TRANSCRIBER'S NOTE: Spelling maintained as closely as possible to the original document, while obvious typos have been corrected. Emdashes in original text for negative temperatures changed to minus signs to standardize temperatures.

CLIMATIC CHANGES

THEIR NATURE AND CAUSES

PUBLISHED ON THE FOUNDATION
ESTABLISHED IN MEMORY OF
THEODORE L. GLASGOW

OTHER BOOKS BY THE SAME AUTHORS

ELLSWORTH HUNTINGTON

1. Four books showing the development of knowledge as to Historical Pulsations of Climate.
   1. The Pulse of Asia. Boston, 1907.
   2. Explorations in Turkestan. Expedition of 1903. Washington, 1905.
   3. Palestine and Its Transformation. Boston, 1911.
   4. The Climatic Factor, as Illustrated in Arid America. Washington, 1914.
2. Two books illustrating the effect of climate on man.
   1. Civilization and Climate. New Haven, 1915.
   2. World Power and Evolution. New Haven, 1919.
3. Four books illustrating the general principles of Geography.
   1. Asia: A Geography Reader. Chicago, 1912.
   2. The Red Man's Continent. New Haven, 1919.
   3. Principles of Human Geography (with S. W. Cushing). New York, 1920.
   4. Business Geography (with F. E. Williams). New York, 1922.
4. A companion to the present volume.
   1. Earth and Sun: An Hypothesis of Weather and Sunspots. New Haven. In press.

STEPHEN SARGENT VISHER

Geography, Geology and Biology of Southern Dakota. Vermilion, 1912.
The Biology of Northwestern South Dakota. Vermilion, 1914.
The Geography of South Dakota. Vermilion, 1918.
Handbook of the Geology of Indiana (with others). Indianapolis, 1922.
Hurricanes of Australia and the South Pacific. Melbourne, 1922.

CLIMATIC CHANGES

THEIR NATURE AND CAUSES
BY
ELLSWORTH HUNTINGTON

Research Associate in Geography in Yale University

AND
STEPHEN SARGENT VISHER

Associate Professor of Geology
in Indiana University

NEW HAVEN
YALE UNIVERSITY PRESS
LONDON: HUMPHREY MILFORD: OXFORD UNIVERSITY PRESS
MDCCCCXXII

COPYRIGHT 1922 BY

YALE UNIVERSITY PRESS

Published 1922.

THE THEODORE L. GLASGOW MEMORIAL

PUBLICATION FUND

The present volume is the fifth work published by the Yale University Press on the Theodore L. Glasgow Memorial Publication Fund. This foundation was established September 17, 1918, by an anonymous gift to Yale University in memory of Flight Sub-Lieutenant Theodore L. Glasgow, R.N. He was born in Montreal, Canada, and was educated at the University of Toronto Schools and at the Royal Military College, Kingston. In August, 1916, he entered the Royal Naval Air Service and in July, 1917, went to France with the Tenth Squadron attached to the Twenty-second Wing of the Royal Flying Corps. A month later, August 19, 1917, he was killed in action on the Ypres front.

TO
THOMAS CHROWDER CHAMBERLIN
OF THE UNIVERSITY OF CHICAGO
WHOSE CLEAR AND MASTERLY DISCUSSION OF THE GREAT PROBLEMS OF TERRESTRIAL EVOLUTION HAS BEEN ONE OF THE MOST INSPIRING FACTORS IN THE WRITING OF THIS BOOK

There is a toy, which I have heard, and I would not have it given over, but waited upon a little. They say it is observed in the Low Countries (I know not in what part), that every five and thirty years the same kind and suit of years and weathers comes about again; as great frosts, great wet, great droughts, warm winters, summers with little heat, and the like, and they call it the prime; it is a thing I do the rather mention, because, computing backwards, I have found some concurrence.

FRANCIS BACON

PREFACE

Unity is perhaps the keynote of modern science. This means unity in time, for the present is but the outgrowth of the past, and the future of the present. It means unity of process, for there seems to be no sharp dividing line between organic and inorganic, physical and mental, mental and spiritual. And the unity of modern science means also a growing tendency toward coöperation, so that by working together scientists discover much that would else have remained hid.

This book illustrates the modern trend toward unity in all of these ways. First, it is a companion volume to Earth and Sun. That volume is a discussion of the causes of weather, but a consideration of the weather of the present almost inevitably leads to a study of the climate of the past. Hence the two books were written originally as one, and were only separated from considerations of convenience. Second, the unity of nature is so great that when a subject such as climatic changes is considered, it is almost impossible to avoid other subjects, such as the movements of the earth's crust. Hence this book not only discusses climatic changes, but considers the causes of earthquakes and attempts to show how climatic changes may be related to great geological revolutions in the form, location, and altitude of the lands. Thus the book has a direct bearing on all the main physical factors which have molded the evolution of organic life, including man.

In the third place, this volume illustrates the unity of modern science because it is preëminently a coöperative product. Not only have the two authors shared in its production, but several of the Yale Faculty have also coöperated. From the geological standpoint, Professor Charles Schuchert has read the entire manuscript in its final form as well as parts at various stages. He has helped not only by criticisms, suggestions, and facts, but by paragraphs ready for the printer. In the same way in the domain of physics, Professor Leigh Page has repeatedly taken time to assist, and either in writing or by word of mouth has contributed many pages. In astronomy, the same cordial coöperation has come with equal readiness from Professor Frank Schlesinger. Professors Schuchert, Schlesinger, and Page have contributed so materially that they are almost co-authors of the volume. In mathematics, Professor Ernest W. Brown has been similarly helpful, having read and criticised the entire book. In certain chemical problems, Professor Harry W. Foote has been our main reliance. The advice and suggestions of these men have frequently prevented errors, and have again and again started new and profitable lines of thought. If we have made mistakes, it has been because we have not profited sufficiently by their coöperation. If the main hypothesis of this book proves sound, it is largely because it has been built up in constant consultation with men who look at the problem from different points of vision. Our appreciation of their generous and unstinted coöperation is much deeper than would appear from this brief paragraph.

Outside the Yale Faculty we have received equally cordial assistance. Professor T. C. Chamberlin of the University of Chicago, to whom, with his permission, we take great pleasure in dedicating this volume, has read the

entire proof and has made many helpful suggestions. We cannot speak too warmly of our appreciation not only of this, but of the way his work has served for years as an inspiration in the preliminary work of gathering data for this volume. Professor Harlow Shapley of Harvard University has contributed materially to the chapter on the sun and its journey through space; Professor Andrew E. Douglass of the University of Arizona has put at our disposal some of his unpublished results; Professors S. B. Woodworth and Reginald A. Daly, and Mr. Robert W. Sayles of Harvard, and Professor Henry F. Reid of Johns Hopkins have suggested new facts and sources of information; Professor E. R. Cumings of Indiana University has critically read the entire proof; conversations with Professor John P. Buwalda of the University of California while he was teaching at Yale make him another real contributor; and Mr. Wayland Williams has contributed the interesting quotation from Bacon on page x of this book. Miss Edith S. Russell has taken great pains in preparing the manuscript and in suggesting many changes that make for clearness. Many others have also helped, but it is impossible to make due acknowledgment because such contributions have become so thoroughly a part of the mental background of the book that their source is no longer distinct in the minds of the authors.

The division of labor between the two authors has not followed any set rules. Both have had a hand in all parts of the book. The main draft of Chapters VII, VIII, IX, XI, and XIII was written by the junior author; his contributions are also especially numerous in Chapters X and XV; the rest of the book was written originally by the senior author.

[CHAPTER I]

THE UNIFORMITY OF CLIMATE

The rôle of climate in the life of today suggests its importance in the past and in the future. No human being can escape from the fact that his food, clothing, shelter, recreation, occupation, health, and energy are all profoundly influenced by his climatic surroundings. A change of season brings in its train some alteration in practically every phase of human activity. Animals are influenced by climate even more than man, for they have not developed artificial means of protecting themselves. Even so hardy a creature as the dog becomes notably different with a change of climate. The thick-haired "husky" of the Eskimos has outwardly little in common with the small and almost hairless canines that grovel under foot in Mexico. Plants are even more sensitive than animals and men. Scarcely a single species can flourish permanently in regions which differ more than 20°C. in average yearly temperature, and for most the limit of successful growth is 10°.^[1] So far as we yet know every living species of plant and animal, including man, thrives best under definite and limited conditions of temperature, humidity, and sunshine, and of the composition and movement of the atmosphere or water in which it lives. Any departure beyond the limits means lessened efficiency, and in the long run a lower rate of

reproduction and a tendency toward changes in specific characteristics. Any great departure means suffering or death for the individual and destruction for the species.

Since climate has so profound an influence on life today, it has presumably been equally potent at other times. Therefore few scientific questions are more important than how and why the earth's climate has varied in the past, and what changes it is likely to undergo in the future. This book sets forth what appear to be the chief reasons for climatic variations during historic and geologic times. It assumes that causes which can now be observed in operation, as explained in a companion volume entitled Earth and Sun, and in such books as Humphreys' Physics of the Air, should be carefully studied before less obvious causes are appealed to. It also assumes that these same causes will continue to operate, and are the basis of all valid predictions as to the weather or climate of the future.

In our analysis of climatic variations, we may well begin by inquiring how the earth's climate has varied during geological history. Such an inquiry discloses three great tendencies, which to the superficial view seem contradictory. All, however, have a similar effect in providing conditions under which organic evolution is able to make progress. The first tendency is toward uniformity, a uniformity so pronounced and of such vast duration as to stagger the imagination. Superposed upon this there seems to be a tendency toward complexity. During the greater part of geological history the earth's climate appears to have been relatively monotonous, both from place to place and from season to season; but since the Miocene the rule has been diversity and complexity, a condition highly favorable to organic evolution. Finally, the uniformity of the vast eons of the past and the

tendency toward complexity are broken by pulsatory changes, first in one direction and then in another. To our limited human vision some of the changes, such as glacial periods, seem to be waves of enormous proportions, but compared with the possibilities of the universe they are merely as the ripples made by a summer zephyr.

The uniformity of the earth's climate throughout the vast stretches of geological time can best be realized by comparing the range of temperature on the earth during that period with the possible range as shown in the entire solar system. As may be seen in Table 1, the geological record opens with the Archeozoic era, or "Age of Unicellular Life," as it is sometimes called, for the preceding cosmic time has left no record that can yet be read. Practically no geologists now believe that the beginning of the Archeozoic was less than one hundred million years ago; and since the discovery of the peculiar properties of radium many of the best students do not hesitate to say a billion or a billion and a half.^[2] Even in the Archeozoic the rocks testify to a climate seemingly not greatly different from that of the average of geologic time. The earth's surface was then apparently cool enough so that it was covered with oceans and warm enough so that the water teemed with microscopic life. The air must have been charged with water vapor and with carbon dioxide, for otherwise there seems to be no possible way of explaining the formation of mudstones and sandstones, limestones of vast thickness, carbonaceous shales, graphites, and iron ores.^[3] Although the Archeozoic has yielded no generally admitted fossils, yet what seem to be massive algæ and sponges have been

found in Canada. On the other hand, abundant life is believed to have been present in the oceans, for by no other known means would it be possible to take from the air the vast quantities of carbon that now form carbonaceous shales and graphite.

In the next geologic era, the Proterozoic, the researches of Walcott have shown that besides the marine algæ there must have been many other kinds of life. The Proterozoic fossils thus far discovered include not only microscopic radiolarians such as still form the red ooze of the deepest ocean floors, but the much more significant tubes of annelids or worms. The presence of the annelids, which are relatively high in the scale of organization, is generally taken to mean that more lowly forms of animals such as coelenterates and probably even the mollusca and primitive arthropods must already have been evolved. That there were many kinds of marine invertebrates living in the later Proterozoic is indicated by the highly varied life and more especially the trilobites found in the oldest Cambrian strata of the next succeeding period. In fact the Cambrian has sponges, primitive corals, a great variety of brachiopods, the beginnings of gastropods, a wonderful array of trilobites, and other lowly forms of arthropods. Since, under the postulate of evolution, the life of that time forms an unbroken sequence with that of the present, and since many of the early forms differ only in minor details from those of today, we infer that the climate then was not very different from that of today. The same line of reasoning leads to the conclusion that even in the middle of the Proterozoic, when multicellular marine animals must already have been common, the climate of the earth had already for an enormous period been such that all the lower types of oceanic invertebrates had already evolved.

Moreover, they could live in most latitudes, for the indirect evidences of life in the Archeozoic and Proterozoic rocks are widely distributed. Thus it appears that at an almost incredibly early period, perhaps many hundred million years ago, the earth's climate differed only a little from that of the present.

The extreme limits of temperature beyond which the climate of geological times cannot have departed can be approximately determined. Today the warmest parts of the ocean have an average temperature of about 30°C. on the surface. Only a few forms of life live where the average temperature is much higher than this. In deserts, to be sure, some highly organized plants and animals can for a short time endure a temperature as high as 75°C. (167°F.). In certain hot springs, some of the lowest unicellular plant forms exist in water which is only a little below the boiling point. More complex forms, however, such as sponges, worms, and all the higher plants and animals, seem to be unable to live either in water or air where the temperature averages above 45°C. (113°F.) for any great length of time and it is doubtful whether they can thrive permanently even at that temperature. The obvious unity of life for hundreds of millions of years and its presence at all times in middle latitudes so far as we can tell seem to indicate that since the beginning of marine life the temperature of the oceans cannot have averaged much above 50°C. even in the warmest portions. This is putting the limit too high rather than too low, but even so the warmest parts of the earth can scarcely have averaged much more than 20° warmer than at present.

Turning to the other extreme, we may inquire how much colder than now the earth's surface may have been since life first appeared. Proterozoic fossils have been

found in places where the present average temperature approaches 0°C. If those places should be colder than now by 30°C., or more, the drop in temperature at the equator would almost certainly be still greater, and the seas everywhere would be permanently frozen. Thus life would be impossible. Since the contrasts between summer and winter, and between the poles and the equator seem generally to have been less in the past than at present, the range through which the mean temperature of the earth as a whole could vary without utterly destroying life was apparently less than would now be the case.

These considerations make it fairly certain that for at least several hundred million years the average temperature of the earth's surface has never varied more than perhaps 30°C. above or below the present level. Even this range of 60°C. (108°F.) may be double or triple the range that has actually occurred. That the temperature has not passed beyond certain narrow limits, whatever their exact degree, is clear from the fact that if it had done so, all the higher forms of life would have been destroyed. Certain of the lowest unicellular forms might indeed have persisted, for when dormant they can stand great extremes of dry heat and of cold for a long time. Even so, evolution would have had to begin almost anew. The supposition that such a thing has happened is untenable, for there is no hint of any complete break in the record of life during geological times,—no sudden disappearance of the higher organisms followed by a long period with no signs of life other than indirect evidence such as occurs in the Archeozoic.

A change of 60°C. or even of 20° in the average temperature of the earth's surface may seem large when viewed from the limited standpoint of terrestrial experience.

Viewed, however, from the standpoint of cosmic evolution, or even of the solar system, it seems a mere trifle. Consider the possibilities. The temperature of empty space is the absolute zero, or -273°C. To this temperature all matter must fall, provided it exists long enough and is not appreciably heated by collisions or by radiation. At the other extreme lies the temperature of the stars. As stars go, our sun is only moderately hot, but the temperature of its surface is calculated to be nearly 7000°C., while thousands of miles in the interior it may rise to 20,000° or 100,000° or some other equally unknowable and incomprehensible figure. Between the limits of the absolute zero on the one hand, and the interior of a sun or star on the other, there is almost every conceivable possibility of temperature. Today the earth's surface averages not far from 14°C., or 287° above the absolute zero. Toward the interior, the temperature in mines and deep wells rises about 1°C. for every 100 meters. At this rate it would be over 500°C. at a depth of ten miles, and over 5000° at 100 miles.

Let us confine ourselves to surface temperatures, which are all that concern us in discussing climate. It has been calculated by Poynting^[5] that if a small sphere absorbed and re-radiated all the heat that fell upon it, its temperature at the distance of Mercury from the sun would average about 210°C.; at the distance of Venus, 85°; the earth 27°; Mars -30°; Neptune 219°. A planet much nearer the sun than is Mercury might be heated to a temperature of a thousand, or even several thousand, degrees, while one beyond Neptune would remain almost at absolute zero. It is well within the range of possibility that the temperature of a planet's surface should be

anywhere from near -273°C. up to perhaps 5000°C. or more, although the probability of low temperature is much greater than of high. Thus throughout the whole vast range of possibilities extending to perhaps 10,000°, the earth claims only 60° at most, or less than 1 per cent. This may be remarkable, but what is far more remarkable is that the earth's range of 60° includes what seem to be the two most critical of all possible temperatures, namely, the freezing point of water, 0°C., and the temperature where water can dissolve an amount of carbon dioxide equal to its own volume. The most remarkable fact of all is that the earth has preserved its temperature within these narrow limits for a hundred million years, or perchance a thousand million.

To appreciate the extraordinary significance of this last fact, it is necessary to realize how extremely critical are the temperatures from about 0° to 40°C., and how difficult it is to find any good reason for a relatively uniform temperature through hundreds of millions of years. Since the dawn of geological time the earth's temperature has apparently always included the range from about the freezing point of water up to about the point where protoplasm begins to disintegrate. Henderson, in The Fitness of the Environment, rightly says that water is "the most familiar and the most important of all things." In many respects water and carbon dioxide form the most unique pair of substances in the whole realm of chemistry. Water has a greater tendency than any other known substance to remain within certain narrowly defined limits of temperature. Not only does it have a high specific heat, so that much heat is needed to raise its temperature, but on freezing it gives up more heat than any substance except ammonia, while none of the common liquids approach it in the amount of additional

heat required for conversion into vapor after the temperature of vaporization has been reached. Again, water substance, as the physicists call all forms of H₂O, is unique in that it not only contracts on melting, but continues to contract until a temperature several degrees above its melting point is reached. That fact has a vast importance in helping to keep the earth's surface at a uniform temperature. If water were like most liquids, the bottoms of all the oceans and even the entire body of water in most cases would be permanently frozen.

Again, as a solvent there is literally nothing to compare with water. As Henderson^[6] puts it: "Nearly the whole science of chemistry has been built up around water and aqueous solution." One of the most significant evidences of this is the variety of elements whose presence can be detected in sea water. According to Henderson they include hydrogen, oxygen, nitrogen, carbon, chlorine, sodium, magnesium, sulphur, phosphorus, which are easily detected; and also arsenic, cæsium, gold, lithium, rubidium, barium, lead, boron, fluorine, iron, iodine, bromine, potassium, cobalt, copper, manganese, nickel, silver, silicon, zinc, aluminium, calcium, and strontium. Yet in spite of its marvelous power of solution, water is chemically rather inert and relatively stable. It dissolves all these elements and thousands of their compounds, but still remains water and can easily be separated and purified. Another unique property of water is its power of ionizing dissolved substances, a property which makes it possible to produce electric currents in batteries. This leads to an almost infinite array of electro-chemical reactions which play an almost dominant rôle in the processes of life. Finally, no common liquid except mercury equals water in its power

of capillarity. This fact is of enormous moment in biology, most obviously in respect to the soil.

Although carbon dioxide is far less familiar than water, it is almost as important. "These two simple substances," says Henderson, "are the common source of every one of the complicated substances which are produced by living beings, and they are the common end products of the wearing away of all the constituents of protoplasm, and of the destruction of those materials which yield energy to the body." One of the remarkable physical properties of carbon dioxide is its degree of solubility in water. This quality varies enormously in different substances. For example, at ordinary pressures and temperatures, water can absorb only about 5 per cent of its own volume of oxygen, while it can take up about 1300 times its own volume of ammonia. Now for carbon dioxide, unlike most gases, the volume that can be absorbed by water is nearly the same as the volume of the water. The volumes vary, however, according to temperature, being absolutely the same at a temperature of about 15°C. or 59°F., which is close to the ideal temperature for man's physical health and practically the same as the mean temperature of the earth's surface when all seasons are averaged together. "Hence, when water is in contact with air, and equilibrium has been established, the amount of free carbonic acid in a given volume of water is almost exactly equal to the amount in the adjacent air. Unlike oxygen, hydrogen, and nitrogen, carbonic acid enters water freely; unlike sulphurous oxide and ammonia, it escapes freely from water. Thus the waters can never wash carbonic acid completely out of the air, nor can the air keep it from the waters. It is the one substance which thus, in considerable quantities relative to its total amount, everywhere accompanies

water. In earth, air, fire, and water alike these two substances are always associated.

"Accordingly, if water be the first primary constituent of the environment, carbonic acid is inevitably the second,—because of its solubility possessing an equal mobility with water, because of the reservoir of the atmosphere never to be depleted by chemical action in the oceans, lakes, and streams. In truth, so close is the association between these two substances that it is scarcely correct logically to separate them at all; together they make up the real environment and they never part company."^[7]

The complementary qualities of carbon dioxide and water are of supreme importance because these two are the only known substances which are able to form a vast series of complex compounds with highly varying chemical formulæ. No other known compounds can give off or take on atoms without being resolved back into their elements. No others can thus change their form freely without losing their identity. This power of change without destruction is the fundamental chemical characteristic of life, for life demands complexity, change, and growth.

In order that water and carbon dioxide may combine to form the compounds on which life is based, the water must be in the liquid form, it must be able to dissolve carbon dioxide freely, and the temperature must not be high enough to break up the highly complex and delicate compounds as soon as they are formed. In other words, the temperature must be above freezing, while it must not rise higher than some rather indefinite point between 50°C. and the boiling point, where all water finally turns into vapor. In the whole range of temperature, so far as

we know, there is no other interval where any such complex reactions take place. The temperature of the earth for hundreds of millions of years has remained firmly fixed within these limits.

The astonishing quality of the earth's uniformity of temperature becomes still more apparent when we consider the origin of the sun's heat. What that origin is still remains a question of dispute. The old ideas of a burning sun, or of one that is simply losing an original supply of heat derived from some accident, such as collision with another body, were long ago abandoned. The impact of a constant supply of meteors affords an almost equally unsatisfactory explanation. Moulton^[8] states that if the sun were struck by enough meteorites to keep up its heat, the earth would almost certainly be struck by enough so that it would receive about half of 1 per cent as much heat from them as from the sun. This is millions of times more heat than is now received from meteors. If the sun owes its heat to the impact of larger bodies at longer intervals, the geological record should show a series of interruptions far more drastic than is actually the case.

It has also been supposed that the sun owes its heat to contraction. If a gaseous body contracts it becomes warmer. Finally, however, it must become so dense that its rate of contraction diminishes and the process ceases. Under the sun's present condition of size and density a radial contraction of 120 feet per year would be enough to supply all the energy now radiated by that body. This seems like a hopeful source of energy, but Kelvin calculated that twenty million years ago it was ineffective and ten million years hence it will be equally so. Moreover, if this is the source of heat, the amount of radiation

from the sun would have to vary enormously. Twenty million years ago the sun would have extended nearly to the earth's orbit and would have been so tenuous that it would have emitted no more heat than some of the nebulæ in space. Some millions of years later, when the sun's radius was twice as great as at present, that body would have emitted only one-fourth as much heat as now, which would mean that on the earth's surface the theoretical temperature would have been 200° below the present level. This is utterly out of accord with the uniformity of climate shown by the geological record. In the future, if the sun's contraction is the only source of heat, the sun can supply the present amount for only ten million years, which would mean a change utterly unlike anything of which the geological record holds even the faintest hint.^[9]

Altogether the problem of how the sun can have remained so uniform and how the earth's atmosphere and other conditions can also have remained so uniform throughout hundreds of millions of years is one of the most puzzling in the whole realm of nature. If appeal is taken to radioactivity and the breaking up of uranium into radium and helium, conditions can be postulated which will give the required amount of energy. Such is also the case if it be supposed that there is some unknown process which may induce an atomic change like radioactivity in bodies which are now supposed to be stable elements. In either case, however, there is as yet no satisfactory explanation of the uniformity of the earth's climate. A hundred million or a thousand million years ago the temperature of the earth's surface was very much the same as now. The earth had then presumably ceased to emit any great amount of heat, if we may judge

from the fact that its surface was cool enough so that great ice sheets could accumulate on low lands within 40° of the equator. The atmosphere was apparently almost like that of today, and was almost certainly not different enough to make up for any great divergence of the sun from its present condition. We cannot escape the stupendous fact that in those remote times the sun must have been essentially the same as now, or else that some utterly unknown factor is at work.

[CHAPTER II]

THE VARIABILITY OF CLIMATE

The variability of the earth's climate is almost as extraordinary as its uniformity. This variability is made up partly of a long, slow tendency in one direction and partly of innumerable cycles of every conceivable duration from days, or even hours, up to millions of years. Perhaps the easiest way to grasp the full complexity of the matter is to put the chief types of climatic sequence in the form of a table.

In assigning names to the various types an attempt has been made to indicate something of the nature of the sequence so far as duration, periodicity, and general tendencies are concerned. Not even the rich English language of the twentieth century, however, furnishes words with enough shades of meaning to express all that

is desired. Moreover, except in degree, there is no sharp distinction between some of the related types, such as glacial fluctuations and historic pulsations. Yet, taken as a whole, the table brings out the great contrast between two absolutely diverse extremes. At the one end lies well-nigh eternal uniformity, or an extremely slow progress in one direction throughout countless ages; at the other, rapid and regular vibrations from day to day, or else irregular and seemingly unsystematic vacillations due to cyclonic storms, both of which types are repeated millions of times during even a single glacial fluctuation.

The meaning of cosmic uniformity has been explained in the preceding chapter. Its relation to the other types of climatic sequences seems to be that it sets sharply defined limits beyond which no changes of any kind have ever gone since life, as we know it, first began. Secular progression, on the other hand, means that in spite of all manner of variations, now this way and then the other, the normal climate of the earth, if there is such a thing, has on the whole probably changed a little, perhaps becoming more complex. After each period of continental uplift and glaciation—for such are preëminently the times of complexity—it is doubtful whether the earth has ever returned to quite its former degree of monotony. Today the earth has swung away from the great diversity of the glacial period. Yet we still have contrasts of what seem to us great magnitude. In low depressions, such as Turfan in the central deserts of Eurasia, the thermometer sometimes ranges from 0°F. in the morning to 60° in the shade at noon. On a cloudy day in the Amazon forest close to the seashore, on the contrary, the temperature for months may rise to 85° by day and sink no lower than 75° at night.

The reasons for the secular progression of the earth's

climate appear to be intimately connected with those which have caused the next, and, in many respects, more important type of climatic sequence, which consists of geological oscillations. Both the progression and the oscillations seem to depend largely on three purely terrestrial factors: first, the condition of the earth's interior, including both internal heat and contraction; second, the salinity and movement of the ocean; and third, the composition and amount of the atmosphere. To begin with the earth's interior—its loss of heat appears to be an almost negligible factor in explaining either secular progression or geologic oscillation. According to both the nebular and the planetesimal hypotheses, the earth's crust appears to be colder now than it was hundreds or thousands of millions of years ago. The emission of internal heat, however, had probably ceased to be of much climatic significance near the beginning of the geological record, for in southern Canada glaciation occurred very early in the Proterozoic era. On the other hand, the contraction of the earth has produced remarkable effects throughout the whole of geological time. It has lessened the earth's circumference by a thousand miles or more, as appears from the way in which the rocks have been folded and thrust bodily over one another. According to the laws of dynamics this must have increased the speed of the earth's rotation, thus shortening the day, and also having the more important effect of increasing the bulge at the equator. On the other hand, recent investigations indicate that tidal retardation has probably diminished the earth's rate of rotation more than seemed probable a few years ago, thus lengthening the day and diminishing the bulge at the equator. Thus two opposing forces have been at work, one causing acceleration and one retardation. Their combined

effect may have been a factor in causing secular progression of climate. It almost certainly was of much importance in causing pronounced oscillations first one way and then the other. This matter, together with most of those touched in these first chapters, will be expanded in later parts of the book. On the whole the tendency appears to have been to create climatic diversity in place of uniformity.

The increasing salinity of the oceans may have been another factor in producing secular progression, although of slight importance in respect to oscillations. While the oceans were still growing in volume, it is generally assumed that they must have been almost fresh for a vast period, although Chamberlin thinks that the change in salinity has been much less than is usually supposed. So far as the early oceans were fresher than those of today, their deep-sea circulation must have been less hampered than now by the heavy saline water which is produced by evaporation in warm regions. Although this saline water is warm, its weight causes it to descend, instead of moving poleward in a surface current; this descent slows up the rise of the cold water which has moved along in the depths of the ocean from high latitudes, and thus checks the general oceanic circulation. If the ancient oceans were fresher and hence had a freer circulation than now, a more rapid interchange of polar and equatorial water presumably tended to equalize the climate of all latitudes.

Again, although the earth's atmosphere has probably changed far less during geological times than was formerly supposed, its composition has doubtless varied. The total volume of nitrogen has probably increased, for that gas is so inert that when it once becomes a part of the air it is almost sure to stay there. On the other hand,

the proportions of oxygen, carbon dioxide, and water vapor must have fluctuated. Oxygen is taken out constantly by animals and by all the processes of rock weathering, but on the other hand the supply is increased when plants break up new carbon dioxide derived from volcanoes. As for the carbon dioxide, it appears probable that in spite of the increased supply furnished by volcanoes the great amounts of carbon which have gradually been locked up in coal and limestone have appreciably depleted the atmosphere. Water vapor also may be less abundant now than in the past, for the presence of carbon dioxide raises the temperature a little and thereby enables the air to hold more moisture. When the area of the oceans has diminished, and this has recurred very often, this likewise would tend to reduce the water vapor. Moreover, even a very slight diminution in the amount of heat given off by the earth, or a decrease in evaporation because of higher salinity in the oceans, would tend in the same direction. Now carbon dioxide and water vapor both have a strong blanketing effect whereby heat is prevented from leaving the earth. Therefore, the probable reduction in the carbon dioxide and water vapor of the earth's atmosphere has apparently tended to reduce the climatic monotony and create diversity and complexity. Hence, in spite of many reversals, the general tendency of changes, not only in the earth's interior and in the oceans, but also in the atmosphere, appears to be a secular progression from a relatively monotonous climate in which the evolution of higher organic forms would scarcely be rapid to an extremely diverse and complex climate highly favorable to progressive evolution. The importance of these purely terrestrial agencies must not be lost sight of when we come to discuss other agencies outside the earth.

In Table 2 the next type of climatic sequence is geologic oscillation. This means slow swings that last millions of years. At one extreme of such an oscillation the climate all over the world is relatively monotonous; it returns, as it were, toward the primeval conditions at the beginning of the secular progression. At such times magnolias, sequoias, figs, tree ferns, and many other types of subtropical plants grew far north in places like Greenland, as is well known from their fossil remains of middle Cenozoic time, for example. At these same times, and also at many others before such high types of plants had evolved, reef-making corals throve in great abundance in seas which covered what is now Wisconsin, Michigan, Ontario, and other equally cool regions. Today these regions have an average temperature of only about 70°F. in the warmest month, and average well below freezing in winter. No reef-making corals can now live where the temperature averages below 68°F. The resemblance of the ancient corals to those of today makes it highly probable that they were equally sensitive to low temperature. Thus, in the mild portions of a geologic oscillation the climate seems to have been so equable and uniform that many plants and animals could live 1500 and at other times even 4000 miles farther from the equator than now.

At such times the lands in middle and high latitudes were low and small, and the oceans extended widely over the continental platforms. Thus unhampered ocean currents had an opportunity to carry the heat of low latitudes far toward the poles. Under such conditions, especially if the conception of the great subequatorial continent of Gondwana land is correct, the trade winds and the westerlies must have been stronger and steadier than now. This would not only enable the westerlies,

which are really southwesterlies, to carry more heat than now to high latitudes, but would still further strengthen the ocean currents. At the same time, the air presumably contained an abundance of water vapor derived from the broad oceans, and an abundance of atmospheric carbon dioxide inherited from a preceding time when volcanoes contributed much carbon dioxide to the air. These two constituents of the atmosphere may have exercised a pronounced blanketing effect whereby the heat of the earth with its long wave lengths was kept in, although the energy of the sun with its shorter wave lengths was not markedly kept out. Thus everything may have combined to produce mild conditions in high latitudes, and to diminish the contrast between equator and pole, and between summer and winter.

Such conditions perhaps carry in themselves the seeds of decay. At any rate while the lands lie quiet during a period of mild climate great strains must accumulate in the crust because of the earth's contraction and tidal retardation. At the same time the great abundance of plants upon the lowlying plains with their mild climates, and the marine creatures upon the broad continental platforms, deplete the atmospheric carbon dioxide. Part of this is locked up as coal and part as limestone derived from marine plants as well as animals. Then something happens so that the strains and stresses of the crust are released. The sea floors sink; the continents become relatively high and large; mountain ranges are formed; and the former plains and emergent portions of the continental platforms are eroded into hills and valleys. The large size of the continents tends to create deserts and other types of climatic diversity; the presence of mountain ranges checks the free flow of winds and also creates diversity; the ocean currents are likewise

checked, altered, and diverted so that the flow of heat from low to high latitudes is diminished. At the same time evaporation from the ocean diminishes so that a decrease in water vapor combines with the previous depletion of carbon dioxide to reduce the blanketing effect of the atmosphere. Thus upon periods of mild monotony there supervene periods of complexity, diversity, and severity. Turn to Table 1 and see how a glacial climate again and again succeeds a time when relative mildness prevailed almost everywhere. Or examine Fig. 1 and notice how the lines representing temperatures go up and down. In the figure Schuchert makes it clear that when the lands have been large and mountain-making has been important, as shown by the high parts of the lower shaded area, the climate has been severe, as shown by the descent of the snow line, the upper shaded area. In the diagram the climatic oscillations appear short, but this is merely because they have been crowded together, especially in the left hand or early part. There an inch in length may represent a hundred million years. Even at the right-hand end an inch is equivalent to several million years.

The severe part of a climatic oscillation, as well as the mild part, will be shown in later chapters to bear in itself certain probable seeds of decay. While the lands are being uplifted, volcanic activity is likely to be vigorous and to add carbon dioxide to the air. Later, as the mountains are worn down by the many agencies of water, wind, ice, and chemical decay, although much carbon dioxide is locked up by the carbonation of the rocks, the carbon locked up in the coal is set free and increases the carbon dioxide of the air. At the same time the continents settle slowly downward, for the earth's crust though rigid as steel is nevertheless slightly viscous and will flow if subjected to sufficiently great and enduring pressure.

The area from which evaporation can take place is thereby increased because of the spread of the oceans over the continents, and water vapor joins with the carbon dioxide to blanket the earth and thus tends to keep it uniformly warm. Moreover, the diminution of the lands frees the ocean currents from restraint and permits them to flow more freely from low latitudes to high. Thus in the course of millions of years there is a return toward monotony. Ultimately, however, new stresses accumulate in the earth's crust, and the way is prepared for another great oscillation. Perhaps the setting free of the stresses takes place simply because the strain at last becomes irresistible. It is also possible, as we shall see, that an external agency sometimes adds to the strain and thereby determines the time at which a new oscillation shall begin.

In Table 2 the types of climatic sequences which follow "geologic oscillations" are "glacial fluctuations," "orbital precessions" and "historical pulsations." Glacial fluctuations and historical pulsations appear to be of the same type, except as to severity and duration, and hence may be considered together. They will be treated briefly here because the theories as to their causes are outlined in the next two chapters. Oddly enough, although the historic pulsations lie much closer to us than do the glacial fluctuations, they were not discovered until two or three generations later, and are still much less known. The most important feature of both sequences is the swing from a glacial to an inter-glacial epoch or from the arsis or accentuated part of an historical pulsation to the thesis or unaccented part. In a glacial epoch or in the arsis of an historic pulsation, storms are usually abundant and severe, the mean temperature is lower than usual, snow accumulates in high

latitudes or upon lofty mountains. For example, in the last such period during the fourteenth century, great floods and droughts occurred alternately around the North Sea; it was several times possible to cross the Baltic Sea from Germany to Sweden on the ice, and the ice of Greenland advanced so much that shore ice caused the Norsemen to change their sailing route between Iceland and the Norse colonies in southern Greenland. At the same time in low latitudes and in parts of the continental interior there is a tendency toward diminished rainfall and even toward aridity and the formation of deserts. In Yucatan, for example, a diminution in tropical rainfall in the fourteenth century seems to have given the Mayas a last opportunity for a revival of their decaying civilization.

Fig. 1. Climatic changes and mountain building.
(After Schuchert, in The Evolution of the Earth and Its Inhabitants, edited by R. S. Lull.)

Diagram showing the times and probable extent of the more or less marked climate changes in the geologic history of North America, and of its elevation into chains of mountains.

Among the climatic sequences, glacial fluctuations are perhaps of the most vital import from the standpoint of organic evolution; from the standpoint of human history the same is true of climatic pulsations. Glacial epochs have repeatedly wiped out thousands upon thousands of species and played a part in the origin of entirely new types of plants and animals. This is best seen when the life of the Pennsylvanian is contrasted with that of the Permian. An historic pulsation may wipe out an entire civilization and permit a new one to grow up with a radically different character. Hence it is not strange that the causes of such climatic phenomena have been discussed with extraordinary vigor. In few realms of science has there been a more imposing or more interesting array of theories. In this book we shall consider the more important of these theories. A new solar or cyclonic hypothesis and the hypothesis of changes in the form and altitude of the land will receive the most attention, but the other

chief hypotheses are outlined in the next chapter, and are frequently referred to throughout the volume.

Between glacial fluctuations and historical pulsations in duration, but probably less severe than either, come orbital precessions. These stand in a group by themselves and are more akin to seasonal alternations than to any other type of climatic sequence. They must have occurred with absolute regularity ever since the earth began to revolve around the sun in its present elliptical orbit. Since the orbit is elliptical and since the sun is in one of the two foci of the ellipse, the earth's distance from the sun varies. At present the earth is nearest the sun in the northern winter. Hence the rigor of winter in the northern hemisphere is mitigated, while that of the southern hemisphere is increased. In about ten thousand years this condition will be reversed, and in another ten thousand the present conditions will return once more. Such climatic precessions, as we may here call them, must have occurred unnumbered times in the past, but they do not appear to have been large enough to leave in the fossils of the rocks any traces that can be distinguished from those of other climatic sequences.

We come now to Brückner periods and sunspot cycles. The Brückner periods have a length of about thirty-three years. Their existence was suggested at least as long ago as the days of Sir Francis Bacon, whose statement about them is quoted on the flyleaf of this book. They have since been detected by a careful study of the records of the time of harvest, vintage, the opening of rivers to navigation, and the rise or fall of lakes like the Caspian Sea. In his book on Klimaschwankungen seit 1700, Brückner has collected an uncommonly interesting assortment of facts as to the climate of Europe for more than two centuries. More recently, by a study of the rate of

growth of trees, Douglass, in his book on Climatic Cycles and Tree Growth, has carried the subject still further. In general the nature of the 33-year periods seems to be identical with that of the 11- or 12- year sunspot cycle, on the one hand, and of historic pulsations on the other. For a century observers have noted that the variations in the weather which everyone notices from year to year seem to have some relation to sunspots. For generations, however, the relationship was discussed without leading to any definite conclusion. The trouble was that the same change was supposed to take place in all parts of the world. Hence, when every sort of change was found somewhere at any given sunspot stage, it seemed as though there could not be a relationship. Of late years, however, the matter has become fairly clear. The chief conclusions are, first, that when sunspots are numerous the average temperature of the earth's surface is lower than normal. This does not mean that all parts are cooler, for while certain large areas grow cool, others of less extent become warm at times of many sunspots. Second, at times of many sunspots storms are more abundant than usual, but are also confined somewhat closely to certain limited tracks so that elsewhere a diminution of storminess may be noted. This whole question is discussed so fully in Earth and Sun that it need not detain us further in this preliminary view of the whole problem of climate. Suffice it to say that a study of the sunspot cycle leads to the conclusion that it furnishes a clue to many of the unsolved problems of the climate of the past, as well as a key to prediction of the future.

Passing by the seasonal alternations which are fully explained as the result of the revolution of the earth around the sun, we may merely point out that, like the daily vibrations which bring Table 2 to a close, they

emphasize the outstanding fact that the main control of terrestrial climate is the amount of energy received from the sun. This same principle is illustrated by pleionian migrations. The term "pleion" comes from a Greek word meaning "more." It was taken by Arctowski to designate areas or periods where there is an excess of some climatic element, such as atmospheric pressure, rainfall, or temperature. Even if the effect of the seasons is eliminated, it appears that the course of these various elements does not run smoothly. As everyone knows, a period like the autumn of 1920 in the eastern United States may be unusually warm, while a succeeding period may be unseasonably cool. These departures from the normal show a certain rough periodicity. For example, there is evidence of a period of about twenty-seven days, corresponding to the sun's rotation and formerly supposed to be due to the moon's revolution which occupies almost the same length of time. Still other periods appear to have an average duration of about three months and of between two and three years. Two remarkable discoveries have recently been made in respect to such pleions. One is that a given type of change usually occurs simultaneously in a number of well-defined but widely separated centers, while a change of an opposite character arises in another equally well-defined, but quite different, set of centers. In general, areas of high pressure have one type of change and areas of low pressure the other type. So systematic are these relationships and so completely do they harmonize in widely separated parts of the earth, that it seems certain that they must be due to some outside cause, which in all probability can be only the sun. The second discovery is that pleions, when once formed, travel irregularly along the earth's surface. Their paths have not yet been worked out in detail, but a general

migration seems well established. Because of this, it is probable that if unusually warm weather prevails in one part of a continent at a given time, the "thermo-pleion," or excess of heat, will not vanish but will gradually move away in some particular direction. If we knew the path that it would follow we might predict the general temperature along its course for some months in advance. The paths are often irregular, and the pleions frequently show a tendency to break up or suddenly revive. Probably this tendency is due to variations in the sun. When the sun is highly variable, the pleions are numerous and strong, and extremes of weather are frequent. Taken as a whole the pleions offer one of the most interesting and hopeful fields not only for the student of the causes of climatic variations, but for the man who is interested in the practical question of long-range weather forecasts. Like many other climatic phenomena they seem to represent the combined effect of conditions in the sun and upon the earth itself.

The last of the climatic sequences which require explanation is the cyclonic vacillations. These are familiar to everyone, for they are the changes of weather which occur at intervals of a few days, or a week or two, at all seasons, in large parts of the United States, Europe, Japan, and some of the other progressive parts of the earth. They do not, however, occur with great frequency in equatorial regions, deserts, and many other regions. Up to the end of the last century, it was generally supposed that cyclonic storms were purely terrestrial in origin. Without any adequate investigation it was assumed that all irregularities in the planetary circulation of the winds arise from an irregular distribution of heat due to conditions within or upon the earth itself. These irregularities were supposed to produce cyclonic storms

in certain limited belts, but not in most parts of the world. Today this view is being rapidly modified. Undoubtedly, the irregularities due to purely terrestrial conditions are one of the chief contributory causes of storms, but it begins to appear that solar variations also play a part. It has been found, for example, that not only the mean temperature of the earth's surface varies in harmony with the sunspot cycle, but that the frequency and severity of storms vary in the same way. Moreover, it has been demonstrated that the sun's radiation is not constant, but is subject to innumerable variations. This does not mean that the sun's general temperature varies, but merely that at some times heated gases are ejected rapidly to high levels so that a sudden wave of energy strikes the earth. Thus, the present tendency is to believe that the cyclonic variations, the changes of weather which come and go in such a haphazard, irresponsible way, are partly due to causes pertaining to the earth itself and partly to the sun.

From this rapid survey of the types of climatic sequences, it is evident that they may be divided into four great groups. First comes cosmic uniformity, one of the most marvelous and incomprehensible of all known facts. We simply have no explanation which is in any respect adequate. Next come secular progression and geologic oscillations, two types of change which seem to be due mainly to purely terrestrial causes, that is, to changes in the lands, the oceans, and the air. The general tendency of these changes is toward complexity and diversity, thus producing progression, but they are subject to frequent reversals which give rise to oscillations lasting millions of years. The processes by which the oscillations take place are fully discussed in this book. Nevertheless, because they are fairly well understood, they are deferred

until after the third group of sequences has been discussed. This group includes glacial fluctuations, historic pulsations, Brückner periods, sunspot cycles, pleionian migrations, and cyclonic vacillations. The outstanding fact in regard to all of these is that while they are greatly modified by purely terrestrial conditions, they seem to owe their origin to variations in the sun. They form the chief subject of Earth and Sun and in their larger phases are the most important topic of this book also. The last group of sequences includes orbital precessions, seasonal alternations, and daily variations. These may be regarded as purely solar in origin. Yet their influence, like that of each of the other groups, is much modified by the earth's own conditions. Our main problem is to separate and explain the two great elements in climatic changes,—the effects of the sun, on the one hand, and of the earth on the other.

[CHAPTER III]

HYPOTHESES OF CLIMATIC CHANGE

The next step in our study of climate is to review the main hypotheses as to the causes of glaciation. These hypotheses apply also to other types of climatic changes. We shall concentrate on glacial periods, however, not only because they are the most dramatic and well-known types of change, but because they have been more discussed than any other and have also had great influence on evolution. Moreover, they stand near the middle of the types of climatic sequences, and an understanding of them does much to explain the others. In reviewing the various theories we shall not attempt to cover all the ground, but shall merely state the main ideas of the few theories which have had an important influence upon scientific thought.

The conditions which any satisfactory climatic hypothesis must satisfy are briefly as follows:

1. Due weight must be given to the fact that changes of climate are almost certainly due to the combined effect of a variety of causes, both terrestrial and solar or cosmic.
2. Attention must also be paid to both sides in the long controversy as to whether glaciation is due primarily to a diminution in the earth's supply of heat or to a redistribution of the heat through changes in atmospheric and oceanic circulation. At present the great
3. majority of authorities are on the side of a diminution of heat, but the other view also deserves study.
4. A satisfactory hypothesis must explain the frequent synchronism between two great types of phenomena; first, movements of the earth's crust whereby continents are uplifted and mountains upheaved; and, second, great changes of climate which are usually marked by relatively rapid oscillations from one extreme to another.
5. No hypothesis can find acceptance unless it satisfies the somewhat exacting requirements of the geological record, with its frequent but irregular repetition of long, mild periods, relatively cool or intermediate periods like the present, and glacial periods of more or less severity and perhaps accompanying the more or less widespread uplifting of continents. At least during the later glacial periods the hypothesis must explain numerous climatic epochs and stages superposed upon a single general period of continental upheaval. Moreover, although historical geology demands cycles of varied duration and magnitude, it does not furnish evidence of any rigid periodicity causing the cycles to be uniform in length or intensity.
6. Most important of all, a satisfactory explanation of climatic changes and crustal deformation must take account of all the agencies which are now causing similar phenomena. Whether any other agencies should be considered is open to question, although the relative importance of existing agencies may have varied.

I. Croll's Eccentricity Theory. One of the most ingenious and most carefully elaborated scientific hypotheses is Croll's^[10] precessional hypothesis as to the effect of the earth's own motions. So well was this worked out that it was widely accepted for a time and still finds a

place in popular but unscientific books, such as Wells' Outline of History, and even in scientific works like Wright's Quaternary Ice Age. The gist of the hypothesis has already been given in connection with the type of climatic sequence known as orbital precessions. The earth is 93 million miles away from the sun in January and 97 million in July. The earth's axis "precesses," however, just as does that of a spinning top. Hence arises what is known as the precession of the equinoxes, that is, a steady change in the season at which the earth is in perihelion, or nearest to the sun. In the course of 21,000 years the time of perihelion varies from early in January through the entire twelve months and back to January. Moreover, the earth's orbit is slightly more elliptical at certain periods than at others, for the planets sometimes become bunched so that they all pull the earth in one direction. Hence, once in about one hundred thousand years the effect of the elliptical shape of the earth's orbit is at a maximum.

Croll argued that these astronomical changes must alter the earth's climate, especially by their effect on winds and ocean currents. His elaborate argument contains a vast amount of valuable material. Later investigation, however, seems to have proven the inadequacy of his hypothesis. In the first place, the supposed cause does not seem nearly sufficient to produce the observed results. Second, Croll's hypothesis demands that glaciation in the northern and southern hemisphere take place alternately. A constantly growing collection of facts, however, indicates that glaciation does not occur in the two hemispheres alternately, but at the same time. Third, the hypothesis calls for the constant and frequent repetition of glaciation at absolutely regular intervals. The geological

record shows no such regularity, for sometimes several glacial epochs follow in relatively close succession at irregular intervals of perhaps fifty to two hundred thousand years, and thus form a glacial period; and then for millions of years there are none. Fourth, the eccentricity hypothesis provides no adequate explanation for the glacial stages or subepochs, the historic pulsations, and the other smaller climatic variations which are superposed upon glacial epochs and upon one another in bewildering confusion. In spite of these objections, there can be little question that the eccentricity of the earth's orbit and the precession of the equinoxes with the resulting change in the season of perihelion must have some climatic effect. Hence Croll's theory deserves a permanent though minor place in any full discussion of the causes of climatic changes.

II. The Carbon Dioxide Theory. At about the time that the eccentricity theory was being relegated to a minor niche, a new theory was being developed which soon exerted a profound influence upon geological thought. Chamberlin,^[11] adopting an idea suggested by

Tyndall, fired the imagination of geologists by his skillful exposition of the part played by carbon dioxide in causing climatic changes. Today this theory is probably more widely accepted than any other. We have already seen that the amount of carbon dioxide gas in the atmosphere has a decided climatic importance. Moreover, there can be little doubt that the amount of that gas in the atmosphere varies from age to age in response to the extent to which it is set free by volcanoes, consumed by plants, combined with rocks in the process of weathering, dissolved in the ocean or locked up in the form of coal and limestone. The main question is whether such variations can produce changes so rapid as glacial epochs and historical pulsations.

Abundant evidence seems to show that the degree to which the air can be warmed by carbon dioxide is sharply limited. Humphreys, in his excellent book on the Physics of the Air, calculates that a layer of carbon dioxide forty centimeters thick has practically as much blanketing effect as a layer indefinitely thicker. In other words, forty centimeters of carbon dioxide, while having no appreciable

effect on sunlight coming toward the earth, would filter out and thus retain in the atmosphere all the outgoing terrestrial heat that carbon dioxide is capable of absorbing. Adding more would be like adding another filter when the one in operation has already done all that that particular kind of filter is capable of doing. According to Humphreys' calculations, a doubling of the carbon dioxide in the air would in itself raise the average temperature about 1.3°C. and further carbon dioxide would have practically no effect. Reducing the present supply by half would reduce the temperature by essentially the same amount.

The effect must be greater, however, than would appear from the figures given above, for any change in temperature has an effect on the amount of water vapor, which in turn causes further changes of temperature. Moreover, as Chamberlin points out, it is not clear whether Humphreys allows for the fact that when the 40 centimeters of CO₂ nearest the earth has been heated by terrestrial radiation, it in turn radiates half its heat outward and half inward. The outward half is all absorbed in the next layer of carbon dioxide, and so on. The process is much more complex than this, but the end result is that even the last increment of CO₂, that is, the outermost portions in the upper atmosphere, must apparently absorb an infinitesimally small amount of heat. This fact, plus the effect of water vapor, would seem to indicate that a doubling or halving of the amount of CO₂, would have an effect of more than 1.3°C. A change of even 2°C. above or below the present level of the earth's mean temperature would be of very appreciable climatic significance, for it is commonly believed that during the height of the glacial period the mean temperature was only 5° to 8°C. lower than now.

Nevertheless, variations in atmospheric carbon dioxide do not necessarily seem competent to produce the relatively rapid climatic fluctuations of glacial epochs and historic pulsations as distinguished from the longer swings of glacial periods and geological eras. In Chamberlin's view, as in ours, the elevation of the land, the modification of the currents of the air and of the ocean, and all that goes with elevation as a topographic agency constitute a primary cause of climatic changes. A special effect of this is the removal of carbon dioxide from the air by the enhanced processes of weathering. This, as he carefully states, is a very slow process, and cannot of itself lead to anything so sudden as the oncoming of glaciation. But here comes Chamberlin's most distinctive contribution to the subject, namely, the hypothesis that changes in atmospheric temperature arising from variations in atmospheric carbon dioxide are able to cause a reversal of the deep-sea oceanic circulation.

According to Chamberlin's view, the ordinary oceanic circulation of the greater part of geological time was the reverse of the present circulation. Warm water descended to the ocean depths in low latitudes, kept its heat while creeping slowly poleward, and rose in high latitudes producing the warm climate which enabled corals, for example, to grow in high latitudes. Chamberlin holds this opinion largely because there seems to him to be no other reasonable way to account for the enormously long warm periods when heat-loving forms of life lived in what are now polar regions of ice and snow. He explains this reversed circulation by supposing that an abundance of atmospheric carbon dioxide, together with a broad distribution of the oceans, made the atmosphere so warm that the evaporation in low latitudes was far more rapid than now. Hence the surface water of the ocean became

a relatively concentrated brine. Such a brine is heavy and tends to sink, thereby setting up an oceanic circulation the reverse of that which now prevails. At present the polar waters sink because they are cold and hence contract. Moreover, when they freeze a certain amount of salt leaves the ice and thereby increases the salinity of the surrounding water. Thus the polar water sinks to the depths of the ocean, its place is taken by warmer and lighter water from low latitudes which moves poleward along the surface, and at the same time the cold water of the ocean depths is forced equatorward below the surface. But if the equatorial waters were so concentrated that a steady supply of highly saline water kept descending to low levels, the direction of the circulation would have to be reversed. The time when this would occur would depend upon the delicate balance between the downward tendencies of the cold polar water and of the warm saline equatorial water.

Suppose that while such a reversed circulation prevailed, the atmospheric CO₂ should be depleted, and the air cooled so much that the concentration of the equatorial waters by evaporation was no longer sufficient to cause them to sink. A reversal would take place, the present type of circulation would be inaugurated, and the whole earth would suffer a chill because the surface of the ocean would become cool. The cool surface-water would absorb carbon dioxide faster than the previous warm water had done, for heat drives off gases from water. This would hasten the cooling of the atmosphere still more, not only directly but by diminishing the supply of atmospheric moisture. The result would be glaciation. But ultimately the cold waters of the higher latitudes would absorb all the carbon dioxide they could hold, the slow equatorward creep would at length permit

the cold water to rise to the surface in low latitudes. There the warmth of the equatorial sun and the depleted supply of carbon dioxide in the air would combine to cause the water to give up its carbon dioxide once more. If the atmosphere had been sufficiently depleted by that time, the rising waters in low latitudes might give up more carbon dioxide than the cold polar waters absorbed. Thus the atmospheric supply would increase, the air would again grow warm, and a tendency toward deglaciation, or toward an inter-glacial condition would arise. At such times the oceanic circulation is not supposed to have been reversed, but merely to have been checked and made slower by the increasing warmth. Thus inter-glacial conditions like those of today, or even considerably warmer, are supposed to have been produced with the present type of circulation.

The emission of carbon dioxide in low latitudes could not permanently exceed the absorption in high latitudes. After the present type of circulation was finally established, which might take tens of thousands of years, the two would gradually become equal. Then the conditions which originally caused the oceanic circulation to be reversed would again destroy the balance; the atmospheric carbon dioxide would be depleted; the air would grow cooler; and the cycle of glaciation would be repeated. Each cycle would be shorter than the last, for not only would the swings diminish like those of a pendulum, but the agencies that were causing the main depletion of the atmospheric carbon dioxide would diminish in intensity. Finally as the lands became lower through erosion and submergence, and as the processes of weathering became correspondingly slow, the air would gradually be able to accumulate carbon dioxide; the temperature would increase; and at length the oceanic circulation would be

reversed again. When the warm saline waters of low latitudes finally began to sink and to set up a flow of warm water poleward in the depths of the ocean, a glacial period would definitely come to an end.

This hypothesis has been so skillfully elaborated, and contains so many important elements that one can scarcely study it without profound admiration. We believe that it is of the utmost value as a step toward the truth, and especially because it emphasizes the great function of oceanic circulation. Nevertheless, we are unable to accept it in full for several reasons, which may here be stated very briefly. Most of them will be discussed fully in later pages.

(1) While a reversal of the deep-sea circulation would undoubtedly be of great climatic importance and would produce a warm climate in high latitudes, we see no direct evidence of such a reversal. It is equally true that there is no conclusive evidence against it, and the possibility of a reversal must not be overlooked. There seem, however, to be other modifications of atmospheric and oceanic circulation which are able to produce the observed results.

(2) There is much, and we believe conclusive, evidence that a mere lowering of temperature would not produce glaciation. What seems to be needed is changes in atmospheric circulation and in precipitation. The carbon dioxide hypothesis has not been nearly so fully developed on the meteorological side as in other respects.

(3) The carbon dioxide hypothesis seems to demand that the oceans should have been almost as saline as now in the Proterozoic era at the time of the first known glaciation. Chamberlin holds that such was the case, but the constant supply of saline material brought to the ocean by rivers and the relatively small deposition of

such material on the sea floor seem to indicate that the early oceans must have been much fresher than those of today.

(4) The carbon dioxide hypothesis does not attempt to explain minor climatic fluctuations such as post-glacial stages and historic pulsations, but these appear to be of the same nature as glacial epochs, differing only in degree.

(5) Another reason for hesitation in accepting the carbon dioxide hypothesis as a full explanation of glacial fluctuations is the highly complex and non-observational character of the explanation of the alternation of glacial and inter-glacial epochs and of their constantly decreasing length.

(6) Most important of all, a study of the variations of weather and of climate as they are disclosed by present records and by the historic past suggests that there are now in action certain other causes which are competent to explain glaciation without recourse to a process whose action is beyond the realm of observation.

These considerations lead to the conclusion that the carbon dioxide hypothesis and the reversal of the oceanic circulation should be regarded as a tentative rather than a final explanation of glaciation. Nevertheless, the action of carbon dioxide seems to be an important factor in producing the longer oscillations of climate from one geological era to another. It probably plays a considerable part in preparing the way for glacial periods and in making it possible for other factors to produce the more rapid changes which have so deeply influenced organic evolution.

III. The Form of the Land. Another great cause of climatic change consists of a group of connected phenomena dependent upon movements of the earth's crust.

As to the climatic potency of changes in the lands there is practical agreement among students of climatology and glaciation. That the height and extent of the continents, the location, size, and orientation of mountain ranges, and the opening and closing of oceanic gateways at places like Panama, and the consequent diversion of oceanic currents, exert a profound effect upon climate can scarcely be questioned. Such changes may be introduced rapidly, but their disappearance is usually slow compared with the rapid pulsations to which climate has been subject during historic times and during stages of glacial retreat and advance, or even in comparison with the epochs into which the Pleistocene, Permian, and perhaps earlier glacial periods have been divided. Hence, while crustal movements appear to be more important than the eccentricity of the earth's orbit or the amount of carbon dioxide in the air, they do not satisfactorily explain glacial fluctuations, historic pulsations, and especially the present little cycles of climatic change. All these changes involve a relatively rapid swing from one extreme to another, while an upheaval of a continent, which is at best a slow geologic process, apparently cannot be undone for a long, long time. Hence such an upheaval, if acting alone, would lead to a relatively long-lived climate of a somewhat extreme type. It would help to explain the long swings, or geologic oscillations between a mild and uniform climate at one extreme, and a complex and varied climate at the other, but it would not explain the rapid climatic pulsations which are closely associated with great movements of the earth's crust. It might prepare the way for them, but could not cause them. That this conclusion is true is borne out by the fact that vast mountain ranges, like those at the close of the Jurassic and Cretaceous, are upheaved without bringing

on glacial climates. Moreover, the marked Permian ice age follows long after the birth of the Hercynian Mountains and before the rise of others of later Permian origin.

IV. The Volcanic Hypothesis. In the search for some cause of climatic change which is highly efficient and yet able to vary rapidly and independently, Abbot, Fowle, Humphreys, and others,^[12] have concluded that volcanic eruptions are the missing agency. In Physics of the Air, Humphreys gives a careful study of the effect of volcanic dust upon terrestrial temperature. He begins with a mathematical investigation of the size of dust particles, and their quantity after certain eruptions. He demonstrates that the power of such particles to deflect light of short wave-lengths coming from the sun is perhaps thirty times more than their power to retain the heat radiated in long waves from the earth. Hence it is estimated that if a Krakatoa were to belch forth dust every year or two, the dust veil might cause a reduction of about 6°C. in the earth's surface temperature. As in every such complicated problem, some of the author's assumptions are open to question, but this touches their quantitative and not their qualitative value. It seems certain that if volcanic explosions were frequent enough and violent enough, the temperature of the earth's surface would be considerably lowered.

Actual observation supports this theoretical conclusion. Humphreys gathers together and amplifies all that he and Abbot and Fowle have previously said as to observations of the sun's thermal radiation by means of the

pyrheliometer. This summing up of the relations between the heat received from the sun, and the occurrence of explosive volcanic eruptions leaves little room for doubt that at frequent intervals during the last century and a half a slight lowering of terrestrial temperature has actually occurred after great eruptions. Nevertheless, it does not justify Humphreys' final conclusion that "phenomena within the earth itself suffice to modify its own climate,... that these and these alone have actually caused great changes time and again in the geologic past." Humphreys sees so clearly the importance of the purely terrestrial point of view that he unconsciously slights the cosmic standpoint and ignores the important solar facts which he himself adduces elsewhere at considerable length.

In addition to this the degree to which the temperature of the earth as a whole is influenced by volcanic eruptions is by no means so clear as is the fact that there is some influence. Arctowski,^[13] for example, has prepared numerous curves showing the march of temperature month after month for many years. During the period from 1909 to 1913, which includes the great eruption of Katmai in Alaska, low temperature is found to have prevailed at the time of the eruption, but, as Arctowski puts it, on the basis of the curves for 150 stations in all parts of the world: "The supposition that these abnormally low temperatures were due to the veil of volcanic dust produced by the Katmai eruption of June 6, 1912, is completely out of the question. If that had been the case, temperature would have decreased from that date on, whereas it was decreasing for more than a year before that date."

Köppen,^[14] in his comprehensive study of temperature for a hundred years, also presents a strong argument against the idea that volcanic eruptions have an important place in determining the present temperature of the earth. A volcanic eruption is a sudden occurrence. Whatever effect is produced by dust thrown into the air must occur within a few months, or as soon as the dust has had an opportunity to be wafted to the region in question. When the dust arrives, there will be a rapid drop through the few degrees of temperature which the dust is supposed to be able to account for, and thereafter a slow rise of temperature. If volcanic eruptions actually caused a frequent lowering of terrestrial temperature in the hundred years studied by Köppen, there should be more cases where the annual temperature is decidedly below the normal than where it shows a large departure in the opposite direction. The contrary is actually the case.

A still more important argument is the fact that the earth is now in an intermediate condition of climate. Throughout most of geologic time, as we shall see again and again, the climate of the earth has been milder than now. Regions like Greenland have not been the seat of glaciers, but have been the home of types of plants which now thrive in relatively low latitudes. In other words, the earth is today only part way from a glacial epoch to what may be called the normal, mild climate of the earth—a climate in which the contrast from zone to zone was much less than now, and the lower air averaged warmer. Hence it seems impossible to avoid the conclusion that the cause of glaciation is still operating with considerable

although diminished efficiency. But volcanic dust is obviously not operating to any appreciable extent at present, for the upper air is almost free from dust a large part of the time.

Again, as Chamberlin suggests, let it be supposed that a Krakatoan eruption every two years would produce a glacial period. Unless the most experienced field workers on the glacial formations are quite in error, the various glacial epochs of the Pleistocene glacial period had a joint duration of at least 150,000 years and perhaps twice as much. That would require 75,000 Krakatoan eruptions. But where are the pits and cones of such eruptions? There has not been time to erode them away since the Pleistocene glaciation. Their beds of volcanic ash would presumably be as voluminous as the glacial beds, but there do not seem to be accumulations of any such size. Even though the same volcano suffered repeated explosions, it seems impossible to find sufficient fresh volcanic debris. Moreover, the volcanic hypothesis has not yet offered any mechanism for systematic glacial variations. Hence, while the hypothesis is important, we must search further for the full explanation of glacial fluctuations, historic pulsations, and the earth's present quasi-glacial climate.

V. The Hypothesis of Polar Wandering. Another hypothesis, which has some adherents, especially among geologists, holds that the position of the earth's axis has shifted repeatedly during geological times, thus causing glaciation in regions which are not now polar. Astrophysicists, however, are quite sure that no agency could radically change the relation between the earth and its axis without likewise altering the orbits of the planets to a degree that would be easily recognized. Moreover, the distribution of the centers of glaciation both in the Permian

and Pleistocene periods does not seem to conform to this hypothesis.

VI. The Thermal Solar Hypothesis. The only other explanations of the climatic changes of glacial and historic times which now seem to have much standing are two distinct and almost antagonistic solar hypotheses. One is the idea that changes in the earth's climate are due to variations in the heat emitted by the sun and hence in the temperature of the earth. The other is the entirely different idea that climatic changes arise from solar conditions which cause a redistribution of the earth's atmospheric pressure and hence produce changes in winds, ocean currents, and especially storms. This second, or "cyclonic," hypothesis is the subject of a book entitled Earth and Sun, which is to be published as a companion to the present volume. It will be outlined in the next chapter. The other, or thermal, hypothesis may be dismissed briefly. Unquestionably a permanent change in the amount of heat emitted by the sun would permanently alter the earth's climate. There is absolutely no evidence, however, of any such change during geologic time. The evidence as to the earth's cosmic uniformity and as to secular progression is all against it. Suppose that for thirty or forty thousand years the sun cooled off enough so that the earth was as cool as during a glacial epoch. As glaciation is soon succeeded by a mild climate, some agency would then be needed to raise the sun's temperature. The impact of a shower of meteorites might accomplish this, but that would mean a very sudden heating, such as there is no evidence of in geological history. In fact, there is far more evidence of sudden cooling than of sudden heating. Moreover, it is far beyond the bounds of probability that such an impact should be repeated again and again with just such force as to bring the climate

back almost to where it started and yet to allow for the slight changes which cause secular progression. Another and equally cogent objection to the thermal form of solar hypothesis is stated by Humphreys as follows: "A change of the solar constant obviously alters all surface temperatures by a roughly constant percentage. Hence a decrease of the heat from the sun would in general cause a decrease of the interzonal temperature gradients; and this in turn a less vigorous atmospheric circulation, and a less copious rain or snowfall—exactly the reverse of the condition, namely, abundant precipitation, most favorable to extensive glaciation."

This brings us to the end of the main hypotheses as to climatic changes, aside from the solar cyclonic hypothesis which will be discussed in the next chapter. It appears that variations in the position of the earth at perihelion have a real though slight influence in causing cycles with a length of about 21,000 years. Changes in the carbon dioxide of the air probably have a more important but extremely slow influence upon geologic oscillations. Variations in the size, shape, and height of the continents are constantly causing all manner of climatic complications, but do not cause rapid fluctuations and pulsations. The eruption of volcanic dust appears occasionally to lower the temperature, but its potency to explain the complex climatic changes recorded in the rocks has probably been exaggerated. Finally, although minor changes in the amount of heat given out by the sun occur constantly and have been demonstrated to have a climatic effect, there is no evidence that such changes are the main cause of the climatic phenomena which we are trying to explain. Nevertheless, in connection with other solar changes they may be of high importance.

[CHAPTER IV]

THE SOLAR CYCLONIC HYPOTHESIS

The progress of science is made up of a vast succession of hypotheses. The majority die in early infancy. A few live and are for a time widely accepted. Then some new hypothesis either destroys them completely or shows that, while they contain elements of truth, they are not the whole truth. In the previous chapter we have discussed a group of hypotheses of this kind, and have tried to point out fairly their degree of truth so far as it can yet be determined. In this chapter we shall outline still another hypothesis, the relation of which to present climatic conditions has been fully developed in Earth and Sun; while its relation to the past will be explained in the present volume. This hypothesis is not supposed to supersede the others, for so far as they are true they cannot be superseded. It merely seems to explain some of the many conditions which the other hypotheses apparently fail to explain. To suppose that it will suffer a fate more glorious than its predecessors would be presumptuous. The best that can be hoped is that after it has been pruned, enriched, and modified, it may take its place among the steps which finally lead to the goal of truth.

In this chapter the new hypothesis will be sketched in broad outline in order that in the rest of this book the reader may appreciate the bearing of all that is said. Details of proof and methods of work will be omitted,

since they are given in Earth and Sun. For the sake of brevity and clearness the main conclusions will be stated without the qualifications and exceptions which are fully explained in that volume. Here it will be necessary to pass quickly over points which depart radically from accepted ideas, and which therefore must arouse serious question in the minds of thoughtful readers. That, however, is a necessary consequence of the attempt which this book makes to put the problem of climate in such form that the argument can be followed by thoughtful students in any branch of knowledge and not merely by specialists. Therefore, the specialist can merely be asked to withhold judgment until he has read all the evidence as given in Earth and Sun, and then to condemn only those parts that are wrong and not the whole argument.

Without further explanation let us turn to our main problem. In the realm of climatology the most important discovery of the last generation is that variations in the weather depend on variations in the activity of the sun's atmosphere. The work of the great astronomer, Newcomb, and that of the great climatologist, Köppen, have shown beyond question that the temperature of the earth's surface varies in harmony with variations in the number and area of sunspots.^[15] The work of Abbot has shown that the amount of heat radiated from the sun also varies, and that in general the variations correspond with those of the sunspots, although there are exceptions, especially when the spots are fewest. Here, however, there at once arises a puzzling paradox. The earth certainly

owes its warmth to the sun. Yet when the sun emits the most energy, that is, when sunspots are most numerous, the earth's surface is coolest. Doubtless the earth receives more heat than usual at such times, and the upper air may be warmer than usual. Here we refer only to the air at the earth's surface.

Another large group of investigators have shown that atmospheric pressure also varies in harmony with the number of sunspots. Some parts of the earth's surface have one kind of variation at times of many sunspots and other parts the reverse. These differences are systematic and depend largely on whether the region in question happens to have high atmospheric pressure or low. The net result is that when sunspots are numerous the earth's storminess increases, and the atmosphere is thrown into commotion. This interferes with the stable planetary winds, such as the trades of low latitudes and the prevailing westerlies of higher latitudes. Instead of these regular winds and the fair weather which they bring, there is a tendency toward frequent tropical hurricanes in the lower latitudes and toward more frequent and severe storms of the ordinary type in the latitudes where the world's most progressive nations now live. With the change in storminess there naturally goes a change in rainfall. Not all parts of the world, however, have increased storminess and more abundant rainfall when sunspots are numerous. Some parts change in the opposite way. Thus when the sun's atmosphere is particularly disturbed, the contrasts between different parts of the earth's surface are increased. For example, the northern United States and southern Canada become more stormy and rainy, as appears in Fig. 2, and the same is true of the Southwest and along the south Atlantic coast. In a crescent-shaped central area, however,

extending from Wyoming through Missouri to Nova Scotia, the number of storms and the amount of rainfall decrease.

Fig. 2. Storminess of sunspot maxima vs. minima.
(After Kullmer.)

Based on nine years' nearest sunspot minima and nine years' nearest sunspot maxima in the three sunspot cycles from 1888 to 1918. Heavy shading indicates excess of storminess when sunspots are numerous. Figures indicate average yearly number of storms by which years of maximum sunspots exceed those of minimum sunspots.

The two controlling factors of any climate are the temperature and the atmospheric pressure, for they determine the winds, the storms, and thus the rainfall. A study of the temperature seems to show that the peculiar paradox of a hot sun and a cool earth is due largely to the increased storminess during times of many sunspots. The earth's surface is heated by the rays of the sun, but

most of the rays do not in themselves heat the air as they pass through it. The air gets its heat largely from the heat absorbed by the water vapor which is intimately mingled with its lower portions, or from the long heat waves sent out by the earth after it has been warmed by the sun. The faster the air moves along the earth's surface the less it becomes heated, and the more heat it takes away. This sounds like a contradiction, but not to anyone who has tried to heat a stove in the open air. If the air is still, the stove rapidly becomes warm and so does the air around it. If the wind is blowing, the cool air delays the heating of the stove and prevents the surface from ever becoming as hot as it would otherwise. That seems to be what happens on a large scale when sunspots are numerous. The sun actually sends to the earth more energy than usual, but the air moves with such unusual rapidity that it actually cools the earth's surface a trifle by carrying the extra heat to high levels where it is lost into space.

There has been much discussion as to why storms are numerous when the sun's atmosphere is disturbed. Many investigators have supposed it was due entirely and directly to the heating of the earth's surface by the sun. This, however, needs modification for several reasons. In the first place, recent investigations show that in a great many cases changes in barometric pressure precede changes in temperature and apparently cause them by altering the winds and producing storms. This is the opposite of what would happen if the effect of solar heat upon the earth's surface were the only agency. In the second place, if storms were due exclusively to variations in the ordinary solar radiation which comes to the earth as light and is converted into heat, the solar effect ought

to be most pronounced when the center of the sun's visible disk is most disturbed. As a matter of fact the storminess is notably greatest when the edges of the solar disk are most disturbed. These facts and others lead to the conclusion that some agency other than heat must also play some part in producing storminess.

The search for this auxiliary agency raises many difficult questions which cannot yet be answered. On the whole the weight of evidence suggests that electrical phenomena of some kind are involved, although variations in the amount of ultra-violet light may also be important. Many investigators have shown that the sun emits electrons. Hale has proved that the sun, like the earth, is magnetized. Sunspots also have magnetic fields the strength of which is often fifty times as great as that of the sun as a whole. If electrons are sent to the earth, they must move in curved paths, for they are deflected by the sun's magnetic field and again by the earth's magnetic field. The solar deflection may cause their effects to be greatest when the spots are near the sun's margin; the terrestrial deflection may cause concentration in bands roughly concentric with the magnetic poles of the earth. These conditions correspond with the known facts.

Farther than this we cannot yet go. The calculations of Humphreys seem to indicate that the direct electrical effect of the sun's electrons upon atmospheric pressure is too small to be of appreciable significance in intensifying storms. On the other hand the peculiar way in which activity upon the margins of the sun appears to be correlated not only with atmospheric electricity, but with barometric pressure, seems to be equally strong evidence in the other direction. Possibly the sun's electrons and its electrical waves produce indirect effects by being

converted into heat, or by causing the formation of ozone and the condensation of water vapor in the upper air. Any one of these processes would raise the temperature of the upper air, for the ozone and the water vapor would be formed there and would tend to act as a blanket to hold in the earth's heat. But any such change in the temperature of the upper air would influence the lower air through changes in barometric pressure. These considerations are given here because the thoughtful reader is likely to inquire how solar activity can influence storminess. Moreover, at the end of this book we shall take up certain speculative questions in which an electrical hypothesis will be employed. For the main portions of this book it makes no difference how the sun's variations influence the earth's atmosphere. The only table 6 essential point is that when the solar atmosphere is active the storminess of the earth increases, and that is a matter of direct observation.

Let us now inquire into the relation between the small cyclonic vacillations of the weather and the types of climatic changes known as historic pulsations and glacial fluctuations. One of the most interesting results of recent investigations is the evidence that sunspot cycles on a small scale present almost the same phenomena as do historic pulsations and glacial fluctuations. For instance, when sunspots are numerous, storminess increases markedly in a belt near the northern border of the area of greatest storminess, that is, in southern Canada and thence across the Atlantic to the North Sea and Scandinavia. (See Figs. 2 and 3.) Corresponding with this is the fact that the evidence as to climatic pulsations in historic times indicates that regions along this path, for instance Greenland, the North Sea region, and southern Scandinavia,

were visited by especially frequent and severe storms at the climax of each pulsation. Moreover, the greatest accumulations of ice in the glacial period were on the poleward border of the general regions where now the storms appear to increase most at times of solar activity.

Fig. 3.a Relative rainfall at times of increasing and decreasing sunspots

Heavy shading, more rain with increasing spots. Light shading, more rain with decreasing spots. No data for unshaded areas.

Figures indicate percentages of the average rainfall by which the rainfall during periods of increasing spots exceeds or falls short of rainfall during periods of decreasing spots. The excess or deficiency is stated in percentages of the average. Rainfall data from Walker: Sunspots and Rainfall.

Fig. 3.b Relative rainfall at times of increasing and decreasing sunspots.

Heavy shading, more rain with increasing spots. Light shading, more rain with decreasing spots. No data for unshaded areas.

Figures indicate percentages of the average rainfall by which the rainfall during periods of increasing spots exceeds or falls short of rainfall during periods of decreasing spots. The excess or deficiency is stated in percentages of the average. Rainfall data from Walker: Sunspots and Rainfall.

Even more clear is the evidence from other regions where storms increase at times of many sunspots. One such region includes the southwestern United States, while another is the Mediterranean region and the semi-arid or desert parts of Asia farther east. In these regions innumerable ruins and other lines of evidence show that at the climax of each climatic pulsation there was more storminess and rainfall than at present, just as there now is when the sun is most active. In still earlier times, while ice was accumulating farther north, the basins of these semi-arid regions were filled with lakes whose strands still remain to tell the tale of much-increased rainfall and presumable storminess. If we go back still further in geological times to the Permian glaciation, the areas where ice accumulated most abundantly appear to be the regions where tropical hurricanes produce the greatest rainfall and the greatest lowering of temperature at times of many sunspots. From these and many other lines of evidence it seems probable that historic pulsations and glacial fluctuations are nothing more than sunspot cycles on a large scale. It is one of the fundamental rules of science to reason from the known to the unknown, from the near to the far, from the present to the past. Hence it seems advisable to investigate whether any of the climatic phenomena of the past may have arisen from an intensification of the solar conditions which now appear to give rise to similar phenomena on a small scale.

The rest of this chapter will be devoted to a résumé of certain tentative conclusions which have no bearing on the main part of this book, but which apply to the closing chapters. There we shall inquire into the periodicity of the climatic phenomena of geological times, and shall ask whether there is any reason to suppose that the sun's activity has exhibited similar periodicity. This leads to an investigation of the possible causes of disturbances in the sun's atmosphere. It is generally assumed that sunspots, solar prominences, the bright clouds known as faculæ, and other phenomena denoting a perturbed state of the solar atmosphere, are due to some cause within the sun. Yet the limitation of these phenomena, especially the sunspots, to restricted latitudes, as has been shown in Earth and Sun, does not seem to be in harmony with an internal solar origin, even though a banded arrangement may be normal for a rotating globe. The fairly regular periodicity of the sunspots seems equally out of harmony with an internal origin. Again, the solar atmosphere has two kinds of circulation, one the so-called "rice grains," and the other the spots and their attendant phenomena. Now the rice grains present the appearance that would be expected in an atmospheric circulation arising from the loss of heat by the outer part of a gaseous body like the sun. For these reasons and others numerous good thinkers from Wolf to Schuster have held that sunspots owe their periodicity to causes outside the sun. The only possible cause seems to be the planets, acting either through gravitation, through forces of an electrical origin, or through some other agency. Various new investigations which are described in Earth and Sun support this conclusion. The chief difficulty in accepting it hitherto has been that although Jupiter, because of its size, would be

expected to dominate the sunspot cycle, its period of 11.86 years has not been detected. The sunspot cycle has appeared to average 11.2 years in length, and has been called the 11-year cycle. Nevertheless, a new analysis of the sunspot data shows that when attention is concentrated upon the major maxima, which are least subject to retardation or acceleration by other causes, a periodicity closely approaching that of Jupiter is evident. Moreover, when the effects of Jupiter, Saturn, and the other planets are combined, they produce a highly variable curve which has an extraordinary resemblance to the sunspot curve. The method by which the planets influence the sun's atmosphere is still open to question. It may be through tides, through the direct effect of gravitation, through electro-magnetic forces, or in some other way. Whichever it may be, the result may perhaps be slight differences of atmospheric pressure upon the sun. Such differences may set in motion slight whirling movements analogous to terrestrial storms, and these presumably gather momentum from the sun's own energy. Since the planetary influences vary in strength because of the continuous change in the relative distances and positions of the planets, the sun's atmosphere appears to be swayed by cyclonic disturbances of varying degrees of severity. The cyclonic disturbances known as sunspots have been proved by Hale to become more highly electrified as they increase in intensity. At the same time hot gases presumably well up from the lower parts of the solar atmosphere and thereby cause the sun to emit more heat. Thus by one means or another, the earth's atmosphere appears to be set in commotion and cycles of climate are inaugurated.

If the preceding reasoning is correct, any disturbance of the solar atmosphere must have an effect upon the

earth's climate. If the disturbance were great enough and of the right nature it might produce a glacial epoch. The planets are by no means the only bodies which act upon the sun, for that body sustains a constantly changing relation to millions of other celestial bodies of all sizes up to vast universes, and at all sorts of distances. If the sun and another star should approach near enough to one another, it is certain that the solar atmosphere would be disturbed much more than at present.

Here we must leave the cyclonic hypothesis of climate and must refer the reader once more to Earth and Sun for fuller details. In the rest of this book we shall discuss the nature of the climatic changes of past times and shall inquire into their relation to the various climatic hypotheses mentioned in the last two chapters. Then we shall inquire into the possibility that the solar system has ever been near enough to any of the stars to cause appreciable disturbances of the solar atmosphere. We shall complete our study by investigating the vexed question of why movements of the earth's crust, such as the uplifting of continents and mountain chains, have generally occurred at the same time as great climatic fluctuations. This would not be so surprising were it not that the climatic phenomena appear to have consisted of highly complex cycles while the uplift has been a relatively steady movement in one direction. We shall find some evidence that the solar disturbances which seem to cause climatic changes also have a relation to movements of the crust.

[CHAPTER V]

THE CLIMATE OF HISTORY^[16]

We are now prepared to consider the climate of the past. The first period to claim attention is the few thousand years covered by written history. Strangely enough, the conditions during this time are known with less accuracy than are those of geological periods hundreds of times more remote. Yet if pronounced changes have occurred since the days of the ancient Babylonians and since the last of the post-glacial stages, they are of great importance not only because of their possible historic effects, but because they bridge the gap between the little variations of climate which are observable during a single lifetime and the great changes known as glacial epochs. Only by bridging the gap can we determine whether there is any genetic relation between the great changes and the small. A full discussion of the climate of historic times is not here advisable, for it has been considered in detail in numerous other publications.^[17] Our most profitable course would seem to be to consider first the general trend of opinion and then to take up the chief objections to each of the main hypotheses.

In the hot debate over this problem during recent

decades the ideas of geographers seem to have gone through much the same metamorphosis as have those of geologists in regard to the climate of far earlier times.

As every geologist well knows, at the dawn of geology people believed in climatic uniformity—that is, it was supposed that since the completion of an original creative act there had been no important changes. This view quickly disappeared and was superseded by the hypothesis of progressive cooling and drying, an hypothesis which had much to do with the development of the nebular hypothesis, and which has in turn been greatly strengthened by that hypothesis. The discovery of evidence of widespread continental glaciation, however, necessitated a modification of this view, and succeeding years have brought to light a constantly increasing number of glacial, or at least cool, periods distributed throughout almost the whole of geological time. Moreover, each year, almost, brings new evidence of the great complexity of glacial periods, epochs, and stages. Thus, for many decades, geologists have more and more been led to believe that in spite of surprising uniformity, when viewed in comparison with the cosmic possibilities, the climate of the past has been highly unstable from the viewpoint of organic evolution, and its changes have been of all degrees of intensity.

Geographers have lately been debating the reality of historic changes of climate in the same way in which geologists debated the reality of glacial epochs and stages. Several hypotheses present themselves but these may all be grouped under three headings; namely, the hypotheses of (1) progressive desiccation, (2) climatic uniformity, and (3) pulsations. The hypothesis of progressive desiccation has been widely advocated. In many of the drier portions of the world, especially between 30°

and 40° from the equator, and preëminently in western and central Asia and in the southwestern United States, almost innumerable facts seem to indicate that two or three thousand years ago the climate was distinctly moister than at present. The evidence includes old lake strands, the traces of desiccated springs, roads in places now too dry for caravans, other roads which make detours around dry lake beds where no lakes now exist, and fragments of dead forests extending over hundreds of square miles where trees cannot now grow for lack of water. Still stronger evidence is furnished by ancient ruins, hundreds of which are located in places which are now so dry that only the merest fraction of the former inhabitants could find water. The ruins of Palmyra, in the Syrian Desert, show that it must once have been a city like modern Damascus, with one or two hundred thousand inhabitants, but its water supply now suffices for only one or two thousand. All attempts to increase the water supply have had only a slight effect and the water is notoriously sulphurous, whereas in the former days, when it was abundant, it was renowned for its excellence. Hundreds of pages might be devoted to describing similar ruins. Some of them are even more remarkable for their dryness than is Niya, a site in the Tarim Desert of Chinese Turkestan. Yet there the evidence of desiccation within 2000 years is so strong that even so careful and conservative a man as Hann,^[18] pronounces it "überzeugend."

A single quotation from scores that might be used will illustrate the conclusions of some of the most careful archæologists.^[19]

Among the regions which were once populous and highly civilized, but which are now desert and deserted, there are few which were more closely connected with the beginnings of our own civilization than the desert parts of Syria and northern Arabia. It is only of recent years that the vast extent and great importance of this lost civilization has been fully recognized and that attempts have been made to reduce the extent of the unexplored area and to discover how much of the territory which has long been known as desert was formerly habitable and inhabited. The results of the explorations of the last twenty years have been most astonishing in this regard. It has been found that practically all of the wide area lying between the coast range of the eastern Mediterranean and the Euphrates, appearing upon the maps as the Syrian Desert, an area embracing somewhat more than 20,000 square miles, was more thickly populated than any area of similar dimensions in England or in the United States is today if one excludes the immediate vicinity of the large modern cities. It has also been discovered that an enormous desert tract lying to the east of Palestine, stretching eastward and southward into the country which we know as Arabia, was also a densely populated country. How far these settled regions extended in antiquity is still unknown, but the most distant explorations in these directions have failed to reach the end of ruins and other signs of former occupation.

The traveler who has crossed the settled, and more or less populous, coast range of northern Syria and descended into the narrow fertile valley of the Orontes, encounters in any farther journey toward the east an irregular range of limestone hills lying north and south and stretching to the northeast almost halfway to the Euphrates. These hills are about 2,500 feet high, rising in occasional peaks from 3,000 to 3,500 feet above sea level. They are gray and unrelieved by any visible vegetation. On ascending into the hills the traveler is astonished to find at every turn remnants of the work of men's hands, paved roads, walls which divided fields, terrace walls of massive structure. Presently he comes upon a small deserted and partly ruined town composed of buildings large and small constructed of beautifully wrought blocks of limestone, all rising out of the barren rock which forms the ribs of the hills. If he mounts an eminence in the vicinity, he will be still further astonished to behold similar ruins lying in all directions. He may count ten or fifteen or twenty, according to the commanding position of his lookout. From a distance it is often difficult to believe that these are not inhabited places; but closer inspection reveals that the gentle hand of time or the rude touch of earthquake has been laid upon every building. Some of the towns are better preserved than others; some buildings are quite perfect but for their wooden roofs which time has removed, others stand in picturesque ruins, while others still are level with the ground. On a far-off hilltop stands the ruin of a pagan temple, and crowning some lofty ridge lie the ruins of a great Christian monastery. Mile after mile of this barren gray country may be traversed without encountering a single human being. Day after day may be spent in traveling from one ruined town to another without seeing any green thing save a terebinth tree or two standing among the ruins, which have sent their roots down into earth still preserved in the foundations of some ancient building. No soil is visible anywhere except in a few pockets in the rock from which it could not be washed by the torrential rains of the wet season; yet every ruin is surrounded with the remains of presses for the making of oil and wine. Only one oasis has been discovered in these high plateaus.

Passing eastward from this range of hills, one descends into a gently rolling country that stretches miles away toward the Euphrates. At the eastern foot of the hills one finds oneself in a totally different country, at first quite fertile and dotted with frequent villages of flat-roofed houses. Here practically all the remains of ancient times have been destroyed through ages of building and rebuilding. Beyond this narrow fertile strip the soil grows drier and more barren, until presently another kind of desert is reached, an undulating waste of dead soil. Few walls or towers or arches rise to break the monotony of the unbroken landscape; but the careful explorer will find on closer examination that this region was more thickly populated in antiquity even than the hill country to the west. Every unevenness of the surface marks the site of a town, some of them cities of considerable extent.

We may draw certain very definite conclusions as to the former conditions of the country itself. There was soil upon the northern hills where none now exists, for the buildings now show unfinished foundation courses which were not intended to be seen; the soil in depressions without outlets is deeper than it formerly was; there are hundreds of olive and wine presses in localities where no tree or vine could now find footing; and there are hillsides with ruined terrace walls rising one above the other with no sign of earth near them. There was also a large natural water supply. In the north as well as in the south we find the dry beds of rivers, streams, and brooks with sand and pebbles and well-worn rocks but no water in them from one year's end to the other. We find bridges over these dry streams and crudely made washing boards along their banks directly below deserted towns. Many of the bridges span the beds of streams that seldom or never have water in them and give clear evidence of the great climatic changes that have taken place. There are well heads and well houses, and inscriptions referring to springs; but neither wells nor springs exist today except in the rarest instances. Many of the houses had their rock-hewn cisterns, never large enough to have supplied water for more than a brief period, and corresponding to the cisterns which most of our recent forefathers had which were for convenience rather than for dependence. Some of the towns in southern Syria were provided with large public reservoirs, but these are not large enough to have supplied water to their original populations. The high plateaus were of course without irrigation; but there are no signs, even in the lower flatter country, that irrigation was ever practiced; and canals for this purpose could not have completely disappeared. There were forests in the immediate vicinity, forests producing timbers of great length and thickness; for in the north and northeast practically all the buildings had wooden roofs, wooden intermediate floors, and other features of wood. Costly buildings, such as temples and churches, employed large wooden beams; but wood was used in much larger quantities in private dwellings, shops, stables, and barns. If wood had not been plentiful and cheap—which means grown near by—the builders would have adopted the building methods of their neighbors in the south, who used very little wood and developed the most perfect type of lithic architecture the world has ever seen. And here there exists a strange anomaly: Northern Syria, where so much wood was employed in antiquity, is absolutely treeless now; while in the mountains of southern Syria, where wood must have been scarce in antiquity to have forced upon the inhabitants an almost exclusive use of stone, there are still groves of scrub oak and pine, and travelers of half a century ago reported large forests of chestnut trees.^[20] It is perfectly apparent that large parts of Syria once had soil and forests and springs and rivers, while it has none of these now, and that it had a much larger and better distributed rainfall in ancient times than it has now.

Professor Butler's careful work is especially interesting because of its contrast to the loose statements of those who believe in climatic uniformity. So far as I am aware, no opponent of the hypothesis of climatic changes has ever even attempted to show by careful statistical analysis that the ancient water supply of such ruins was no greater than that of the present. The most that has been done is to suggest that there may have been sources of water which are now unknown. Of course, this might be true in a single instance, but it could scarcely be the case in many hundreds or thousands of ruins.

Although the arguments in favor of a change of climate during the last two thousand years seem too strong to be ignored, their very strength seems to have been a source of error. A large number of people have jumped to the conclusion that the change which appears to have occurred in certain regions occurred everywhere, and that it consisted of a gradual desiccation.

Many observers, quite as careful as those who believe in progressive desiccation, point to evidences of aridity in past times in the very regions where the others find proof of moisture. Lakes such as the Caspian Sea fell to such a low level that parts of their present floors were exposed and were used as sites for buildings whose ruins are still extant. Elsewhere, for instance in the Tian-Shan Mountains, irrigation ditches are found in places where irrigation never seems to be necessary at present. In Syria and North Africa during the early centuries of the Christian era the Romans showed unparalleled activity in building great aqueducts and in watering land which then apparently needed water almost as much as it does today. Evidence of this sort is abundant and is as convincing as is the evidence of moister conditions in the past. It is admirably set forth, for example, in the comprehensive and ably written monograph of Leiter on the climate of North Africa.^[21] The evidence cited there and elsewhere has led many authors strongly to advocate the hypothesis of climatic uniformity. They have done exactly as have the advocates of progressive change, and have extended their conclusions over the whole world and over the whole of historic times.

The hypotheses of climatic uniformity and of progressive

change both seem to be based on reliable evidence. They may seem to be diametrically opposed to one another, but this is only when there is a failure to group the various lines of evidence according to their dates, and according to the types of climate in which they happen to be located. When the facts are properly grouped in both time and space, it appears that evidence of moist conditions in the historic Mediterranean lands is found during certain periods; for instance, four or five hundred years before Christ, at the time of Christ, and 1000 A. D. The other kind of evidence, on the contrary, culminates at other epochs, such as about 1200 B. C. and in the seventh and thirteenth centuries after Christ. It is also found during the interval from the culmination of a moist epoch to the culmination of a dry one, for at such times the climate was growing drier and the people were under stress. This was seemingly the case during the period from the second to the fourth centuries of our era. North Africa and Syria must then have been distinctly better watered than at present, as appears from Butler's vivid description; but they were gradually becoming drier, and the natural effect on a vigorous, competent people like the Romans was to cause them to construct numerous engineering works to provide the necessary water.

The considerations which have just been set forth have led to a third hypothesis, that of pulsatory climatic changes. According to this, the earth's climate is not stable, nor does it change uniformly in one direction. It appears to fluctuate back and forth not only in the little waves which we see from year to year or decade to decade, but in much larger waves, which take hundreds of years or even a thousand. These in turn seem to merge into and be imposed on the greater waves which form glacial stages, glacial epochs, and glacial periods. At the

present time there seems to be no way of determining whether the general tendency is toward aridity or toward glaciation. The seventh century of our era was apparently the driest time during the historic period—distinctly drier than the present—but the thirteenth century was almost equally dry, and the twelfth or thirteenth before Christ may have been very dry.

The best test of an hypothesis is actual measurements. In the case of the pulsatory hypothesis we are fortunately able to apply this test by means of trees. The growth of vegetation depends on many factors—soil, exposure, wind, sun, temperature, rain, and so forth. In a dry region the most critical factor in determining how a tree's growth shall vary from year to year is the supply of moisture during the few months of most rapid growth.^[22] The work of Douglass^{[23] ]} and others has shown that in Arizona and California the thickness of the annual rings affords a reliable indication of the amount of moisture available during the period of growth. This is especially true when the growth of several years is taken as the unit and is compared with the growth of a similar number of years before or after. Where a long series of years is used, it is necessary to make corrections to eliminate the effects of age, but this can be done by mathematical methods of considerable accuracy. It is difficult to determine whether the climate at the beginning

and end of a tree's life was the same, but it is easily possible to determine whether there have been pulsations while the tree was making its growth. If a large number of trees from various parts of a given district all formed thick rings at a certain period and then formed thin ones for a hundred years, after which the rings again become thick, we seem to be safe in concluding that the trees have lived through a long, dry period. The full reasons for this belief and details as to the methods of estimating climate from tree growth are given in The Climatic Factor.

The results set forth in that volume may be summarized as follows: During the years 1911 and 1912, under the auspices of the Carnegie Institution of Washington, measurements were made of the thickness of the rings of growth on the stumps of about 450 sequoia trees in California. These trees varied in age from 250 to nearly 3250 years. The great majority were over 1000 years of age, seventy-nine were over 2000 years, and three over 3000. Even where only a few trees are available the record is surprisingly reliable, except where occasional accidents occur. Where the number approximates 100, accidental variations are largely eliminated and we may accept the record with considerable confidence. Accordingly, we may say that in California we have a fairly accurate record of the climate for 2000 years and an approximate record for 1000 years more. The final results of the measurements of the California trees are shown in Fig. 4, where the climatic variations for 3000 years in California are indicated by the solid line. The high parts of the line indicate rainy conditions, the low parts, dry. An examination of this curve shows that during 3000 years there have apparently been climatic variations more important than any which have taken place during the past century. In order to bring out the

details more clearly, the more reliable part of the California curve, from 100 B. C. to the present time, has been reproduced in Fig. 5. This is identical with the corresponding part of Fig. 4, except that the vertical scale is three times as great.

Fig. 4. Changes of climate in California (solid line) and in western and central Asia (dotted line).

Note. The curves of Figs. 4 and 5 are reproduced as published in The Solar Hypothesis in 1914. Later work, however, has indicated that in the Asiatic curve the dash lines, which were tentatively inserted in 1914, are probably more nearly correct than the dotted lines. Still further evidence indicates that the Asiatic curve is nearly like that of California in its main features.

The curve of tree growth in California seems to be a true representation of the general features of climatic pulsations in the Mediterranean region. This conclusion was originally based on the resemblance between the solid line of Fig. 4, representing tree growth, and the dotted line representing changes of climate in the eastern Mediterranean region as inferred from the study of ruins and of history before any work on this subject had been done in America.^[24] The dotted line is here reproduced for its historical significance as a stage in the study of climatic changes. If it were to be redrawn today on the basis of the knowledge acquired in the last twelve years, it would be much more like the tree curve. For example, the period of aridity suggested by the dip of the dotted line about 300 A. D. was based largely on Professor Butler's data as to the paucity of inscriptions and ruins dating from that period in Syria. In the recent article, from which a long quotation has been given, he shows that later work proves that there is no such paucity. On the other hand, it has accentuated the marked and sudden decay in civilization and population which occurred shortly after 600 A. D. He reached the same conclusion to which the present authors had come on wholly different grounds, namely, that the dip in the dotted line about 300 A. D. is not warranted, whereas the dip about 630 A. D. is extremely important. In similar fashion the work of

Stein^[25] in central Asia makes it clear that the contrast between the water supply about 200 B. C. and in the preceding and following centuries was greater than was supposed on the basis of the scanty evidence available when the dotted line of Fig. 4 was drawn in 1910.

Fig. 5. Changes in California climate for 2000 years, as measured by growth of Sequoia trees.

Fig. 5 is the same as the later portion of Fig. 4, except that the vertical scale has been magnified threefold. It seems probable that the dotted line at the right is more nearly correct than the solid line. During the thirty years since the end of the curve the general tendency appears in general to have been somewhat upward.

Since the curve of the California trees is the only continuous and detailed record yet available for the climate of the last three thousand years, it deserves most careful study. It is especially necessary to determine the degree of accuracy with which the growth of the trees represents (1) the local rainfall and (2) the rainfall of remote regions such as Palestine. Perhaps the best way to determine these matters is the standard mathematical method of correlation coefficients. If two phenomena vary in perfect unison, as in the case of the turning of the wheels and the progress of an automobile when the brakes are not applied, the correlation coefficient is 1.00, being positive when the automobile goes forward and negative when it goes backward. If there is no relation between two phenomena, as in the case of the number of miles run by a given automobile each year and the number of chickens hatched in the same period, the coefficient is zero. A partial relationship where other factors enter into the matter is represented by a coefficient between zero and one, as in the case of the movement of the automobile and the consumption of gasoline. In this case the relation is very obvious, but is modified by other factors, including the roughness and grade of the road, the amount of traffic, the number of stops, the skill of the driver, the condition and load of the automobile, and the state of the weather. Such partial relationships are the kind for which correlation coefficients are most useful, for the size of the coefficients shows the relative importance

of the various factors. A correlation coefficient four times the probable error, which can always be determined by a formula well known to mathematicians, is generally considered to afford evidence of some kind of relation between two phenomena. When the ratio between coefficient and error rises to six, the relationship is regarded as strong.

Few people would question that there is a connection between tree growth and rainfall, especially in a climate with a long summer dry season like that of California. But the growth of the trees also depends on their position, the amount of shading, the temperature, insect pests, blights, the wind with its tendency to break the branches, and a number of other factors. Moreover, while rain commonly favors growth, great extremes are relatively less helpful than more moderate amounts. Again, the roots of a tree may tap such deep sources of water that neither drought nor excessive rain produces much effect for several years. Hence in comparing the growth of the huge sequoias with the rainfall we should expect a correlation coefficient high enough to be convincing, but decidedly below 1.00. Unfortunately there is no record of the rainfall where the sequoias grow, the nearest long record being that of Sacramento, nearly 200 miles to the northwest and close to sea level instead of at an altitude of about 6000 feet.

Applying the method of correlation coefficients to the annual rainfall of Sacramento and the growth of the sequoias from 1863 to 1910, we obtain the results shown in Table 3. The trees of Section A of the table grew in moderately dry locations although the soil was fairly deep, a condition which seems to be essential to sequoias. In this case, as in all the others, the rainfall is reckoned from July to June, which practically means from October to May, since there is almost no summer rain. Thus the tree growth in 1861 is compared with the rainfall of the preceding rainy season, 1860-1861, or of several preceding rainy seasons as the table indicates.

In the first line of Section A a correlation coefficient of only -0.056, which is scarcely six-tenths of the probable error, means that there is no appreciable relation between the rainfall of a given season and the growth during the following spring and summer. The roots of the sequoias probably penetrate so deeply that the rain and melted snow of the spring months do not sink down rapidly enough to influence the trees before the growing season comes to an end. The precipitation of two preceding seasons, however, has some effect on the trees, as appears in the second line of Section A, where the correlation coefficient is +0.288, or 3.2 times the probable error. When the rainfall of three seasons is taken into account the coefficient rises to +0.570, or 8.7 times the probable error, while with four years of rainfall the coefficient begins to fall off. Thus the growth of these eighteen sequoias on relatively dry slopes appears to have depended chiefly on the rainfall of the second and third preceding rainy seasons. The growth in 1900, for example, depended largely on the rainfall in the rainy seasons of 1897-1898 and 1898-1899.

Section B of the table shows that with 112 trees, growing chiefly in moist depressions where the water supply is at a maximum, the correlation between growth and rainfall, +0.577 for ten years' rainfall, is even higher than with the dry trees. The seepage of the underground water is so slow that not until four years' rainfall is taken into account is the correlation coefficient more than four times the probable error. When only the trees growing in moist locations are employed, the coefficient between

tree growth and the rainfall for ten years rises to the high figure of +0.605, or 9.8 times the probable error, as appears in Section C. These figures, as well as many others not here published, make it clear that the curve of sequoia growth from 1861 to 1910 affords a fairly close indication of the rainfall at Sacramento, provided allowance be made for a delay of three to ten years due to the fact that the moisture in the soil gradually seeps down the mountain-sides and only reaches the sequoias after a considerable interval.

If a rainfall record were available for the place where the trees actually grow, the relationship would probably be still closer.

The record at Fresno, for example, bears out this conclusion so far as it goes. But as Fresno lies at a low altitude and its rainfall is of essentially the Sacramento type, its short record is of less value than that of Sacramento. The only rainfall records among the Sierras at high levels, where the rainfall and temperature are approximately like those of the sequoia region, are found along the main line of the Southern Pacific railroad. This runs from Oakland northeastward seventy miles across the open plain to Sacramento, then another seventy miles, as the crow flies, through Colfax and over a high pass in the Sierras at Summit, next twenty miles or so down through Truckee to Boca, on the edge of the inland basin of Nevada, and on northeastward another 160 miles to Winnemucca, where it turns east toward Ogden and Salt Lake City. Section D of Table 3 shows the correlation coefficients between the rainfall along the railroad and the growth of the sequoias. At Sacramento, which lies fairly open to winds from the Pacific and thus represents the general climate of central California, the coefficient is nearly five times the probable error, thus indicating a

real relation to sequoia growth. Then among the foothills of the Sierras at Colfax, the coefficient drops till it is scarcely larger than the probable error. It rises rapidly, however, as one advances among the mountains, until at Boca it attains the high figure of +0.604 or eight times the probable error, and continues high in the dry area farther east. In other words the growth of the sequoias is a good indication of the rainfall where the trees grow and in the dry region farther east.

In order to determine the degree to which the sequoia record represents the rainfall of other regions, let us select Jerusalem for comparison. The reasons for this selection are that Jerusalem furnishes the only available record that satisfies the following necessary conditions: (1) its record is long enough to be important; (2) it is located fairly near the latitude of the sequoias, 32°N versus 37°N; (3) it is located in a similar type of climate with winter rains and a long dry summer; (4) it lies well above sea level (2500 feet) and somewhat back from the seacoast, thus approximating although by no means duplicating the condition of the sequoias; and (5) it lies in a region where the evidence of climatic changes during historic times is strongest. The ideal place for comparison would be the valley in which grow the cedars of Lebanon. Those trees resemble the sequoias to an extraordinary degree, not only in their location, but in their great age. Some day it will be most interesting to compare the growth of these two famous groups of old trees.

The correlation coefficients for the sequoia growth and the rainfall at Jerusalem are given in Section A, Table 4. They are so high and so consistent that they scarcely leave room for doubt that where a hundred or more sequoias are employed, as in Fig. 5, their curve of growth affords a good indication of the fluctuations of climate in western Asia. The high coefficient for the eleven trees measured by Douglass suggests that where the number of trees falls as low as ten, as in the part of Fig. 4 from 710 to 840 B. C., the relation between tree growth and rainfall is still close even when only one year's growth is considered. Where the unit is ten years of growth, as in Figs. 4 and 5, the accuracy of the tree curve as a measure of rainfall is much greater than when a single year is used as in Table 4. When the unit is raised to thirty years, as in the smoothed part of Fig. 4 previous to 240 B. C., even four trees, as from 960 to 1070, probably give a fair approximation to the general changes in rainfall, while a single tree prior to 1110 B. C. gives a rough indication.

Table 4 shows a peculiar feature in the fact that the correlations of Section A between tree growth and the rainfall of Jerusalem are decidedly higher than those between the rainfall in the two regions. Only at Sacramento and Boca are the rainfall coefficients high enough to be conclusive. This, however, is not surprising, for even between Sacramento and San Bernardino, only 400 miles apart, the correlation coefficient for the rainfall by three-year periods is only 10.7 times the probable error, as appears in Section C of Table 4, while between San Bernardino and Winnemucca 500 miles away, the corresponding figure drops to 2.8. It must be remembered that in some respects the growth of the sequoias is a much better record of rainfall than are the records kept by man. The human record is based on the amount of water caught by a little gauge a few inches in diameter. Every gust of wind detracts from the accuracy of the record; a mile away the rainfall may be double what it is at the gauge. Each sequoia, on the other hand, draws its moisture from an area thousands of times as large as

a rain gauge. Moreover, the trees on which Figs. 4 and 5 are based were scattered over an area fifty miles long and several hundred square miles in extent. Hence they represent the summation of the rainfall over an area millions of times as large as that of a rain gauge. This fact and the large correlation coefficients between sequoia growth and Jerusalem rainfall should be considered in connection with the fact that all the coefficients between the rainfall of California and Nevada and that of Jerusalem are positive. If full records of the complete rainfall of California and Nevada on the one hand and of the eastern Mediterranean region on the other were available for a long period, they would probably agree closely.

Just how widely the sequoias can be used as a measure of the climate of the past is not yet certain. In some regions, as will shortly be explained, the climatic changes seem to have been of an opposite character from those of California. In others the Californian or eastern Mediterranean type of change seems sometimes to prevail but is not always evident. For example, at Malta the rainfall today shows a distinct relation to that of Jerusalem and to the growth of the sequoias. But the correlation coefficient between the rainfall of eight-year periods at Naples, a little farther north, and the growth of the sequoias at the end of the periods is -0.132, or only 1.4 times the probable error and much too small to be significant. This is in harmony with the fact that although Naples has summer droughts, they are not so pronounced as in California and Palestine, and the prevalence of storms is much greater. Jerusalem receives only 8 per cent of its rain in the seven months from April to October, and Sacramento 13, while Malta receives 31 per cent and Naples 43. Nevertheless, there is some evidence that in the past the climatic fluctuations of southern Italy followed

nearly the same course as those of California and Palestine. This apparent discrepancy seems to be explained by our previous conclusion that changes of climate are due largely to a shifting of storm tracks. When sunspots are numerous the storms which now prevail in northern Italy seem to be shifted southward and traverse the Mediterranean to Palestine just as similar storms are shifted southward in the United States. This perhaps accounts for the agreement between the sequoia curve and the agricultural and social history of Rome from about 400 B. C. to 100 A. D., as explained in World Power and Evolution. For our present purposes, however, the main point is that since rainfall records have been kept the fluctuations of climate indicated by the growth of the sequoias have agreed closely with fluctuations in the rainfall of the eastern Mediterranean region. Presumably the same was true in the past. In that case, the sequoia curve not only is a good indication of climatic changes or pulsations in regions of similar climate, but may serve as a guide to coincident but different changes in regions of other types.

An enormous body of other evidence points to the same conclusion. It indicates that while the average climate of the present is drier than that of the past in regions having the Mediterranean type of winter rains and summer droughts, there have been pronounced pulsations during historic times so that at certain times there has actually been greater aridity than at present. This conclusion is so important that it seems advisable to examine the only important arguments that have been raised against it, especially against the idea that the general rainfall of the eastern Mediterranean was greater in the historic past than at present. The first objection is the unquestionable fact that droughts and famines have

occurred at periods which seem on other evidence to have been moister than the present. This argument has been much used, but it seems to have little force. If the rainfall of a given region averages thirty inches and varies from fifteen to forty-five, a famine will ensue if the rainfall drops for a few years to the lower limit and does not rise much above twenty for a few years. If the climate of the place changes during the course of centuries, so that the rainfall averages only twenty inches, and ranges from seven to thirty-five, famine will again ensue if the rainfall remains near ten inches for a few years. The ravages of the first famine might be as bad as those of the second. They might even be worse, because when the rainfall is larger the population is likely to be greater and the distress due to scarcity of food would affect a larger number of people. Hence historic records of famines and droughts do not indicate that the climate was either drier or moister than at present. They merely show that at the time in question the climate was drier than the normal for that particular period.

The second objection is that deserts existed in the past much as at present. This is not a real objection, however, for, as we shall see more fully, some parts of the world suffer one kind of change and others quite the opposite. Moreover, deserts have always existed, and when we talk of a change in their climate we merely mean that their boundaries have shifted. A concrete example of the mistaken use of ancient dryness as proof of climatic uniformity is illustrated by the march of Alexander from India to Mesopotamia. Hedin gives an excellent presentation of the case in the second volume of his Overland to India. He shows conclusively that Alexander's army suffered terribly from lack of water and provisions. This certainly proves that the climate was dry, but it by no

means indicates that there has been no change from the past to the present. We do not know whether Alexander's march took place during an especially dry or an especially wet year. In a desert region like Makran, in southern Persia and Beluchistan, where the chief difficulties occurred, the rainfall varies greatly from year to year. We have no records from Makran, but the conditions there are closely similar to those of southern Arizona and New Mexico. In 1885 and 1905 the rainfall for five stations in that region was as follows:

These stations are distributed over an area nearly 500 miles east and west. Manifestly a traveler who spent the year 1885 in that region would have had much more difficulty in finding water and forage than one who traveled in the same places in 1905. During 1885 the rainfall was 42 per cent less than the average, and during 1905 it was 134 per cent more than the average. Let us suppose, for the sake of argument, that the average rainfall of southeastern Persia is six inches today and was ten inches in the days of Alexander. If the rainfall from year to year varied as much in the past in Persia as it does now in New Mexico and Arizona, the rainfall during an ancient

dry year, corresponding in character to 1885, would have been about 5.75 inches. On the other hand, if we suppose that the rainfall then averaged less than at present,—let us say four inches,—a wet year corresponding to 1905 in the American deserts might have had a rainfall of about ten inches. This being the case, it is clear that our estimate of what Alexander's march shows as to climate must depend largely on whether 325 B. C. was a wet year or a dry year. Inasmuch as we know nothing about this, we must fall back on the fact that a large army accomplished a journey in a place where today even a small caravan usually finds great difficulty in procuring forage and water. Moreover, elephants were taken 180 miles across what is now an almost waterless desert, and yet the old historians make no comment on such a feat which today would be practically impossible. These things seem more in harmony with a change of climate than with uniformity. Nevertheless, it is not safe to place much reliance on them except when they are taken in conjunction with other evidence, such as the numerous ruins, which show that Makran was once far more densely populated than now seems possible. Taken by itself, such incidents as Alexander's march cannot safely be used either as an argument for or against changes of climate.

The third and strongest objection to any hypothesis of climatic changes during historic times is based on vegetation. The whole question is admirably set forth by J. W. Gregory,^[29] who gives not only his own results, but those of the ablest scholars who have preceded him. His conclusions are important because they represent one of the few cases where a definite statistical attempt has been made to prove the exact condition of the climate of the

past. After stating various less important reasons for believing that the climate of Palestine has not changed, he discusses vegetation. The following quotation indicates his line of thought. A sentence near the beginning is italicized in order to call attention to the importance which Gregory and others lay on this particular kind of evidence:

Some more certain test is necessary than the general conclusions which can be based upon the historical and geographical evidence of the Bible. In the absence of rain gauge and thermometric records, the most precise test of climate is given by the vegetation; and fortunately the palm affords a very delicate test of the past climate of Palestine and the eastern Mediterranean.... The date palm has three limits of growth which are determined by temperature; thus it does not reach full maturity or produce ripe fruit of good quality below the mean annual temperature of 69°F. The isothermal of 69° crosses southern Algeria near Biskra; it touches the northern coasts of Cyrenaica near Derna and passes Egypt near the mouth of the Nile, and then bends northward along the coast lands of Palestine.

To the north of this line the date palm grows and produces fruit, which only ripens occasionally, and its quality deteriorates as the temperature falls below 69°. Between the isotherms of 68° and 64°, limits which include northern Algeria, most of Sicily, Malta, the southern parts of Greece and northern Syria, the dates produced are so unripe that they are not edible. In the next cooler zone, north of the isotherm of 62°, which enters Europe in southwestern Portugal, passes through Sardinia, enters Italy near Naples, crosses northern Greece and Asia Minor to the east of Smyrna, the date palm is grown only for its foliage, since it does not fruit.

Hence at Benghazi, on the north African coast, the date palm is fertile, but produces fruit of poor quality. In Sicily and at Algiers the fruit ripens occasionally and at Rome and Nice the palm is grown only as an ornamental tree.

The date palm therefore affords a test of variations in mean annual temperature of three grades between 62° and 69°.

This test shows that the mean annual temperature of Palestine has not altered since Old Testament times. The palm tree now grows dates on the coast of Palestine and in the deep depression around the Dead Sea, but it does not produce fruit on the highlands of Judea. Its distribution in ancient times, as far as we can judge from the Bible, was exactly the same. It grew at "Jericho, the city of palm trees" (Deut. xxxiv: 3 and 2 Chron. xxviii: 15), and at Engedi, on the western shore of the Dead Sea (2 Chron. xx: 2; Sirach xxiv: 14); and though the palm does not still live at Jericho—the last apparently died in 1838—its disappearance must be due to neglect, for the only climatic change that would explain it would be an increase in cold or moisture. In olden times the date palm certainly grew on the highlands of Palestine; but apparently it never produced fruit there, for the Bible references to the palm are to its beauty and erect growth: "The righteous shall flourish like the palm" (Ps. xcii: 12); "They are upright as the palm tree" (Jer. x: 5); "Thy stature is like to a palm tree" (Cant. vii: 7). It is used as a symbol of victory (Rev. vii: 9), but never praised as a source of food.

Dates are not once referred to in the text of the Bible, but according to the marginal notes the word translated "honey" in 2 Chron. xxxi: 5 may mean dates....

It appears, therefore, that the date palm had essentially the same distribution in Palestine in Old Testament times as it has now; and hence we may infer that the mean temperature was then the same as now. If the climate had been moister and cooler, the date could not have flourished at Jericho. If it had been warmer, the palms would have grown freely at higher levels and Jericho would not have held its distinction as the city of palm trees.^[30]

In the main Gregory's conclusions seem to be well grounded, although even according to his data a change

of 2° or 3° in mean temperature would be perfectly feasible. It will be noticed, however, that they apply to temperature and not to rainfall. They merely prove that two thousand years ago the mean temperature of Palestine and the neighboring regions was not appreciably different from what it is today. This, however, is in no sense out of harmony with the hypothesis of climatic pulsations. Students of glaciation believe that during the last glacial epoch the mean temperature of the earth as a whole was only 5° or 6°C. lower than at present. If the difference between the climate of today and of the time of Christ is a tenth as great as the difference between the climate of today and that which prevailed at the culmination of the last glacial epoch, the change in two thousand years has been of large dimensions. Yet this would require a rise of only half a degree Centigrade in the mean temperature of Palestine. Manifestly, so slight a change would scarcely be detectable in the vegetation.

The slightness of changes in mean temperature as compared with changes in rainfall may be judged from a comparison of wet and dry years in various regions. For example, at Berlin between 1866 and 1905 the ten most rainy years had an average precipitation of 670 mm. and a mean temperature of 9.15°C. On the other hand, the ten years of least rainfall had an average of 483 mm. and a mean temperature of 9.35°. In other words, a difference of 137 mm., or 39 per cent, in rainfall was accompanied by a difference of only 0.2°C. in temperature. Such contrasts between the variability of mean rainfall and mean temperature are observable not only when individual years are selected, but when much longer periods are taken. For instance, in the western Gulf region of the United States the two inland stations of Vicksburg, Mississippi, and Shreveport, Louisiana, and the two maritime

stations of New Orleans, Louisiana, and Galveston, Texas, lie at the margins of an area about 400 miles long. During the ten years from 1875 to 1884 their rainfall averaged 59.4 inches,^[31] while during the ten years from 1890 to 1899 it averaged only 42.4 inches. Even in a region so well watered as the Gulf States, such a change—40 per cent more in the first decade than in the second—is important, and in drier regions it would have a great effect on habitability. Yet in spite of the magnitude of the change the mean temperature was not appreciably different, the average for the four stations being 67.36°F. during the more rainy decade and 66.94°F. during the less rainy decade—a difference of only 0.42°F. It is worth noticing that in this case the wetter period was also the warmer, whereas in Berlin it was the cooler. This is probably because a large part of the moisture of the Gulf States is brought by winds having a southerly component. Similar relationships are apparent in other places. We select Jerusalem because we have been discussing Palestine. At the time of writing, the data available in the Quarterly Journal of the Palestine Exploration Fund cover the years from 1882-1899 and 1903-1909. Among these twenty-five years the thirteen which had most rain had an average of 34.1 inches and a temperature of 62.04°F. The twelve with least rain had 24.4 inches and a temperature of 62.44°. A difference of 40 per cent in rainfall was accompanied by a difference of only 0.4°F. in temperature.

The facts set forth in the preceding paragraphs seem to show that extensive changes in precipitation and storminess can take place without appreciable changes of mean temperature. If such changed conditions can persist

for ten years, as in one of our examples, there is no logical reason why they cannot persist for a hundred or a thousand. The evidence of changes in climate during the historic period seems to suggest changes in precipitation much more than in temperature. Hence the strongest of all the arguments against historic changes of climate seems to be of relatively little weight, and the pulsatory hypothesis seems to be in accord with all the known facts.

Before the true nature of climatic changes, whether historic or geologic, can be rightly understood, another point needs emphasis. When the pulsatory hypothesis was first framed, it fell into the same error as the hypotheses of uniformity and of progressive change—that is, the assumption was made that the whole world is either growing drier or moister with each pulsation. A study of the ruins of Yucatan, in 1912, and of Guatemala, in 1913, as is explained in The Climatic Factor, has led to the conclusion that the climate of those regions has changed in the opposite way from the changes which appear to have taken place in the desert regions farther south. These Maya ruins in Central America are in many cases located in regions of such heavy rainfall, such dense forests, and such malignant fevers that habitation is now practically impossible. The land cannot be cultivated except in especially favorable places. The people are terribly weakened by disease and are among the lowest in Central America. Only a hundred miles from the unhealthful forests we find healthful areas, such as the coasts of Yucatan and the plateau of Guatemala. Here the vast majority of the population is gathered, the large towns are located, and the only progressive people are found. Nevertheless, in the past the region of the forests was the home of by far the most progressive people who are ever known to have lived in America previous to the

days of Columbus. They alone brought to high perfection the art of sculpture; they were the only American people who invented the art of writing. It seems scarcely credible that such a people would have lived in the worst possible habitat when far more favored regions were close at hand. Therefore it seems as if the climate of eastern Guatemala and Yucatan must have been relatively dry at some past time. The Maya chronology and traditions indicate that this was probably at the same time when moister conditions apparently prevailed in the subarid or desert portions of the United States and Asia. Fig. 3 shows that today at times of many sunspots there is a similar opposition between a tendency toward storminess and rain in subtropical regions and toward aridity in low latitudes near the heat equator.

Thus our final conclusion is that during historic times there have been pulsatory changes of climate. These changes have been of the same type in regions having similar kinds of climate, but of different and sometimes opposite types in places having diverse climates. As to the cause of the pulsations, they cannot have been due to the precession of the equinoxes nor apparently to any allied astronomical cause, for the time intervals are too short and too irregular. They cannot have been due to changes in the percentage of carbon dioxide in the atmosphere, for not even the strongest believers in the climatic efficacy of that gas hold that its amount could fluctuate in any such violent way as would be necessary to explain the pulsations shown in the California curve of tree growth. Volcanic activity seems more probable as at least a partial cause, and it would be worth while to investigate the matter more fully. Nevertheless, it can apparently be only a minor cause. In the first place, the main effect of a cloud of dust is to alter the temperature, but

Gregory's summary of the palm and the vine shows that variations in temperature are apparently of very slight importance during historic times. Again, ruins on the bottoms of enclosed salt lakes, old beaches now under the water, and signs of irrigation ditches where none are now needed indicate a climate drier than the present. Volcanic dust, however, cannot account for such a condition, for at present the air seems to be practically free from such dust for long periods. Thus we now experience the greatest extreme which the volcanic hypothesis permits in one direction, but there have been greater extremes in the same direction. The thermal solar hypothesis is likewise unable to explain the observed phenomena, for neither it nor the volcanic hypothesis offers any explanation of why the climate varies in one way in Mediterranean climates and in an opposite way in regions near the heat equator.

[CHAPTER VI]

THE CLIMATIC STRESS OF THE FOURTEENTH CENTURY

In order to give concreteness to our picture of the climatic pulsations of historic times let us take a specific period and see how its changes of climate were distributed over the globe and how they are related to the little changes which now take place in the sunspot cycle. We will take the fourteenth century of the Christian era, especially the first half. This period is chosen because it is the last and hence the best known of the times when the climate of the earth seems to have taken a considerable swing toward the conditions which now prevail when the sun is most active, and which, if intensified, would apparently lead to glaciation. It has already been discussed in World Power and Evolution, but its importance and the fact that new evidence is constantly coming to light warrant a fuller discussion.

To begin with Europe; according to the careful account of Pettersson^[32] the fourteenth century shows

a record of extreme climatic variations. In the cold winters the rivers Rhine, Danube, Thames, and Po were frozen for weeks and months. On these cold winters there followed violent floods, so that the rivers mentioned inundated their valleys. Such floods are recorded in 55 summers in the 14th century. There is, of course, nothing astonishing in the fact that the inundations of the great rivers of Europe were more devastating 600 to 700 years ago than in our days, when the flow of the rivers has been regulated by canals, locks, etc.; but still the inundations in the 13th and 14th centuries must have surpassed everything of that kind which has occurred since then. In 1342 the waters of the Rhine rose so high that they inundated the city of Mayence and the Cathedral "usque ad cingulum hominis." The walls of Cologne were flooded so that they could be passed by boats in July. This occurred also in 1374 in the midst of the month of February, which is of course an unusual season for disasters of the kind. Again in other years the drought was so intense that the same rivers, the Danube, Rhine, and others, nearly dried up, and the Rhine could be forded at Cologne. This happened at least twice in the same century. There is one exceptional summer of such evil record that centuries afterwards it was spoken of as "the old hot summer of 1357."

Pettersson goes on to speak of two oceanic phenomena on which the old chronicles lay greater stress than on all others:

The first [is] the great storm-floods on the coast of the North Sea and the Baltic, which occurred so frequently that not less than nineteen floods of a destructiveness unparalleled in later times are recorded from the 14th century. The coastline of the North Sea was completely altered by these floods. Thus on January 16, 1300, half of the island Heligoland and many other islands were engulfed by the sea. The same fate overtook the island of Borkum, torn into several islands by the storm-flood of January 16, which remoulded the Frisian Islands into their present shape, when also Wendingstadt, on the island of Sylt, and Thiryu parishes were engulfed. This flood is known under the name of "the great man-drowning." The coasts of the Baltic also were exposed to storm-floods of unparalleled violence. On November 1, 1304, the island of Ruden was torn asunder from Rugen by the force of the waves. Time does not allow me to dwell upon individual disasters of this kind, but it will be well to note that of the nineteen great floods on record eighteen occurred in the cold season between the autumnal and vernal equinoxes.

The second remarkable phenomenon mentioned by the chronicles is the freezing of the entire Baltic, which occurred many times during the cold winters of these centuries. On such occasions it was possible to travel with carriages over the ice from Sweden to Bornholm and from Denmark to the German coast (Lubeck), and in some cases even from Gotland to the coast of Estland.

Norlind^[33] says that "the only authentic accounts" of the complete freezing of the Baltic in the neighborhood of the Kattegat are in the years 1296, 1306, 1323, and 1408. Of these 1296 is "much the most uncertain," while 1323 was the coldest year ever recorded, as appears from the fact that horses and sleighs crossed regularly from Sweden to Germany on the ice.

Not only central Europe and the shores of the North Sea were marked by climatic stress during the fourteenth century, but Scandinavia also suffered. As Pettersson puts it:

On examining the historic (data) from the last centuries of the Middle Ages, Dr. Bull of Christiania has come to the conclusion that the decay of the Norwegian kingdom was not so much a consequence of the political conditions at that time, as of the frequent failures of the harvest so that corn [wheat] for bread had to be imported from Lübeck, Rostock, Wismar and so forth. The Hansa Union undertook the importation and obtained political power by its economic influence. The Norwegian land-owners were forced to lower their rents. The population decreased and became impoverished. The revenue sank 60 to 70 per cent. Even the income from Church property decreased. In 1367 corn was imported from Lübeck to a value of one-half million kroner. The trade balance inclined to the disadvantage of Norway whose sole article of export at that time was dried fish. (The production of fish increased enormously in the Baltic regions off south Sweden because of the same changes which were influencing the lands, but this did not benefit Norway.) Dr. Bull draws a comparison with the conditions described in the Sagas when Nordland [at the Arctic Circle] produced enough corn to feed the inhabitants of the country. At the time of Asbjörn Selsbane the chieftains in Trondhenäs [still farther north in latitude 69°] grew so much corn that they did not need to go southward to buy corn unless three successive years of dearth had occurred. The province of Trondheim exported wheat to Iceland and so forth. Probably the turbulent political state of Scandinavia at the end of the Middle Ages was in a great measure due to unfavorable climatic conditions, which lowered the standard of life, and not entirely to misgovernment and political strife as has hitherto been taken for granted.

During this same unfortunate first half of the fourteenth century England also suffered from conditions which, if sufficiently intensified, might be those of a glacial period. According to Thorwald Rogers^[34] the severest famine ever experienced in England was that of 1315-1316, and the next worst was in 1321. In fact, from 1308 to 1322 great scarcity of food prevailed most of the time. Other famines of less severity occurred in 1351 and 1369. "The same cause was at work in all these cases," says Rogers, "incessant rain, and cold, stormy summers. It is said that the inclemency of the seasons affected the cattle, and that numbers perished from disease and want." After the bad harvest of 1315 the price of wheat, which was already high, rose rapidly, and in May, 1316, was about five times the average. For a year or more thereafter it remained at three or four times the ordinary

level. The severity of the famine may be judged from the fact that previous to the Great War the most notable scarcity of wheat in modern England and the highest relative price was in December, 1800. At that time wheat cost nearly three times the usual amount, instead of five as in 1316. During the famine of the early fourteenth century "it is said that people were reduced to subsist upon roots, upon horses and dogs, and stories are told of even more terrible acts by reason of the extreme famine." The number of deaths was so great that the price of labor suffered a permanent rise of at least 10 per cent. There simply were not people enough left among the peasants to do the work demanded by the more prosperous class who had not suffered so much.

After the famine came drought. The year 1325 appears to have been peculiarly dry, and 1331, 1344, 1362, 1374, and 1377 were also dry. In general these conditions do little harm in England. They are of interest chiefly as showing how excessive rain and drought are apt to succeed one another.

These facts regarding northern and central Europe during the fourteenth century are particularly significant when compared with the conclusions which we have drawn in Earth and Sun from the growth of trees in Germany and from the distribution of storms. A careful study of all the facts shows that we are dealing with two distinct types of phenomena. In the first place, the climate of central Europe seems to have been peculiarly continental during the fourteenth century. The winters were so cold that the rivers froze, and the summers were so wet that there were floods every other year or oftener. This seems to be merely an intensification of the conditions which prevail at the present time during periods of many sunspots, as indicated by the growth of trees at

Eberswalde in Germany and by the number of storms in winter as compared with summer. The prevalence of droughts, especially in the spring, is also not inconsistent with the existence of floods at other seasons, for one of the chief characteristics of a continental climate is that the variations from one season to another are more marked than in oceanic climates. Even the summer droughts are typically continental, for when continental conditions prevail, the difference between the same season in different years is extreme, as is well illustrated in Kansas. It must always be remembered that what causes famine is not so much absolute dryness as a temporary diminution of the rainfall.

The second type of phenomena is peculiarly oceanic in character. It consists of two parts, both of which are precisely what would be expected if a highly continental climate prevailed over the land. In the first place, at certain times the cold area of high pressure, which is the predominating characteristic of a continent during the winter, apparently spread out over the neighboring oceans. Under such conditions an inland sea, such as the Baltic, would be frozen, so that horses could cross the ice even in the Far West. In the second place, because of the unusually high pressure over the continent, the barometric gradients apparently became intensified. Hence at the margin of the continental high-pressure area the winds were unusually strong and the storms of corresponding severity. Some of these storms may have passed entirely along oceanic tracks, while others invaded the borders of the land, and gave rise to the floods and to the wearing away of the coast described by Pettersson.

Turning now to the east of Europe, Brückner's^[35] study

of the Caspian Sea shows that that region as well as western Europe was subject to great climatic vicissitudes in the first half of the fourteenth century. In 1306-1307 the Caspian Sea, after rising rapidly for several years, stood thirty-seven feet above the present level and it probably rose still higher during the succeeding decades. At least it remained at a high level, for Hamdulla, the Persian, tells us that in 1325 a place called Aboskun was under water.^[36]

Still further east the inland lake of Lop Nor also rose at about this time. According to a Chinese account the Dragon Town on the shore of Lop Nor was destroyed by a flood. From Himley's translation it appears that the level of the lake rose so as to overwhelm the city completely. This would necessitate the expansion of the lake to a point eighty miles east of Lulan, and fully fifty from the present eastern end of the Kara Koshun marsh. The water would have to rise nearly, or quite, to a strand which is now clearly visible at a height of twelve feet above the modern lake or marsh.

In India the fourteenth century was characterized by what appears to have been the most disastrous drought in all history. Apparently the decrease in rainfall here was as striking as the increase in other parts of the world. No statistics are available but we are told that in the great famine which began in 1344 even the Mogul emperor was unable to obtain the necessaries of life for his household. No rain worth mentioning fell for years. In some places the famine lasted three or four years, and in some twelve, and entire cities were left without an inhabitant. In a later famine, 1769-1770, which occurred in Bengal shortly after the foundation of British rule in

India, but while the native officials were still in power, a third of the population, or ten out of thirty millions, perished. The famine in the first half of the fourteenth century seems to have been far worse. These Indian famines were apparently due to weak summer monsoons caused presumably by the failure of central Asia to warm up as much as usual. The heavier snowfall, and the greater cloudiness of the summer there, which probably accompanied increased storminess, may have been the reason.

The New World as well as the Old appears to have been in a state of climatic stress during the first half of the fourteenth century. According to Pettersson, Greenland furnishes an example of this. At first the inhabitants of that northland were fairly prosperous and were able to approach from Iceland without much hindrance from the ice. Today the North Atlantic Ocean northeast of Iceland is full of drift ice much of the time. The border of the ice varies from season to season, but in general it extends westward from Iceland not far from the Arctic circle and then follows the coast of Greenland southward to Cape Farewell at the southern tip and around to the western side for fifty miles or more. Except under exceptional circumstances a ship cannot approach the coast until well northward on the comparatively ice-free west coast. In the old Sagas, however, nothing is said of ice in this region. The route from Iceland to Greenland is carefully described. In the earliest times it went from Iceland a trifle north of west so as to approach the coast of Greenland after as short an ocean passage as possible. Then it went down the coast in a region where approach is now practically impossible because of the ice. At that time this coast was icy close to the shore, but there is no sign that navigation was rendered difficult as is now the

case. Today no navigator would think of keeping close inland. The old route also went north of the island on which Cape Farewell is located, although the narrow channel between the island and the mainland is now so blocked with ice that no modern vessel has ever penetrated it. By the thirteenth century, however, there appears to have been a change. In the Kungaspegel or Kings' Mirror, written at that time, navigators are warned not to make the east coast too soon on account of ice, but no new route is recommended in the neighborhood of Cape Farewell or elsewhere. Finally, however, at the end of the fourteenth century, nearly 150 years after the Kungaspegel, the old sailing route was abandoned, and ships from Iceland sailed directly southwest to avoid the ice. As Pettersson says:

... At the end of the thirteenth and the beginning of the fourteenth century the European civilization in Greenland was wiped out by an invasion of the aboriginal population. The colonists in the Vesterbygd were driven from their homes and probably migrated to America leaving behind their cattle in the fields. So they were found by Ivar Bardsson, steward to the Bishop of Gardar, in his official journey thither in 1342.

The Eskimo invasion must not be regarded as a common raid. It was the transmigration of a people, and like other big movements of this kind [was] impelled by altered conditions of nature, in this case the alterations of climate caused by [or which caused?] the advance of the ice. For their hunting and fishing the Eskimos require an at least partially open arctic sea. The seal, their principal prey, cannot live where the surface of the sea is entirely frozen over. The cause of the favorable conditions in the Viking-age was, according to my hypothesis, that the ice then melted at a higher latitude in the arctic seas.

The Eskimos then lived further north in Greenland and North America. When the climate deteriorated and the sea which gave them their living was closed by ice the Eskimos had to find a more suitable neighborhood. This they found in the land colonized by the Norsemen whom they attacked and finally annihilated.

Finally, far to the south in Yucatan the ancient Maya civilization made its last flickering effort at about this time. Not much is known of this but in earlier periods the history of the Mayas seems to have agreed quite closely with the fluctuations in climate.^[37] Among the Mayas, as we have seen, relatively dry periods were the times of greatest progress.

Let us turn now to Fig. 3 once more and compare the climatic conditions of the fourteenth century with those of periods of increasing rainfall. Southern England, Ireland, and Scandinavia, where the crops were ruined by extensive rain and storms in summer, are places where storminess and rainfall now increase when sunspots are numerous. Central Europe and the coasts of the North Sea, where flood and drought alternated, are regions which now have relatively less rain when sunspots increase than when they diminish. However, as appears from the trees measured by Douglass, the winters become more continental and hence cooler, thus corresponding to the cold winters of the fourteenth century when people walked on the ice from Scandinavia to Denmark. When such high pressure prevails in the winter, the total rainfall is diminished, but nevertheless the storms are more severe than usual, especially in the spring. In southeastern Europe, the part of the area whence the Caspian derives its water, appears to have less rainfall during times of increasing sunspots than when sunspots are few, but in an equally large area to the south, where the mountains

are higher and the run-off of the rain is more rapid, the reverse is the case. This seems to mean that a slight diminution in the water poured in by the Volga would be more than compensated by the water derived from Persia and from the Oxus and Jaxartes rivers, which in the fourteenth century appear to have filled the Sea of Aral and overflowed in a large stream to the Caspian. Still farther east in central Asia, so far as the records go, most of the country receives more rain when sunspots are many than when they are few, which would agree with what happened when the Dragon Town was inundated. In India, on the contrary, there is a large area where the rainfall diminishes at times of many sunspots, thus agreeing with the terrible famine from which the Moguls suffered so severely. In the western hemisphere, Greenland, Arizona, and California are all parts of the area where the rain increases with many sunspots, while Yucatan seems to lie in an area of the opposite type. Thus all the evidence seems to show that at times of climatic stress, such as the fourteenth century, the conditions are essentially the same as those which now prevail at times of increasing sunspots.

As to the number of sunspots, there is little evidence previous to about 1750. Yet that little is both interesting and important. Although sunspots have been observed with care in Europe only a little more than three centuries, the Chinese have records which go back nearly to the beginning of the Christian era. Of course the records are far from perfect, for the work was done by individuals and not by any great organization which continued the same methods from generation to generation. The mere fact that a good observer happened to use his smoked glass to advantage may cause a particular period to appear to have an unusual number of spots. On the

other hand, the fact that such an observer finds spots at some times and not at others tends to give a valuable check on his results, as does the comparison of one observer's work with that of another. Hence, in spite of many and obvious defects, most students of the problem agree that the Chinese record possesses much value, and that for a thousand years or more it gives a fairly true idea of the general aspect of the sun. In the Chinese records the years with many spots fall in groups, as would be expected, and are sometimes separated by long intervals. Certain centuries appear to have been marked by unusual spottedness. The most conspicuous of these is the fourteenth, when the years 1370 to 1385 were particularly noteworthy, for spots large enough to be visible to the naked eye covered the sun much of the time. Hence Wolf,^[38] who has made an exhaustive study of the matter, concludes that there was an absolute maximum of spots about 1372. While this date is avowedly open to question, the great abundance of sunspots at that time makes it probable that it cannot be far wrong. If this is so, it seems that the great climatic disturbances of which we have seen evidence in the fourteenth century occurred at a time when sunspots were increasing, or at least when solar activity was under some profoundly disturbing influence. Thus the evidence seems to show not merely that the climate of historic times has been subject to important pulsations, but that those pulsations were magnifications of the little climatic changes which now take place in sunspot cycles. The past and the present are apparently a unit except as to the intensity of the changes.

[CHAPTER VII]

GLACIATION ACCORDING TO THE SOLAR-CYCLONIC HYPOTHESIS^[39]

The remarkable phenomena of glacial periods afford perhaps the best available test to which any climatic hypothesis can be subjected. In this chapter and the two that follow, we shall apply this test. Since much more is known about the recent Great Ice Age, or Pleistocene glaciation, than about the more ancient glaciations, the problems of the Pleistocene will receive especial attention. In the present chapter the oncoming of glaciation and the subsequent disappearance of the ice will be outlined in the light of what would be expected according to the solar-cyclonic hypothesis. Then in the next chapter several problems of especial climatic significance will be considered, such as the localization of ice sheets, the succession of severe glacial and mild inter-glacial epochs, the sudden commencement of glaciation and the peculiar variations in the height of the snow line. Other topics to be considered are the occurrence of pluvial or rainy climates in non-glaciated regions, and glaciation near sea level in subtropical latitudes during the Permian and Proterozoic. Then in Chapter IX we shall consider the development and distribution of the remarkable deposits of wind-blown material known as loess.

Facts not considered at the time of framing an hypothesis

are especially significant in testing it. In this particular case, the cyclonic hypothesis was framed to explain the historic changes of climate revealed by a study of ruins, tree rings, and the terraces of streams and lakes, without special thought of glaciation or other geologic changes. Indeed, the hypothesis had reached nearly its present form before much attention was given to geological phases of the problem. Nevertheless, it appears to meet even this severe test.

According to the solar-cyclonic hypothesis, the Pleistocene glacial period was inaugurated at a time when certain terrestrial conditions tended to make the earth especially favorable for glaciation. How these conditions arose will be considered later. Here it is enough to state what they were. Chief among them was the fact that the continents stood unusually high and were unusually large. This, however, was not the primary cause of glaciation, for many of the areas which were soon to be glaciated were little above sea level. For example, it seems clear that New England stood less than a thousand feet higher than now. Indeed, Salisbury^[40] estimates that eastern North America in general stood not more than a few hundred feet higher than now, and W. B. Wright^[41] reaches the same conclusion in respect to the British Isles. Nevertheless, widespread lands, even if they are not all high, lead to climatic conditions which favor glaciation. For example, enlarged continents cause low temperature in high latitudes because they interfere with the ocean currents that carry heat polewards. Such continents also cause relatively cold winters, for lands cool much sooner than does the ocean. Another result is a

diminution of water vapor, not only because cold air cannot hold much vapor, but also because the oceanic area from which evaporation takes place is reduced by the emergence of the continents. Again, when the continents are extensive the amount of carbonic acid gas in the atmosphere probably decreases, for the augmented erosion due to uplift exposes much igneous rock to the air, and weathering consumes the atmospheric carbon dioxide. When the supply of water vapor and of atmospheric carbon dioxide is small, an extreme type of climate usually prevails. The combined result of all these conditions is that continental emergence causes the climate to be somewhat cool and to be marked by relatively great contrasts from season to season and from latitude to latitude.

When the terrestrial conditions thus permitted glaciation, unusual solar activity is supposed to have greatly increased the number and severity of storms and to have altered their location, just as now happens at times of many sunspots. If such a change in storminess had occurred when terrestrial conditions were unfavorable for glaciation, as, for example, when the lands were low and there were widespread epicontinental seas in middle and high latitudes, glaciation might not have resulted. In the Pleistocene, however, terrestrial conditions permitted glaciation, and therefore the supposed increase in storminess caused great ice sheets.