the physical and astronomical investigations of the question yield smaller estimates than those drawn from stratigraphic and paleontologic data. Sir William Thomson (now Lord Kelvin) long ago estimated, from his study of the earth's internal heat, its increase from the surface downward, and the rate of its loss by radiation into space, that the time since the consolidation of the surface of the globe has been somewhere between twenty million and four hundred million years, and that most probably this time and all the geologic record must be limited within one hundred million years. Mr. Clarence King, from very careful experiments on the volcanic rock diabase, supposed to represent the average constitution of the earth's crust, when subjected to extremes of heat and pressure, applying his results in the same way as Lord Kelvin, has within the present year published his conclusion that the earth's duration measures only about twenty-four million years. Prof. George H. Darwin computes, from the influence of tidal friction in retarding the earth's rotation, that probably only fifty-seven million years have elapsed since the moon's mass was shed from the revolving molten earth, long before the formation of its crust. From the same arguments and the rate at which the sun is losing its store of heat, Prof. Guthrie Tait affirms that apparently ten million years are as much as physical science can allow to the geologist. Prof. Newcomb writes: "If the sun had, in the beginning, filled all space, the amount of heat generated by his contraction to his present volume would have been sufficient to last eighteen million years at his present rate of radiation. . . . Ten million years. . . is, therefore, near the extreme limit of time that we can suppose water to have existed on the earth in the fluid state." Not only the earth but even the whole solar system, according to Newcomb, "must have had a beginning within a certain number of years which we can not yet calculate with certainty, but which can not much exceed twenty million, and it must end."

The geologist demurs against these latter far too meager allotments of time for the wonderful, diversified, and surely vastly long history which he has patiently made out in his perusal of the volume of science disclosed by the rocks. He can apparently do very well with Lord Kelvin's original estimate, but must respectfully dissent from the less liberal opinions noted. Somewhere in the assumed premises which yield to mathematicians these narrow limits of time, there must be conditions which do not accord with the actual constitution of the sun and earth. It must be gratefully acknowledged, however, in the camp of the geologists, that we owe to these researches a beneficial check against the notion once prevalent that geologic time extends back practically without limit; and it is most becoming for us carefully to inquire how closely the apparently conflicting testimonies of geology and physics may be brought into harmony by revision of each.

Among all the means afforded by geology for direct estimates of the earth's duration, doubtless the most reliable is through comparing the present measured rate of denudation of continental areas with the aggregate of the greatest determined thickness of the strata referable to the successive time divisions. The factors of this method of estimate, however, are in considerable part uncertain, or dependent on the varying opinions of different geologists. According to Sir Archibald Geikie, in his presidential address a year ago before the British Association, the time thus required for the formation of all the stratified rocks of the earth's crust may range from a minimum of seventy-three million up to a maximum of six hundred and eighty million years. Prof. Samuel Haughton obtains in this way, "for the whole duration of geological time a minimum of two hundred million years." On the other hand, smaller results are reached through the same method by Dana, who conjectures that the earth's age may be about forty-eight million years since the formation of the oldest fossiliferous rocks; and by Alfred Russel Wallace, who concludes that this time has probably been only about twenty-eight million years. With these, rather than with the foregoing, we may also place Mr. T. Mellard Reade's recent estimate of ninety-five million years, similarly derived. Again, Mr. C. D. Walcott, in his vice-presidential address before Section E of the American Association for the Advancement of Science, in its meeting last August, gave his opinion, from a study of the sedimentary rocks of the western Cordilleran area of the United States, that the duration of time since the Archæan era has been probably some forty-five million years.

Selecting the lowest of these various estimates as the nearest in accord with the conclusions of physical and astronomical science, let us scrutinize the processes of Wallace's measurements and computations. It has been found that the rates at which rivers are lowering the altitudes of their basins by the transportation of sediments to the sea vary from an average of one foot taken from the land surface of its hydrographic basin by the river Po in seven hundred and thirty years, to one foot by the Danube in six thousand eight hundred years. As a mean for all the rivers of the world, Wallace assumes that the erosion from all the land surface is one foot in three thousand years. The sediments are laid down in the sea on an average within thirty miles from the coast, and all the coast lines of the earth have a total measured length, according to Dr. James Croll and Mr. Wallace, of about one hundred thousand miles, so that the deposition is almost wholly confined to an area of about three million square miles. This area is one nineteenth as large as the earth's total land area; hence it will receive sediment nineteen times as fast as the land is denuded, or at the rate of about nineteen feet of stratified beds in three thousand years, which would give one foot in one hundred and fifty-eight years. With this Wallace compares the total maxima of all the sedimentary rocks of the series of geologic epochs, measured in whatever part of the earth they are found to have their greatest development. Prof. Haughton estimates their aggregate to be one hundred and seventy-seven thousand two hundred feet, which, multiplied by one hundred and fifty-eight, gives approximately twenty-eight million years as the time required for the deposition of the rock strata in the various districts where they are thickest and have most fully escaped erosion and redeposition.

Most readers, following this argument, would infer that it must give too large rather than too scanty an estimate of geologic duration; but to many students of the earth's stratigraphy it seems more probably deficient than excessive. All must confess that the argument rests upon many indeterminate premises, since the total extent of the land areas and the depths of the oceans have probably been increasing through the geologic eras, and the effects of tides have probably diminished. The imperfection of the geologic record, so impressively shown by Charles Darwin in respect to the sequence of plants and animals found fossil in the rocks, will also be appealed to as opposing the assumption that the one hundred and seventy-seven thousand two hundred feet, or thirty-three and a half miles, of strata represent the whole, or indeed any more than a small fraction, of the earth's history. To myself, however, this last objection seems unfounded, since in many extensive and clearly conformable sections observed on a grand scale in crossing broad areas, there is seen to have been evidently continuous deposition during several or many successive geologic epochs; and by combining such sections from different regions a record of sedimentation is made well-nigh complete from the earliest Palæozoic morning of life to its present high noon. But perhaps we may do better to change somewhat the premises of our computation, in view of the extensive regions where the rock strata remain yet to be thoroughly explored, and because of certain large inland tracts having little rain and therefore no drainage into the sea. If we assume that the total maxima of strata amount to fifty miles, and that the mean rate of the land denudation is only one foot in six thousand years, we then obtain a result three times greater than before, or about eighty-four million years for the deposition of the stratified rocks.

Another method of considering this problem is afforded by the determination of one term in a sequence of ratios, whereby the sum of the whole becomes known. Though geologists differ widely in their estimates of the earth's age, up to the seven thousand million years claimed by McGee, in an address last year before the American Association, they are approximately in agreement as to the ratios of the several great divisions of geologic time. From the thicknesses of the strata and the changes in the animal and plant life, it is comparatively easy to determine the relative lengths of the successive eras, while yet it is very difficult to decide beyond doubt even the approximate length in years of any part of the record. The portions for which we have the best means of determining their lengths are the Glacial and recent periods, the latter extending from the Champlain epoch, or closing stage of the ice age to the present time, while these two divisions, the Glacial or Pleistocene period and the recent, make up the Quaternary era. If we can only ascertain somewhat nearly what has been the duration of this era, from the oncoming of the Ice age until now, it will serve as a known quantity to be used as the multiplier for giving us the approximate or probable measures in years for the recedingly earlier and far longer Tertiary, Mesozoic or Secondary, Palæozoic or Primary, and Archæan or Beginning eras, which last takes us back almost or quite to the time when the cooling molten earth became first enveloped with a solid crust.

Haughton has estimated time ratios from two series of data. His results deduced from the maximum thickness of the strata for the three grand divisions of Archæan, Palæozoic, and subsequent time, expressed in percentages, are 34·3:42·5:23·2; and from his computations as to the secular cooling of the earth, 33·0:41·0:26·0. The ratios reached by Profs. J. D. Dana and Alexander Winchell from the thicknesses of the rock strata are closely harmonious, the durations of Palæozoic, Mesozoic, and Cenozoic time being to each other as 12:3:1. The Tertiary and Quaternary ages, the latter extending to the present day, which are here united as the Cenozoic era, Dana would rank approximately in the ratio of 3:1, giving to the Quaternary a sixty-fourth part of all time since the beginning of the Cambrian period, to which our earliest well-preserved fossil faunas belong. For reasons to be stated later, I think that this estimate of the relative length of Quaternary time is greatly exaggerated; but this would not sensibly affect the general ratios.

Prof. W. M. Davis, of Harvard University, without speaking definitely of the lapse of time by years, has endeavored to give some conception of what these and like estimates of geologic ratios really mean, through a translation of them into terms of a linear scale. Starting with the representation of the postglacial or recent period, since the North American ice-sheet was melted away, as two inches, he estimates that the beginning of the Tertiary erosion of the Hudson River gorge through the Highlands would be expressed by a distance of ten feet; that the Triassic reptilian tracks in the sandstone of the Connecticut Valley would be probably fifty feet distant; that the formation of the coal beds of Pennsylvania would be eighty or one hundred feet back from the present time; and that the Middle Cambrian trilobites of Braintree, Mass, would be two hundred, three hundred, or four hundred feet from us.

Having such somewhat definite and agreeing ratios, derived from various data by different investigators, can we secure the factor by which they should be multiplied to yield the approximate duration of geologic epochs, periods, and eras, in years? If on the scale used by Prof. Davis we could substitute a certain time for the period since the departure of the ice-sheet, we should thereby at once determine, albeit with some vagueness and acknowledged latitude for probable error, how much time has passed since the Triassic tracks were made, the coal deposited, and the trilobites entombed in the Cambrian slates. Now, just this latest and present division of the geologic record, following the Ice age, is the only one for which geologists find sufficient data to permit direct measurements or estimates of its duration. "The glacial invasion from which New England and other northern countries have lately escaped," remarks Davis, "was prehistoric, and yet it should not be regarded as ancient."

In various localities we are able to measure the present rate of erosion of gorges below waterfalls, and the length of the postglacial gorge divided by the rate of recession of the falls gives approximately the time since the Ice age. Such measurements of the gorge and Falls of St. Anthony by Prof. N. H. Winchell show the length of the postglacial or recent period to have been about eight thousand years; and from the surveys of Niagara Falls Mr. G. K. Gilbert believes it to have been seven thousand years, more or less. From the rates of wave-cutting along the sides of Lake Michigan and the consequent accumulation of sand around the south end of the lake, Dr. E. Andrews estimates that the land there became uncovered from its ice-sheet not more than seventy-five hundred years ago. Prof. G. Frederick Wright obtains a similar result from the rate of filling of kettle-holes among the gravel knolls and ridges called kames and eskers, and likewise from the erosion of valleys by streams tributary to Lake Erie; and Prof. Benjamin K. Emerson, from the rate of deposition of modified drift in the Connecticut Valley at Northampton, Mass., thinks that the time since the Glacial period can not exceed ten thousand years. An equally small estimate is also indicated by the studies of Gilbert and Russell for the time since the last great rise of the Quaternary lakes Bonneville and Lahontan, lying in Utah and Nevada, within the arid Great Basin of interior drainage, which are believed to have been contemporaneous with the great extension of ice-sheets upon the northern part of our continent.

Prof. James Geikie maintains that the use of palæolithic implements had ceased, and that early man in Europe made neolithic (polished) implements, before the recession of the ice-sheet from Scotland, Denmark, and the Scandinavian peninsula; and Prestwich suggests that the dawn of civilization in Egypt, China, and India may have been coeval with the glaciation of northwestern Europe. In Wales and Yorkshire the amount of denudation of limestone rocks on which bowlders lie has been regarded by Mr. D. Mackintosh as proof that a period of not more than six thousand years has elapsed since the bowlders were left in their positions. The vertical extent of this denudation, averaging about six inches, is nearly the same with that observed in the southwest part of the province of Quebec by Sir William Logan and Dr. Robert Bell, where veins of quartz marked with glacial stride stand out to various heights not exceeding one foot above the weathered surface of the inclosing limestone.

From this wide range of concurrent but independent testimonies we may accept it as practically demonstrated that the icesheets disappeared from North America and Europe some six to ten thousand years ago. But having thus found the value of one term in our ratios of geologic time divisions, we may know them all approximately by its substitution. The two inches assumed to represent the postglacial portion of the Quaternary era may be called eight thousand years; then, according to the proportional estimates by Davis, the Triassic period was probably two million four hundred thousand years ago; the time since the Carboniferous period, in the closing part of the Palæozoic era, has been about four or five million years; and since the middle of the Cambrian period, twice or perhaps four times as long. Continuing this series still further back, the earliest Cambrian fossils may be twenty or twenty-five million years old, and the beginning of life on our earth was not improbably twice as long ago.

Seeking to substitute our measure of postglacial time in Dana's ratios, we are met by the difficulty of ascertaining first its proportion to the preceding Glacial period, and then the ratio which these two together bear to the Tertiary era. It would fill a very large volume to rehearse all the diverse opinions current among glacialists concerning the history of the Ice age, its wonderful climatic vicissitudes, and the upward and downward movements of the lands which are covered with the glacial drift. Many eminent glacialists, as James Geikie, Wahnschaffe, Penck, De Geer, Chamberlin, Salisbury, Shaler, McGee, and others, believe that the Ice age was complex, having two, three, or more epochs of glaciation, divided by long interglacial epochs of mild and temperate climate when the ice-sheets were entirely or mainly melted away. Prof. Geikie claims five distinct glacial epochs, as indicated by fossiliferous beds lying between deposits of till or unstratified glacial drift, and by other evidences of great climatic changes. In this country Mr. McGee recognizes at least three glacial epochs. The astronomic theory of Croll attributes the accumulation of ice-sheets to recurrent cycles which bring the winters of each polar hemisphere of the earth alternately into aphelion and perihelion each twenty-one thousand years during the periods of maximum eccentricity of the earth's orbit. Its last period of this kind was from about two hundred and forty thousand to eighty thousand years ago, allowing room for seven or eight such cycles and alternations of glacial and interglacial conditions. The supposed evidence of interglacial epochs therefore gave to this theory a wide credence; but the recent determinations of the geologic brevity of the time since the ice-sheets disappeared from North America and Europe make it clear, in the opinions even of some of the geologists who believe in a duality or plurality of Quaternary Glacial epochs, that not astronomic but geographic causes produced the Ice age.

Glacialists who reject Croll's ingenious and brilliant theory mostly appeal to great preglacial altitude of the land as the chief cause of the ice accumulation, citing as proof of such altitude the fiords and submarine valleys which on the shores of Scandinavia, and the Atlantic, Arctic, and Pacific coasts of North America, descend from one thousand to three thousand and even four thousand feet below the sea level, testifying of former uplifts of these continental areas so much above their present heights. But beneath the enormous weights of their ice-sheets these lands sank, so that when the ice attained its maximum area and thickness and during its departure the areas on which it lay were depressed somewhat lower than now and have since been re-elevated. This view to account for the observed records of the Glacial period is held by Dana, Le Conte, Wright, Jamieson, and others, including the present writer. It is believed to be consistent either with the doctrine of two or more glacial epochs during the Quaternary era, or with the reference of all the glacial drift to a single glacial epoch, which is thought by Wright, Prestwich, Lamplugh, Falsan, Hoist, Nikitin, and others to be more probable. To myself, though formerly accepting two glacial epochs with a long warm interval between them, the essential continuity of the Ice age seems now the better provisional hypothesis, to be held with candor for weighing evidence on either side.

The duration of the Ice age, if there was only one epoch of glaciation, with moderate temporary retreats and readvances of the ice border sufficient to allow stratified beds with the remains of animals and plants to be intercalated between accumulations of till, may only have comprised a few tens of thousands of years. On this point Prof. Prestwich has well written as follows: "For the reasons before given, I think it possible that the Glacial epoch—that is to say, the epoch of extreme cold—may not have lasted longer than from fifteen thousand to twenty-five thousand years, and I would for the same reasons limit the time of. . . the melting away of the ice-sheet to from eight thousand to ten thousand years or less."

From these and foregoing estimates, which seem to me acceptable, we have the probable length of Glacial and postglacial time together thirty thousand or forty thousand years, more or less; but an equal or considerably longer preceding time, while the areas that became covered by ice were being uplifted to great altitudes, may perhaps with good reason be also included in the Quaternary era, which then would comprise some one hundred thousand years. Comparing the Tertiary era with the Quaternary, however, I can not agree with Prof. Dana's estimate that the latter was a third as long as the former, and am quite at a loss to discern evidences justifying that view. The best means for learning their ratio I think to be found in the changes of faunas and floras since the beginning of the Tertiary era, using especially the marine molluscan faunas as most valuable for this comparison. Scarcely any species of marine mollusks have become extinct or undergone important changes during the Glacial and recent periods; but since the Eocene dawn of the Tertiary nearly all of these species have come into existence. Judged upon this basis, the Tertiary era seems probably fifty or a hundred times longer than the Ice age and subsequent time; in other words, it may well have lasted two million or even four million years. Taking the mean of these numbers, or three million years, for Cenozoic time, or the Quaternary and Tertiary ages together, we have precisely the value of Prof. Dana's ratios which he himself assumes for conjectural illustration, namely, forty-eight million years since the Cambrian period began. But the diversified types of animal life in the earliest Cambrian faunas surely imply a long antecedent time for their development, on the assumption that the Creator worked before then as during the subsequent ages in the evolution of all living creatures. According to these ratios, therefore, the time needed for the deposition of the earth's stratified rocks and the unfolding of its plant and animal life must be about a hundred million years.

Reviewing the several results independently reached through the geologic estimates and ratios supplied by Wallace, Dana, and Davis, we are much impressed and convinced of their approximate truth by their somewhat good agreement among themselves, which seems as close as the nature of the problem would lead us to expect, and by their all coming within the limit of one hundred million years which Sir William Thomson estimated on physical grounds. This limit of probable geologic duration seems therefore fully worthy to take the place of the once almost unlimited assumptions of geologists and writers on the evolution of life, that the time at their disposal has been practically infinite. No other more important conclusion in the natural sciences, directly and indirectly modifying our conceptions in a thousand ways, has been reached during this century.