4 ' A lx ' I . t \ i .-v . m * ■» < . . .^ • 1 «'■- , »- * .• ' * ♦ • I .. M i: *;v->;a>> 4 --iwv-' m^. % U ■• ■ •: - .. .:.i.km%W > • ^ :'i^ A. ; •. * - ^ '■•*■ t* , ^ kit f 5 ^\ ' *4 ".' N. .' 2 \'i -''*•% f V ’* ^ . ■ i .'- A » 2 !‘ **’■ --' -X'A’.,'I ,. * ' Li t. * ^ ^ . rr- •« .'*.■•.•■* .t’» yVi>‘ • ■ V .* t " *j^ * ^ ^j?~ * * *' I 4a' V'Ll jqB& *> ’ 1 * i * kvfw$^:r‘ t' |w>' I •% V . •■'*, , '■ ‘ . . , , [jji ' ^ ' A i' '• ^‘ •*'"-■ ^r-- !'!fr_ ■*"''' . >■ .> ' '. ^ , THE CLIMATIC FACTOK AS ILLUSTEATED IE AEID AMERICA BY ELLSWORTH HUNTINGTON Assistant Professor of Geography at Yale University WITH CONTRIBUTIONS BY CHARLES SCHUCHERT, ANDREW E. DOUGLASS, AND CHARLES J. KULLMER WASHINGTON, D. C. Published by the Carnegie Institution of Washington 1914 CARNEGIE INSTITUTION OF WASHINGTON Publication No. 192 Copies of thi# Bask were first MAY I 5 19M PRESS OF THE NEW ERA PRrNTlNO COMPANY LANC.AST^.R. PA« TABLE OF CONTENTS. Chapter. Page. Introduction. 1-5 Part I. The Problem of Climatic Changes. I. The monsoon climate of Arizona and New Mexico. 9 II. The topographic influence of aridity. 15 III. The arboreal vegetation of the monsoon desert. 21 IV. The climatic theory of terraces. 23 V. The fluctuations of the Otero Soda Lake. 37 VI. The relation of alluvial terraces to man. 43 VII. The ancient people of southern Arizona. 47 VIII. Ruins in northern Sonora and southern New Mexico. 65 IX. The successive stages of culture in northern New Mexico. 75 X. Southern Mexico as a test case. 95 XI. A method of estimating rainfall by the growth of trees, by A. E. Douglass. 101 XII. The correction and comparison of curves of growth. 123 XIII. The curve of the big trees. 139 XIV. The interpretation of the curve of the Sequoia. 157 XV. The peninsula of Yucatan. 175 XVI. The shifting of climatic zones (including the Shift of the Storm Track, by Charles J. Kullmer). 189 XVII. Guatemala and the highest native American civilization.211 XVIII. Climatic changes and Maya history. 225 XIX. The solar hypothesis. 233 XX. Crustal deformation as the cause of climatic changes.255 Part II. Climates of Geologic Time. XXI. Climates of Geologic Time, by Charles Schuchert. 265 Part III. Tables. A. Average growth of 451 Sequoia trees in California, by decades and centuries, beginning with their youth ; basis of corrective factor for age. 301 B. Comparative growth of short-lived and long-Uved Sequoias by groups; factor for longevity. 301 C. List of individual Sequoia trees measured in California in 1911 and 1912. 302 D. Summary of Sequoia trees, by groups. 307 E. Combined corrective factors for age and longevity. Sequoia washingtoniana . 308 F. Growth of Sequoia washingtoniana by groups for each decade; uncorrected and corrected.311 G. Summary of growth of Sequoia washingtoniana, corrected and uncorrected, including Caspian factor.323 H. Summary of growth of trees measured by the United States Forest Service. 325 I. Average annual growth of Sequoias.328 J. Errors of ring counting in northern Arizona pines.330 Index. 331 iii LIST OF ILLUSTRATIONS. PLATES AND MAPS. Facing page Plate 1. 21 A. Alluvial deposits burying the bottoms of mesquite trees on the lower Santa Cruz near Charco Yuua. B. Typical vegetation of southern Arizona, giant cactus, Cholla, mesquite bushes, grease-wood, etc. C. Alluvial terraces and typical vegetation of a river valley in Northern Sonora. Plate 2.... . 46 A. Ruins of little stone terraces at Rincon Canyon. B. DefeMive Hohokam walls on a hilltop near San Xavier. C. Looking down from the top of the Trincheras of the Magdalena River. D. Site of an ancient village in Southern Arizona; metate and mani stones for grinding seeds. Plate 3. .... 83 A. Ruins of Tyuonyd in the Canyon de los Frijoles. B. Ruins of Pueblo Bonita in Chaco Canyon. Plate 4. 104 The cross identification of rings of growth. Plate 5. 139 A. The Boule tree, a Sequoia probably 2,500 years old. B. Young and middle-aged Sequoia in a valley bottom. Plate 6. 145 The dying out of rings in a young Sequoia at Dillonwood. Plate?. 146 The effect of injuries upon a young Sequoia. Plates. 183 A. A market-place in Yucatan, showing the best modern architecture. B. A typical house in Yucatan. C. Archway at Labna. D. Farmer’s hut in the midst of Labna. E. Ruins of Chac-multun. Plate 9. 189 A. Carved head at Baul in the Pacific coffee belt of Guatemala. B. A bit of a temple wall at Copan. C. One of the Stelae at Copan. D. Forest in which ruins at Kichen-kanab are located. Plate 10. 211 A. The church of Esquipulas, representing the best Spanish architecture in Guatemala. B. The riverward side of the main citadel at Copan. Plate 11. 218 A. The ruins of Quiche, the most extensive on the Guatemalan plateau. B. Near view of the most imposing ruins of Quiche. Plate 12. 230 A. Stelse inscribed with hieroglyphics at Quirigua. B. The ruins of Copan. Sketch map of Arizona and New Mexico, showing location of places mentioned in the text. Map 2. 175 Sketch map of a part of Central America, showing location of Maya ruins. TEXT CUTS. Page. 1. Rainfall of Arizona and New Mexico. 10 2. Annual rainfall at Tucson, Arizona, 1868-1912. 11 3. Winter and summer rainfall at Tucson. 13 4. Comparison of 3-year means of winter and summer rainfall at Tucson. 13 5. Cross-section illustrating the formation of climatic terraces. 27 6. Profile of climatic terraces. 33 7. Rainfall and emigration in Europe. 89 8. Cross-section of alluvial terraces in mountain valleys near the City of Mexico. 100 9. Annual growth of trees at Prescott. 107 10. Annual rainfall and growth of trees (Group V) at Prescott. 108 11. Annual growth of trees at Flagstaff, and variations in annual rainfall according to month which is reckoned as the beginning of the year.. 109 12. Growth of individual trees compared with precipitation at Flagstaff. Ill 13. 14. Effect of monthly distribution of precipitation on thickness of rings of growth. Ill 15. Monthly and yearly precipitation from 1866 to 1909, and size and character of rings. 112 IV LIST OF ILLUSTRATIONS. V 16. Actual tree growth compared with growth calculated from rainfall. 113 17. Five-year smoothed curves of rainfall and tree growth at Prescott. 113 18. Actual rainfall compared with rainfall calculated from gi’owth of trees, Arizona. 114 19. Annual growth of trees at Flagstaff since 1385 A.D. 116 20. 500-year curve of tree ^owth, 20-year means. 117 21. A possible 150-year period. 117 22. Mean curve of the 21-year cycle. 117 23. Variations of the 11-year cycle. 118 24. Comparison of eleven 4-year cycles in tree growth, rainfall, temperature, and inverted sun-spot numbers... 119 25. Sun-spots and the growth of trees at Eberswalde, Germany. 120 26. Ideal curves illustrating correction for age. 125 27. Ideal curve illustrating correction for longevity. 127 28. Curve of growth and correction for age of yellow pine in New Mexico. 128 29. Curve of growth of 50 yellow pines over 280 years of age. 129 30. Variation in radial growth by decades. 131 31. Curves of growth of American trees. 133 32. Curves of growth of western yellow pine in New Mexico and Idaho. 135 33. Rainfall of Idaho compared with that of New Mexico. 136 34. Ideal diagram to illustrate the dropping of rings. 148 35. Sequoia washingtoniana, corrective factor for age during first 250 years of hfe. 150 36. Sequoia washingtoniana, corrective factor for age, plotted by centuries. 151 37. Sequoia washingtoniana, corrective factor for longevity. 152 38. Curve of growth of the Sequoia washingtoniana in California. 153 39. Effect of flaring buttresses on the measurement of rings of growth. 154 40. Annual rainfall at selected stations in California. 158 41. Monthly distribution of precipitation in California. 160 42. Rainfall at Portersville compared with growth of Sequoias at Dillonwood. 161 43. Annual growth of 111 Sequoias at Hume. 162 44. Growth of trees at Hume, and rainfall at Fresno. 163 45. Mean monthly distribution of rainfall compared with distribution in exceptional years. 164 46. Conservation factor in the relation of growth and rainfall, method 1. 166 47. Conservation factor in the relation of growth and rainfall, method II. 166 48. Tree growth in California calculated from rainfall. 167 49. Rainfall by months in favorable and unfavorable years. 170 50. Changes of climate in California and western Asia dvuing historic times. 172 51. Storm frequency, 1878-1887. 191 52. Storm frequency, January. 194 53. Storm frequency, February. 194 54. Storm frequency, March. 195 55. Storm frequency, April.'. 195 56. Storm frequency. May. 196 57. Storm frequency, June. 196 58. Storm frequency, July. 197 59. Storm frequency, August. 197 60. Storm frequency, September. 198 61. Storm frequency, October. 198 62. Storm frequency, November. 199 63. Storm frequency, December.7. 199 64. Storm frequency. Year maps for 1878-1887, after Dunwoody, and 1899-1908, after Kullmer, showing shift of storm track. 201 65. Changes in storm frequency by months according to longitude.202 66. Changes in storm frequency by months irrespective of longitude and latitude. 202 67. Changes in storm frequency by months according to longitude in latitude 50“-55°. 202 68. Changes in storm frequency by months according to longitude in latitude 45°-50°. 203 69. Changes in storm frequency by months according to longitude in latitude 30°-35°.204 70. Changes in storm frequency by months according to longitude in latitude 25'’-30°.204 71. Summary of differences in storm frequency, 1878-1887 and 1899-1908. 204 72. Changes of climate in California for 2,000 years.209, 231 73. Diagrammatic sections of wall and deposits at Copan Ruins. 214 74. Relation of terraces and ruins at Copan.214 75. The Relation of sun-spots and tree growth in the 11-year cycle. 239 76. The sun-spot cycle and terrestrial phenomena. 240 77. The com crop of the United States, 1901, a “lean” year. 244 78. The com crop of the United States, 1906, a “fat” year. 244 79. The corn crop of the United States, 1908. 245 80. The com crop of the United States, 1909. 245 81. Variations of the solar constant and monthly departures from mean temperature at Arequipa.246 82. Monthly departures of temperature in South Equatorial regions, showing agreement.247 83. Monthly departures of temperature in North and South Equatorial regions, showing disagreement.248 84. Monthly departures of temperature in North America compared with Arequipa in Peru. 249 85. The Relation of volcanoes, sun-spots, and terrestrial temperature. 252 86. Geological changes of climate and movements of the Earth’s crust.256 87. Map of Pleistocene glaciation. 266 88. Paleogeography and glaciation of early Permic times. 267 89. Map of Proterozoic glaciation. 270 90. Chart of geological climates. Paleometeorology. 285 NOTE TO THE READER. This volume as a whole deals with climate, but various parts are concerned with particular aspects of the problem. The reader who is interested in one special phase is referred to the following chapters: Physical aspects of the Southwest, Chapters I to VI. The Ancient People of the Southwest, Chapters VII to X. The Measurement of Rainfall by the Growth of Trees, Chapters XI to XIV. The Maya Civilization and Changes of Climate in the Torrid Zone, Chapters XV to XVIII. Theories of Climatic Changes, Chapters XIX and XX. The Climate of the Geological Past, Chapter XXI. The volume may be divided in another way according to the sciences with which the different chapters are more especially concerned. In this division, however, it must be recognized that there is much overlapping, for the same chapter often deals with several sciences. Climatology, Chapters I, XIV, XVI, XIX. Geology, Chapters II, IV, V, VI, X, XX, XXL Botany, Chapters III, XI, XII, XIII, XIV. Archeology, Chapters VI, VII, VIII, IX, X, XVI, XVII, XVIII. Ethnology, Chapters VII, XV, XVII, XVIII. The reader who desires to understand the main outline of the theories here presented, but does not care to go through all the details of evidence is advised to read the Introduc¬ tion, and Chapters VI, VII, IX, XIII, XIV, XVI, XVII, XIX, and XX. VI THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. mi ■<% ►-fl . .jj * - ^ •-%•■■ 0 > !vr@ |1T- tr4/’'- "S 1 . ' ' ' » ' ■ W* "W ?“:il .r. 'Vi^r '2? I ' ^‘3 a. - I • ^ « • * 4 .-L 'J ,* -J,” • '.;■ “A' *«■ y.if^ ■« .»• r.fx f* •' ,r 'wiV V;M' 9 a^« »• • • . -. • il. • . X ; 1/';, , ^ -v j.! _?> -A v;»%i 1*^ ’ *4 * ■ lA. •*! Y* ... •• '. ' »•■ ■■■■■■ V 55>* . *s f-v^' * \ ^-v ' r-i*^ ^'V Csf.' '■'■'S'-.V:': ‘ V> "'i if^' «?> ■T\ * . ■ -r. ♦ ^ ' A - - , ■ 'f . .*4^7^ 5 *■■%■■ *■■■' k*-' ..vV ?®*c I ■■ 1 ^ ■' v-Ai*;! ■.*«•-« ' ♦'> ^5*^ *'’^i ‘‘I** - i..-'' ,. ''■ ' ..' t fc'’ »A. U ”> I ' r,ri. V ^ vr * it"1 ih k* T r t>- thx i I •’ •. Si ■ ? |«-VIM^J i'.;i ^ i v> »,* Jt' 4*s%-' -tH? L?P? c'liiW n*fr“ i ri>,l A' rAi i r-’ ‘■4i t “V-?* Sr*’\iK ■>_. •) ■sfit Al+V. T ■M » vVi- 1(U> KHKini '!< ‘ IS'-Y ^Vl ^ JW f^’>' £v,:-i «^«ii ~3:*J HIT/ L 5( y -Ai. ..if 9, • V I •i^ 4 ;;i|% i f'' I • ,< r _ S ’ ‘ J-. ’ ^ A*!*'' A' **n^ *f9^j pmnjt^ f ~ -•■ . Hi >^i? I k L;- -x .Ki ■- gM ■ ■ T , .ur «V' nw^ •‘•^V • • ’ ^'‘' oDixjfM */.y<<^ff/:w*f fXiT aHT ’yft CW?gii>yjM/. '■' .* • D., - • / ■■^?'4'-' ’ ^ ^rK'/"-*- fj' ,■•^•■ - '•i, . I ji »M» X. V t- A • ..^ "ST.-* Mif ..■ .•k >4. -V /.,/ MAP I. HUNTINGTON ^ CiifuirroTi , Des Moines | SOUTHERN UTB. Kanai) (RaibabI Wagpi^' Mors ^ MOQUI IND.EES, Csaakar > Mess I / Oraibi. Chico Antoii Trrand 4 Sco/^. Mtn Tlliams ' Thpreau wslottf Holbrook^ 'ONtflO N.F. ‘ Bnowflaiy ■ IwiviKF-yr Magdalena Qparte^te trosaoS IKD.RE PortTK Hillstioro yuma( Rillito^ y .fiare. Nollik ^ Artesa InpWs Pa/ra I Cibr ii Janos El Altai’ [U/’-wCaiorca ■Bnfcani 'Lasco Trinchera SKETCH MAP OF ARIZONA AND NEW MEXICO SHOWING LOCATION OF FLACKS MENTIONED IN THE TEXT JULIUS e»EN CO.PHOTC UTK IjOO 50 0 100 Miles zn INTRODUCTION. Climate as an element of physical environment is so well recognized that there is no need to demonstrate its importance. By common consent it is held to be a primary factor not only in the life of plants, animals, and man as they exist to-day, but in their entire evolution. Moreover, among the main elements of physical environment it alone is subject to pronounced changes in comparatively brief periods. The form of the lands, the location of mountains, and the composition of the atmosphere are doubtless all subject to great changes, but these are too slow to have much effect upon a single generation of living beings or even upon all the generations that have existed during the period since man emerged from barbarism. Small climatic changes, however, such as those from one year to the next, or from one decade to another, are constantly in progress, and their far- reaching results are a matter of every-day experience. Moreover, it is quite possible that larger changes have taken place during the past 2,000 or 3,000 years, and if this is the case their effects must have been correspondingly important. The investigation of this possibility is the purpose of this volume. The study of changes of climate naturall}’' divides itself into two parts, relating to the present and to the past. The facts as to the present are being rapidly gathered by the excellent work of the various National Weather Bureaus of the world. The facts as to the remote past are being studied minutely by geologists so far as they relate to geological times. Comparatively little, however, is known as to the state of affairs during the period covered by history and man’s later development. Yet a knowledge of this period is essential. In the first place, it is only by accurate knowledge of past variations that we may hope to ascertain the causes of present variations, and thus to predict those which will occur in the future. Geological evidence of course tells us much about the past, but it pertains largely to periods too remote to be of great present importance, and its phe¬ nomena can not be dated with accuracy in terms of years. Hence something else is needed to fill the gap between such geological phenomena as the glacial period and our modern climatic records covering scarcely a century. In the second place, a mathematical investi¬ gation of the chief effects of present climatic conditions may do much to show how far human habits, customs, physiological traits, and mental character are influenced by physical environment, but it is impossible to determine the exact effect of present con¬ ditions until we know how long those conditions have lasted and how the environment of the past, especially during the last 2,000 or 3,000 years, differed from that of the present. Hence along many lines the study of the climatic variations of historic times is essential as the foundation of future work. The present volume is an attempt to determine the sequence and character of such variations on the basis of evidence in the drier portions of America from Guatemala on the south to Idaho on the north. A large number of phenomena from the diverse fields of geology, archeology, history, and botany seem to agree in indicating that during the past 3,000 years North America has been subject to pronounced climatic pulsations similar to those which appear to have taken place in Asia and other parts of the Old World. In the temperate portions of the Eastern Hemisphere the climate of the past appears on the whole to have been distinctly moister than that of the present. The change from the past to the present, however, does not seem to have been gradual and regular, but pulsatory or cyclic, so that certain periods have been exceptionally dry, while others have been wet. In America the same appears to be true. 2 1 2 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. The facts set forth in this volume are the result of investigations carried on in coopera¬ tion with the Department of Botanical Research of the Carnegie Institution of Washington, at the invitation of the Director, Doctor D. T. MacDougal. The first field season, March, April, and May, 1910, was devoted to the study of a relatively restricted area centering at Tucson, the site of the Desert Botanical Laboratory in southern Arizona, and extending southwestward for 150 miles to the shores of the Gulf of CaUfornia in northwestern Mexico. During the second season the months of March and April, 1911, were devoted to the investigation of selected sites in various parts of New Mexico, while May and June were occupied with measurements of the rate of growth of about 200 Sequoia trees in the central parts of the Sierra Nevada Mountains in Cahfornia. The third season was divided into two portions. In the first place, six weeks during March and April, 1912, were spent among the lakes and ancient ruins of southern Mexico and Yucatan. In the second place, the work upon the Sequoias in California was carried further, and about 250 more trees were measured. Finally, in March and April, 1913, independently of the Carnegie Institution, the author made a journey to Guatemala for the purpose of investigating the ruins of that region and their relation to the physical surroundings and vegetation of the country. Throughout the work much attention was given to the influence of the present climatic conditions upon physiography and upon the habits and distribution of plants and animals, including man. No attempt will be made to deal with these subjects here, however, except in so far as they bear upon changes of chmate. The purpose of this volume is primarily to investigate the extent and nature of such changes and our attention will be devoted almost exclusively to that subject, while the interesting problems of the relation of climate to human character and history will be left for another volume. The first portion of our investigations will be concerned entirely with New Mexico, Arizona, and the adjacent parts of the Mexican state of Sonora. Inasmuch as an intelligent knowledge of present climatic conditions is essential to a full understanding of the past, the present chmate of those regions and its relation to the great climatic zones of the earth as a whole wiU form our first subject of consideration. A minute knowledge of the land forms of the region is not necessary for our present purpose, but a clear conception of the main types and of their relation to climatic conditions is advisable. Accordingly a chapter will be devoted to the general aspect of New Mexico, Arizona, and Sonora, and to the chief types of physio¬ graphic forms commonly found there. To complete the picture a brief chapter on vegeta¬ tion will be added, not from the point of view of the botanist but of the geographer. Having gained a general idea of the physical aspects of New Mexico and Arizona, we shall be prepared to turn to some of the details which furnish evidence as to past climatic conditions. First we shall take up those fines of inquiry which involve only physical processes without respect to man. Among these a foremost place is occupied by the alluvial terraces which are so widely distributed throughout all arid mountainous regions. Their importance as possible indicators of climatic changes is so great that I shall consider the problem of their formation in detail. The strands of ancient lakes are often closely associated with river terraces, and the conclusions to be drawn from them are similar. Our region has so few lakes and these few have been so little investigated that they are of less importance than many other fines of evidence. Nevertheless, the Otero soda lake near Alamogordo in New Mexico and the group of small lakes near the City of Mexico are of such interest as to warrant a somewhat full discussion. The desiccated bed of the Otero Lake presents evidences of a changing climate not only in its old strands, but in the re¬ markable series of gypsum dunes of various ages which surround it; while in the Mexican lakes natural causes appear to have induced variations in size even during the period since the coming of the Spaniards. From the purely physiographic portion of our investigations in Arizona, New Mexico, and old Mexico, we shall proceed to the main phase of the subject—the study of traces of INTRODUCTION. 3 ancient human occupation. In this part of the work we shall examine a large number of ruins scattered from the shore of the Gulf of Cahfornia to the northern limits of New Mexico, 500 miles away. Except among a few archeologists, the number of ruins is rarely appreciated. The ruins not only indicate a considerable degree of culture, but they show distinctly that different races occupied the same sites in succession. The successive occu¬ pations were separated by periods of abandonment, due possibly to climatic causes, or perhaps to something quite different, but at least well worthy of study. In considering the ancient ruins and their prehistoric inhabitants it is essential to keep fairly in mind two opposing theories. The first, which is usually accepted, holds that the large number of ruins does not indicate a correspondingly large population, and that the physical conditions of the country in the past, just as at present, forbade any great number of inhabitants. The other, which is accepted by only a few scholars, holds that the ruins were occupied by a relatively dense population which persisted for a long time. This could have been possible only on the assumption of greater rainfall than now, and therefore those who hold this view believe in changes of climate. The lines of reasoning followed thus far are similar to those which I have employed in respect to various countries of Asia, and which were first set forth in a volume entitled “Exploration in Turkestan,” published by the Carnegie Institution of Washington in 1905, and have been amplified and revised in “The Pulse of Asia” and “Palestine and its Trans¬ formation,” published in 1907 and 1911, respectively. The conclusions derived from these lines of reasoning are open to the criticism that a preconceived theory may have led to the interpretation of phenomena according to that theory. Hence, before the conclusions here indicated deserve final acceptance, it is necessary to compare them with the results obtained by the observations of other unprejudiced workers, or with the independent results of some new method in which the personal equation plays no part. Fortunately the work of Professor A. E. Douglass, of the University of Arizona, suggests a method by which old trees may be used as a mathematical measuring-rod in order to determine exactly what climatic events have occurred during the last 2,000 or 3,000 years. Accord¬ ingly, considerable space will be devoted to setting forth the results of the measurement of the rate of growth of nearly 500 Sequoia trees among the Sierra Mountains in California. A large number of other measurements of trees by the United States Forest Service have been kindly put at the disposal of the Carnegie Institution of Washington by the Forester, Mr. Henry S. Graves, and a discussion of them is included in this volume. By purely mathematical methods, unaffected by any personal bias, it has been possible to obtain curves indicating the climatic pulsations of the last 3,000 years. A comparison of these curves with the results obtained from other lines of evidence, both in America and Asia, shows that in spite of certain disagreements the general climatic history of both continents appears to have been characterized by similar pulsations having a periodicity of hundreds or thousands of years. If the conclusions outlined above are accepted, they may perhaps furnish a key to the pre-Columbian chronology of America. Apparently the Southwest has been first relatively inhabitable and then relatively uninhabitable during periods lasting hundreds of years. The dates of these periods are ascertainable from ancient trees. Each propitious period has probably been a time of expanding cultme and comparatively dense population, while the unpropitious periods have been times of invasion, disaster, and depopulation. Archeo¬ logical study has begun to differentiate periods of this sort, but has been unable to date or correlate them. If the evidence of climate be considered together with that of archeology, we may perhaps at length be able to overcome the absence of written documents so far as to construct a fairly intelligible record of the history of our ancient predecessors. The broader the area under observation, the greater is the probability of accurate results. Hence, after two field seasons in Arizona, New Mexico, and Sonora, it seemed wise 4 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. to extend the work more than 1,300 miles southward to southern Mexico and Yucatan, and finally 400 miles farther into Guatemala and Honduras. Here again the lines of reasoning previously employed in Asia and later in the United States and northern Mexico gave similar results. In Yucatan and Guatemala, however, a new type of phenomena was also found, which confirmed the previous conclusions most interestingly. In that tropical region the presence of magnificent ruins in the midst of dense forests seems to indicate changes of climate contrary to those of the regions farther north. On the whole the country appears to be moister than it was several thousand years ago, instead of drier, as in regions farther north. This at first sight appears contradictory, but in reality it confirms a conclusion derived from other evidence, namely, that changes of climate are probably characterized by the shifting of the world’s great climatic zones from north to south, and the reverse. Having reached this conclusion, we find ourselves able to test it by means of another of wholly independent nature based on a comparison of historic events and climatic changes in Europe and Asia. This second conclusion is that while the course of history depends upon a vast number of factors and its details are due to what may be designated as purely human causes, yet in its broader outlines it is profoundly affected by climatic changes. These may alter economic conditions, they may disturb the adaptation of a race to its environment by fostering special diseases, or they may force people to move out of a habitat where they can no longer compete with nature. In Yucatan and the neighboring parts of Central America the dense tropical forests and their deadly malarial fevers are man’s chief enemy. Our study of the terraces, ruins, trees, and other phenomena of the United States and northern Mexico leads to the hypothesis that the forests of Yucatan and the surrounding regions have alternately increased and diminished in size, the increase coming at times when aridity prevailed in regions such as CaUfornia, and the decrease during California’s moist times. Our hypothesis as to the relation of changes of climate and history leads to the inference that civilization in Central America would thrive when the forests diminished and would decline when they increased. A comparison of the chmatic periods indicated by the Sequoias with the events of the history of the ancient Mayas of Yucatan, as recorded on monuments and in chronicles, shows that our expectations are realized to a considerable degree. Thus both conclusions are strengthened. The degree of certainty given to our conclusions by the agreement of the evidence of old trees with that derived from other sources brings us to the point where we may reason¬ ably attempt to ascertain the cause of changes of climate. In such an attempt the first matter to claim attention is a certain degree of coincidence between the phenomena of climate and those of the sun. There are many reasons for thinking that the well-known sun-spot cycle of 11 years is related to a distinct climatic cycle. The longer, but less thoroughly estabUshed 35-year cycle of Bruckner also appears to be correlated with the activity of the sun. These things, as has often been pointed out, suggest that our minor climatic fluctuations maybe due to slight variations in the intensity of the sun’s radiation. The work of Langley, Abbott, and others proves that the sun’s radiation actually does vary, while that of Koppen, Newcomb, and many more proves that the temperature of the earth’s atmosphere shows a corresponding variation. The terrestrial variation is so slight, however, and is so irregular outside of equatorial regions, that some of the best authorities doubt whether it is sufficient to produce appreciable results. Opposed to this is the fact that by almost universal consent students of glaciation beheve that a permanent lowering of the earth’s mean temperature to the extent of from 3° to 10° C. would produce a glacial period. The most conservative estimates of the change in terrestrial temperature between the minimum and maximum of sun-spots is about 0.5° C. This is so large a frac¬ tion of the change needed to produce glaciation that it seems as if it must produce some appreciable meteorological results. INTRODUCTION. It is true that attempts to detect such results have hitherto proved contradictory, but this is not surprising. It seems to be due partly to the lack of long, homogeneous records and partly to failure to make due allowance for the different degrees to which different kinds of phenomena, such as temperature, pressure, wind, and rain, must lag behind their cause, especially under conditions such as those of the atmosphere, where large quantities of heat are transferred from one region to another. The new method of investigation of climate by means of the growth of trees, as elaborated by Professor Douglass in his contribution to this volume, furnishes long, homogeneous records, and these show a distinct sun-spot cycle. The careful researches of Arctowski upon '‘pleions” and “anti-pleions’’ seem not only to demonstrate a relation between solar phenomena and terrestrial climate, but also to show why the manifestation of this relationship is irregular and, at first sight, con¬ tradictory. These things lead to the conclusion that the small climatic cycles now in progress upon the earth are in large measure due to variations in the sun. In regard to greater climatic changes, it appears that the pulsations of the past 3,000 years are too large to be due to fortuitous rearrangements of the earth’s atmosphere because of purely terrestrial causes. On the other hand, they occur too rapidly to be due to pre¬ cession of the equinoxes, changes in the carbonic-acid content of the air, or deformation of the earth’s crust. Hence we are led to conclude that they, too, are due to variations in the sun. The same conclusion seems to apply to glacial and interglacial epochs, since their characteristics appear to be identical in nature with those of the pulsations of historic times, although differing greatly in degree. In explanation of still greater changes, how¬ ever, such as the radical difference between the distribution of climatic zones in the Permian and Pleistocene eras, something else is demanded. In Part II of this volume the matter is fully presented by Professor Schuchert from the standpoint of the paleontological geolo¬ gist. We are there led to the conclusion that while changes in the amount of carbonic-acid gas in the atmosphere may explain certain climatic phenomena, they can not explain this particular feature. Crustal deformation, on the contrary, seems fully adequate to cause just such a redistribution of zones as we find from time to time in geologic history. It apparently can not, however, account for the climatic instability which often accompanies or immediately follows periods of crustal deformation, that is, for fluctuations from glacial to interglacial conditions and for minor pulsations. In explanation of these it seems reasonable to turn back to our solar h 3 rpothesis. Thus we are led to the final hypothesis that for some unknown cause both the earth and the sun have been repeatedly thrown into activity at approximately the same time. The activity of the earth seems to manifest itself in the changes of form whereby continents are uplifted and mountain ranges shoved up. That of the sun seemingly displays itself in pulsations which give rise to climatic variations of every grade, beginning with glacial and interglacial epochs and ending with little cycles like that of the sun-spots. Here we leave the matter, but not without a word of caution. Throughout this volume our purpose is not to develop the hypothesis just outlined as to the interrelation of solar activity, crustal deformation, and climatic changes. That hypothesis, important as it may prove, is by its very nature open to grave question. At best it is merely a corol¬ lary of our main conclusion, whose truth or falsity is in no way dependent upon it. The primary purpose of this book is to investigate any possible climatic changes which may have taken place in historic times. Our main conclusion is that such changes have taken place and that they have been of a pulsatory nature. All other questions are here sub¬ ordinate, and the truth or falsity of this conclusion is the point upon which attention should be focused. ^ ' ' I::* -Ilf • 'i«. r • ? lanr'in- ■» ■ ' ,' ^ "3- • ^Jvv-i . • * - f^uLJDil ■ ' ' J» .f5^- .7^ ' :.''!v 'T.i < • *■- '> ftv* j-'A \ iIT" >s # •Kr^s vVl ‘ri'CZK 'r . J=i' ’j* • ^ . Ilf*‘ • ^T *' ■ V'> Jt’ '’i ' '■ Kiv. -)J-.V‘ii!!!'>3f' '.' > V . ,y/T;w ^, ci.-s:.^.£fv,. '>i ;. (y>’ M F- !* IS fT‘ f-‘i;. 4*-fr'.' •''V'-,:1 -4.' •-l.<' ■'-^’ ,1'V fr* .y: ^1'1 ■Sll’ '• ■ 'i*^. •..- .. »* J.:< ‘ T*’T>*' ' *'■ . t.s >5^., - ^ 4 ^ vj r?» fi 4t i i MB’ •ijjtj' ■-^’> H ►' ' 7 Vv’>’ '4 'I* ." '•'■ " ^ • ‘-^ tl li' ■ ' 'V . .j'k , ■*..*■> 'v^ te^‘ .?T J VV <■■7; Ly*': '■^'' ‘^''' -- ^'' '■’ ’■ f ^ .“'■f '.M «C.'4‘: e'F '!« '’Vil ! I. . ,v' :v* i kV*'i «i^. , t V':*M ... V'.-"? .d'ifi ■■ lu J*a' ‘ '-'♦*'■» *'■ ■-■ 'f-ti®- ^^'^''•. ■’;l , AVvf *• . . •.*' k'*^ '»l'* ' ^ '•' . * . *-•* -W ,.m :S3i. kri: |v fi 4 ■'>,* :k^:i PART I. THE PROBLEM OF RECENT CLIMATIC CHANGES. 7 -• _ ■ CHAPTER I. THE MONSOON CLIMATE OF ARIZONA AND NEW MEXICO. The climate of Arizona, New Mexico, and northern Sonora is of a peculiar, transitional type. It may be defined as a subtropical continental climate of the monsoon variety. It resembles that of the provinces of the Punjab, Rajputana, and Sind in northern India more closely than that of any other part of the world. As the region extends from about north latitude 28° in Mexico to 37° in northern New Mexico and Arizona, its subtropical position brings most of it within the great world-zone where high pressure and consequent aridity normally prevail. Here the main movement of the air is downward and outward; and here the northeasterly winds of the trade-wind zone and the southw^esterly winds of the zone of prevailing westerlies find their origin. If the climatic zones of the earth were not interfered with by the different rates at which land and sea are heated and cooled, seasonal changes would bring this region within the range of the prevailing westerlies and their rain-bringing cyclonic storms in winter, and within the trade-wind belt in summer. In winter the country would receive a fair amount of rain for a few months, the edges, so to speak, of the great storms which whirl across more northerly regions not only in winter but in summer. During the rest of the year it would be rainless, for in spring and fall it would be in the subtropical zone of high pressure and descending air, while in summer the trades would blow across it from the northeast. Inasmuch as the trades would come from a dry interior, they would bring no rain. As a matter of fact, however, the trade-winds are never well developed in Aiizona and New Mexico, and herein lies the explanation of the most peculiar characteristics of the climate. The cyclonic storms of the westerlies in winter and the descending air of the subtropical ‘‘horse latitudes” in spring and autumn give rise respectively to the rain and the aridity which would be expected. In summer, however, because of the great size of the continent of North America, the trade winds which would be expected do not appear; their place is taken by relatively moist winds which blow in general from the south, and may be called monsoons for lack of any more appropriate name. In order to give definiteness to our discussion of the climate of the southwest, let us recall the familiar general principles of the effect of continents upon temperature, pressure, winds, and rainfall. Land masses, as is well known, become heated or cooled much more quickly than expanses of water. Hence, in winter the continents become much colder than the oceans, and are therefore the seat of centers of high barometric pressure, a con¬ dition exactly the reverse of that prevalent over the comparatively warm oceans. From the continental areas of high pressure the winds tend to blow outward, especially toward the east and south. Thus the cold waves of the Eastern and Southern States arise, for on the western side of ordinary cyclonic storms the indraft of air occasioned by the storms themselves is strengthened by the general high pressure prevaihng in the cold interior of the continent. Inasmuch as Arizona and New Mexico, unlike the parts of India with which we have compared them, are not protected by an east-and-west range of mountains such as the Himalayas, chill winds from the north sweep over the country in winter, producing frequent frosts. Even as far south as Tucson, in latitude 32° and at an elevation of only 2,300 feet above sea level, the thermometer occasionally falls to 16° F. Except in the warmest and lowest places, such as the Gila Valley around Phoenix, or at Yuma on the Colorado River, this liability to sudden cold prevents the growth of subtropical fruits, such 9 10 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. as oranges, although much farther north in California these grow to perfection. Oddly enough, the warmth of the winter in the intervals between cyclonic storms, when north winds do not prevail, is almost as fatal to such northern fruits as apricots and peaches as is the low temperature to oranges. Since this is a desert region of clear skies and slight humidity, the daily extremes of temperature are naturally great, amounting often to 40°. In Febru¬ ary, or even January, it is not uncommon for the mercury to rise to 70° F., and if the nights are above the freezing-point the fruit trees are stimulated to open their blossoms too early. A blighting wind from the north swoops down, and the flowers are nipped. In summer the conditions are the reverse of those of winter, except that the range of temperature from day to night is still extreme. The whole continental interior becomes greatly heated and in Tucson the temperature rises occasionally to 114° F., and at Yuma still higher. Under such circumstances low barometric pressure must of necessity prevail, and winds from the periphery of the continent tend to blow inward. This tendency is so Fig. 1. —Rainfall of Arizona and New Mexico. The figures show annual rainfall in inches. From two maps published in Bull. No.188, Bureau of Plant Industry, U. S. Dept. Agriculture. strong that toward the end of June the trade winds, which would normally be expected in the district from Arizona southward, are entirely destroyed. They prevail in normal fashion over the adjacent oceans, but on the continent they give place to somewhat irregular winds whose prevailing direction is distinctly northward. They form, as it were, an inward draft blowing from the Gulf of California and the Pacific Ocean on the one hand, and from the Gulf of Mexico on the other, toward the continental center of low pressure. In all essential respects they are like the monsoons of India, although less strong and distinct because of the smaller size of the continent. As the American monsoons approach the land, the first tendency is for them to become heated and hence relatively dry, for the land is hotter than the ocean, and in many places the height of the mountains is too slight to overcome the heating due to the land. The case is like that of the plains of Sind at the mouth of the Indus. As the winds blow inward, however, they are soon forced to rise by the mountains, they reach more northerly and hence cooler latitudes, and they enter the continental area of low pressure where the general tendency of atmospheric movements THE MONSOON CLIMATE OF ARIZONA AND NEW MEXICO. 11 is upward. Thus in the peninsula of Lower California, in the Mexican states of Sonora and Chihuahua, and in Arizona and New Mexico, the summer is characterized by heavy thunder-showers of the kind commonly known as tropical. These usually occur from about the end of June to the early part of September, beginning earlier and ending later in the south than in the north. Thus the country has two rainy seasons, one in winter deriving its rain from cyclonic westerly storms, and one in the summer deriving rain from southerly monsoon thunder-storms. The total rainfall is small, ranging from 5 to 20 inches per year in most parts of the area, as is shown in the map, figure 1, and rising above 20 inches only in the high mountains. The variation from year to year, however, is great, as may be seen in figure 2, where the rainfall of Tucson is plotted by calendar years. The rainfall of the two seasons, summer and winter, is still more variable, a fact evident from figure 3. The average of the winter season at Tucson is 4.5 inches. The amount is small because the moisture comes largely from the Pacific and must cross the high Sierras on the way. The winds, of course, often blow from the east at the actual time of rainfall, but this affords no indication of the source of the moisture. In all cyclonic storms of the northern hemisphere the motion of the air around and toward the centers of low pressure is similar, and the southeastern quadrant of a cyclonic area lying in front of the storm where the air has not yet suffered depletion of its moisture, and where the winds move rapidly from warmer to cooler lati¬ tudes, is apt to have a rainfall more abundant than that of any other quadrant. Much of the rain which accompanies such winds has doubtless come from the oceans to the eastward, but more has probably been brought by the prevailing westerly winds and is merely caught up and prepared for precipitation by the easterly winds. Inches 68 1870 72 74 76 78 1880 82 84 86 88 1890 92 94 96 98 1900 02 04 06 08 1910 3017 I Fig. 2.—Annual Rainfall at Tucson, Arizona, 1868-1912. The paucity of the winter rainfall would not be so harmful were it not for its extreme variability. In the winter of 1903-4 the total precipitation at Tucson for the six months from November to April, inclusive, amounted to only 1.08 inches, while in the succeeding year it amounted to 14.74. Records kept at Tucson and at the neighboring army post of Fort Lowell show that in the years from 1868 to 1912 the winter rainfall was less than 2.5 inches, or practically useless, in 9 winters; it amounted to from 2.5 to 5 inches, that is, it was fair, in 20 winters; it ranged from 5 to 7.5 inches, or was good, in 13 winters, and exceeded 7.5 inches only three times. (See table la.) These figures do not show quite the true state of affairs so far as agriculture is concerned, for 3 inches in February and March, after the chief frosts are over, are worth double that quantity in November and December. Still, the figures serve to give an idea of the extreme variability and uncer¬ tainty of the winter rains. Manifestly, even with the help of irrigation, the prospects of the farmer are not of the rosiest when he may have only one-fourteenth as much rain in one year as in another. After the dry spring season — the fore-summer, as MacDougal has called it*— the southerly monsoon gradually becomes well established by the strong indraft toward the heated continent, and thunder-showers finally begin upon the mountains. Far to the south in Mexico the first showers may come in May or even April. In southern Arizona they usually begin, as we have seen, toward the end of June or early in July, while farther north * D. T. MacDougal; Botanical Features of North American Deserts. Cam. Inst. Wash. Pub. 99, 1908. 12 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. they do not come till mid-July. In exceptionally warm years, however, they may begin unusually early because of the more rapid heating of the continent. Thus in 1910 the mean temperature of the month of May at Tucson was 6° F. above the average of the preceding four years; and showers, light on the plains, but heavy on the mountains, began early in June. Farther north, where the showers begin later, they also end earlier, and instead of lasting into September, terminate in August. Everywhere they are accompanied by vivid lightning and the rainfall is torrential. The summer rains are more abundant and less vari¬ able than those of the winter. At Tucson they average Table 1a. —Summer and Winter Rain¬ fall at Tucson (and Fort Lowell) from 1868 to 1912. 7.14 inches for the six months from May to October inclu¬ sive. During the 45 years embraced in the records at Tucson and Fort Lowell the minimum was 3.01 inches in 1885 and 3.03 in 1900, while the maximum was 14.2 inches in 1876. Fifteen summers had a rainfall of less than 5 inches, 12 from 5 to 7.5 inches, 9 from 7.5 to 10 inches, and 9 over 10 inches. The summer showers are so sudden and the rain falls so rapidly that a large part of the water runs off in great floods, serving no useful pur¬ pose. Nevertheless, the showers support considerable vegetation, and from the earhest times have enabled the inhabitants to cultivate quickly growing crops like corn and beans, but neither these nor any other crops can be grown without irrigation except in a few places at great altitudes; yet in these places there is always danger of failure. Even with the aid of irrigation the arable area is at best extremely limited. The division of the rainfall into two seasons has, as we shall see, a beneficial effect upon native vegetation, but it can scarcely be considered particularly advantageous to agriculture, especially to the type brought by the inhabitants of Europe to the better- watered parts of the United States. From a theoretical standpoint the rainfall of Arizona and New Mexico is peculiarly interesting because the winter rains possess the characteristics of the temperate zone and the summer rains those of equatorial regions. It is not possible to enter into any detailed discussion of the subject at this time, but one or two matters may be pointed out as especially deserving of study. In the first place, at high elevations among the mountains the sea¬ sonal distribution of precipitation is not the same as in the lowlands. For instance, on the Santa Catalina Mountains, over 9,000 feet in eleva¬ tion, Dr. MacDougal has found not only that the rainfall is two or three times as large as down below, as might be expected, but also that the winter rains are heavier than those of summer. This, of course, is the reverse of what prevails in the lowlands. It seems to mean that the climate of the mountains approximates to that of regions farther north, not only in temperature, but in the character of its storms. In other words, it seems as if westerly winds and their attendant conditions prevailed for a longer time, or else during the same time, but more completely at high levels than at low, while in the districts of lower altitude equatorial conditions are predominant so far as precipitation is concerned. To put the matter in another way, it may be that the heating up of the continent in summer disturbs the equilibrium in the upper air so much less than in the lower that [See Figures 3 and 4]. Summer of— Inches of rainfall. Winter of— Inches of rainfall. 1868 8.19 1867- 68 4.43 1869 9.48 1868- 69 4.10 1870 4.86 1869- 70 2.25 1871 7.12 1870- 71 2.10 1872 11.48 1871- 72 1.27 1873 3.43 1872- 73 3.09 1874 7.90 1873- 74 7.31 1875 8.90 1874- 75 2.97 1876 14.20 1875- 76 2.20 1877 6.37 1876- 77 4.21 1878 11.16 1877- 78 6.42 1879 4.29 1878- 79 5.80 1880 4.88 1879- 80 5.07 1881 12.64 1880- 81 2.66 1882 10.27 1881- 82 4.35 1883 4.57 1882- 83 4.08 1884 4.47 1883- 84 6.53 1885 3.01 1884- 85 5.88 1886 4.27 1885- 86 4.32 1887 10.69 1886- 87 2.08 1888 4.25 1887- 88 3.34 1889 11.50 1888- 89 8.98 1890 10.66 1889- 90 5.14 1891 3.93 1890- 91 5.75 1892 4.05 1891- 92 5.56 1893 9.95 1892- 93 2.50 1894 3.12 1893- 94 3.24 1895 6.14 1894- 95 2.26 1896 9.33 1895- 96 5.38 1897 8.66 1896- 97 3.06 1898 7.46 1897- 98 3.32 1899 5.66 1898- 99 4.64 1900 3.03 1899-1900 2.87 1901 7.43 1900- 01 5.66 1902 4.14 1901- 02 1.06 1903 5.80 1902- 03 6.23 1904 6.12 1903- 04 1.08 1905 5.85 1904- 05 14.74 1906 4.78 1905- 06 7.17 1907 10.88 1906- 07 7.77 1908 7.92 1907- 08 4.01 1909 7.19 1908- 09 4.13 1910 7.22 1909- 10 2.88 1911 7.58 1910- 11 4.42 1912 6.68 1911- 12 3.76 THE MONSOON CLIMATE OF ARIZONA AND NEW MEXICO. 13 the rainfall at high altitudes is influenced much less than at low. As yet, data obtained on the subject are not sufiBcient to permit of any trustworthy conclusions. The matter is mentioned here merely as one of the many interesting problems which would repay investigation. Another problem of the same kind is illustrated in figures 2 and 3. The curve of figure 2 represents the total rainfall by years from 1868 to 1912 as given in the Summary of the Climatological Data for the United States and in the Monthly Weather Review. Figure 3 shows the summer rainfall for the six months from May to October inclusive and the winter rainfall for the six months from November to April during the same term of years. The latter two curves seem to indicate a reciprocal relation of some sort between the rainfall of summer and winter. In general, when the summer rains increase in amount the winter rains decrease, and vice versa. This phenomenon is not confined to Tucson, but is apparently characteristic of the Southwest as a whole. For instance, in the two curves at the lower left-hand corner of figure 11, on page 109, it is clearly seen in the rain¬ fall of Flagstaff, 200 miles north of Tucson and 5,000 feet higher. Inasmuch as a similar relation between the rainfall of equatorial and -temperate regions has been inferred from Summer rainfall Winter rainfall Fig. 3.—Winter and Summer Rainfall at Tucson, Aiizona, 1808-1912. - Winter rainfall, Nov .-Apr. -- Summer rainfall, May-Oct. 68 1870 72 74 76 78 1880 82 84 86 88 1890 92 94 96 98 1900 02 04 06 08 1910 Fig. 4. —Comparison of 3-year Means of Winter and Summer Rainfall at Tucson, Arizona, 1868-1912. the comparison of records in various parts of the world, particularly India, it is of great interest to find it so clearly manifest here. Examination of the curves shows that in two cases out of every three a minimum of winter rain is followed by a maximum during the succeeding summer. One would expect to find the reverse also true, and that a summer maximum would be followed by a winter minimum, but this does not hold good. A summer minimum, however, is usually followed by a winter maximum. In other words, if it be permissible to generalize on so small a basis of fact, the minima appear to be the critical points. A maximum, either in summer or winter, is not likely to be followed by especially marked conditions in the succeeding season. A minimum, on the contrary, whether in summer or winter, is likely to be followed immediately by a maximum in the succeeding season. The preceding generalization obviously holds good only about two-thirds of the time. In figure 4 the summer and winter curves have been smoothed by using 3-year means instead of the actually observed rainfall. When the minor fluctuations are thus eliminated, the opposed phases of the summer and winter curves are brought out clearly in the period from 1868 to 1887, and less clearly from 1895 to 1907. In the period from 1888 to 1894 14 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. the diagram presents a wholly different appearance: the two curves show agreement instead of opposition. The effect of such agreement upon the economic life of the country is marked. During the late eighties, when both summer and winter rains were on the increase, the cattle industry flourished as at no other period. In the early nineties, how¬ ever, when the rain of both seasons decreased, dire distress prevailed. Cattle died by the thousand and the industry received such a blow that on many ranches there are now only hundreds of animals where then there were thousands. The peculiar fashion in which the summer and winter curves show opposite phases part of the time, and then suddenly agree, suggests various speculations as to the cause. It looks as if there might be more than one type of cyclical or periodic variation in the activity of the earth’s at¬ mosphere. One type perhaps causes agreement and one type disagreement. Here, as in the case of the contrast between the precipitation of high and low regions, it is too early to attempt to form positive theories. CHAPTER II. THE TOPOGRAPHIC INFLUENCE OF ARIDITY. In the modern science of physiography as developed in recent years, under the leader¬ ship of Professor Davis, one of the most interesting features is the connection between the form of the earth’s surface and the climate of any given region. Not only do glaciated areas possess their own peculiar topography, but so do humid and dry regions. The scenery of Arizona and New Mexico is stamped indelibly with the impress of an aridity which has lasted hundreds of thousands of years. Just when it began we can not tell, but certainly far back in the Tertiary era, and possibly earlier, for deposits characteristic of aridity not only attain a great thickness superficially, but are interbedded with marine strata in formations dating far back in geological time. A full discussion of the effects of aridity upon the form of the land in all parts of New Mexico and Arizona would require a volume and would demand an amount of field work far greater than I have been able to give to the matter. Accordingly, in the following pages I shall limit myself to a few salient features which clearly show evidences of aridity, or are of special importance in relation to changes of climate and the ancient human occupation of the country. Topographically Arizona and New Mexico consist of two chief parts, plateaus of nearly horizontal strata 5,000 to 7,000 feet high and basin regions where mountain ranges, due to faulting or to rapid uplift of relatively small areas, alternate with more or less completely inclosed basins filled with alluvial waste. In Arizona the plateaus and the basin ranges are sharply separated by the Mogollon Escarpment, a line of southward-facing cliffs which extend approximately northwest and southeast across nearly the whole State and pass almost through its center. North of the escarpment lies a high plateau broken in places by fault scarps running north and south, diversified by extinct volcanoes and cut by deep canyons, like that of the Colorado, but preserving almost uniformly the practically level position of its rock formations in spite of thousands of feet of uplift since their original deposition. South of the escarpment the basin-range region lies at a general elevation 3,000 or 4,000 feet less than that of the plateaus. Here the strata by no means lie hori¬ zontal, but have been tipped this way and that, chiefly by means of block faulting along lines running more or less closely north and south. The spaces intervening between the uplifted blocks form basins which have been filled with alluvium. Thus to the eye of the traveler the difference between the plateaus and the plains may be briefly summed up by saying that the plateaus are a region of great plains cut by deep canyons, while the basin- range country is composed of great plains broken by narrow mountain ranges. In New Mexico the separation between the plateaus and the basin ranges is not so distinct as in Arizona. In the elevated regions of the northwest, however, and in the Staked Plains of the eastern part of the State the plateau quality is as well marked as in Arizona, while basins and ranges of mountains due to faulting are almost as characteristic a feature in the south as in the neighboring State to the west. In the center, especially toward the north near Colorado, the main chain of the Rocky Mountains extends down into New Mexico and adds a distinct type of topography. The mountains, however, soon break up into isolated ranges rising from the plateau or bordering waste-filled basins, so that most of the country may fairly be said to belong to one of the two main types with which we are dealing. Inasmuch as the main chain of the Rockies has little to do either with the early inhabitants or with the other evidences of climatic changes with which we are here con¬ cerned, it will not be further discussed. 15 16 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. THE TOPOGRAPHIC FEATURES OF THE PLATEAUS. (1) Mature Uplands .—Where most typically developed the plateaus present three chief types of topographic form, which may be described as mature uplands of ancient origin, young plains of erosion upon soft strata, and young cliffs composed of hard strata and forming the borders either of mesas or canyons. Other features, such as volcanic cones or fault scarps, for example, may be omitted as of secondary importance in spite of their great interest. The plateaus, it is needless to say, were formed by the slow uplifting of large areas of the earth’s surface without any pronounced tilting or bending of the rocks. In all such cases an old topography, brought to a greater or less degree of maturity, must have been carried up to a height far above that under which it was originally developed. In some cases—for example, the Kaibab Plateau in northern Arizona just north of the part of the Grand Canyon most commonly visited—this ancient topography is still pre¬ served. On the edges it is being rapidly dissected and removed by the rapid streams which are the normal result of uphft. The Mescalero Plateau, east of the Otero Basin in the south central part of New Mexico, is another good example. Here a steep fault scarp, gashed by precipitous young canyons, rises on the east side of the basin to a height of about 9,000 feet, nearly 5,000 feet above the basin floor. At the top one emerges from the narrow valleys formed since the last uplift and finds himself in a wooded region of open, mature topography. Gentle slopes rise from broad valleys to round-topped hills of nearly uniform height. Everywhere the soil is deep, and outcrops of naked rock are rare. Often the valleys converge into flat sink-holes, where the water stands for a while before it can seep away through underground passages in the soluble limestone. Everything indicates that the region was subjected to extensive erosion long before it was slowly upheaved to its present situation. Its topography was formed under conditions quite different from those of to-day, and we can as yet draw no satisfactory conclusion as to the climate prevalent during the long ages required for its erosion. (2) Young Plains due to Erosion .—The mature uplands are in most cases so elevated as to be too cold for extensive habitation or agriculture. On their borders, however, the processes of erosion have in many cases given rise to broad and relatively youthful plains of subaerial denudation at altitudes of 6,000 or 7,000 feet. These would be habitable if provided with more water, and many of them seem to have been cultivated in former times. The plains are rarely smooth for any great distance. At frequent intervals they are interrupted by steep-sided mesas, hues of cliffs, or canyons, the product of the same pro¬ cess of erosion which has produced the plains. It is unnecessary here to enter into any detailed description of this well-known process. I would merely call attention to the fact that it reaches a high state of development only in arid regions. Where strata of unequal hardness are exposed to erosion, such soft materials as shales are worn back much faster than hard formations, such as massive sandstones or limestones. If the strata are horizontal the weathering of the soft formation tends to carry it away from under the hard formation wherever a vertical surface is exposed by erosion. The hard rocks of course break off as soon as they are undermined, and thus steep cliffs are formed. This process takes place in a moist chmate quite as much as in a dry, but it can not go so far. In the moist climate two things tend to check it. In the first place, the action of frost, rain, snow, and vegetation tends to cause the weathering of the hard rocks to go on at a rate which approximates that of the soft rocks more nearly than in dry regions. Hence rela¬ tively more talus falls from the cliffs of moist regions than from those of dry regions, and the tops of the cliffs are worn back, while the soft strata at the base are protected by the accu¬ mulation of d6bris. Hence steep cliffs are not common. In the second place, erosion is less hindered in dry regions than in wet. The torrential character of the rains and the absence of vegetation allow the talus to be carried rapidly away in arid countries, while the barren- THE TOPOGRAPHIC INFLUENCE OF ARIDITY. 17 ness and dryness of the surface allow the wind to etch out the soft rocks in a fashion quite unknown in moist lands. Consequently, where strong contrasts of hardness exist in a dry climate the soft rocks may be worn back for miles, leaving the underlying hard rocks to form broad plains of erosion, while the remnants of the overlying hard rocks form mesas. Where the climate is moist the sharp contrast between the hard rocks and the soft is diminished, as we have seen. Moreover, the number of residual hills of hard rock is likely to be large because of the abundance of streams and consequent minute dissection. Thus the plains of erosion are apt to be more broken by hills than in dry regions, while the slopes are gentler because more masked by talus. (3) Cliffs bordering Mesas and Canyons .—The origin of the steep cliffs of the plateau country is evident from what has just been said. The uplifting of the plateaus has caused rapid erosion and the swift deepening of valleys. The differences between hard and soft strata have resulted in a benched topography; the hard layers form cliffs while the soft wear back so as to form benches on top of the hard. Where the chffs wear back far from the streams, leaving plains, the hard formations may still retain their steepness, and thus mesas and buttes arise. For our present purpose this is important, partly because such topography is characteristic of arid regions, and still more because of its relation to human occupation. The ancient chff-dwellers, who figure so largely in American archeology, made most of their dwellings in narrow canyons just at the point where the lowest soft layer makes a hollow under the overlying hard layer. Starting, probably, with no shelter except that of the cliffs overhanging their wind-scoured caves, they gradually learned to dig caves in soft formations such as the volcanic tuff of the Pajarito Plateau near Sante Fe in northern New Mexico, while later they developed the art of building walls in front of the caves, and these in turn led them to build entire rooms, sometimes three or four rows deep, at the base of the chffs. Others among the ancient Americans utilized these same cliffs for protection, building their houses of stone on the tops of great steep-sided mesas, of which the Mesa Verde is the best known. Many of the ancient inhabitants, as may be seen near the remarkable ruins of the Chaco Canyon in northwestern New Mexico, dwelt at the base of the cliffs, but apparently cultivated the plains of erosion high above their heads. This last matter is still in dispute, but there can be no question that the peculiar topography characteristic of arid plateaus was the warp upon which was woven one of the most interesting of all the phases of pre-Columbian American civilization. THE BASIN REGIONS. (1) The Mountain Slopes .—Going down from the plateaus to the basin regions of the south, we find a country where, during the most ancient times, men dwelt as numerously as in the plateaus, although the remaining ruins are less conspicuous. Here, as there, three chief elements of physiographic form dominate the landscape: (1) rough, rocky mountain slopes, usually of steep ascent; (2) gently sloping piedmont deposits of gravel merging imperceptibly into smooth plains and playas of fine silt; and (3) terraces com¬ posed of alluvium, chiefly in the form of gravel. For convenience I shall not attempt a general description of these elements, but shall describe them as they occur in the region of Tucson, where much of our future investigation will center. This will serve as well as a more general description, for in all essential matters there is little difference between the various parts of the basin region. Near Tucson the mountain slopes, the first of our three physiographic elements, form the sides of irregular ranges scattered here and there like islands in the midst of a sea of gravel and silt. In general the mountains run northwest and southeast. They vary in height from 4,000 to 9,000 feet, while the plains lie at an altitude of 2,000 to 3,000 feet, dimin¬ ishing to the west and increasing to the east in New Mexico. Some ranges, such as the Santa Catalinas northeast of Tucson, the Tortolitas farther to the north, and the Sierritas 3 18 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. to the southwest, are of disordered structure and consist of masses of granites and gneisses flanked by sedimentary rocks of Paleozoic or later age. The majority of the ranges, however, are composed of Paleozoic sedimentary or metamorphic rocks, together with later lavas. Most are fault blocks which have been uplifted on the southwest side of lines of faulting running northwest and southeast, and have been tilted in such a way that the back of the block slopes toward the southeast. The structure is not regular, for there has been a large amount of secondary faulting. As none of the faulting is recent, the mountains are maturely dissected. This does not mean that sharp forms of peak and cliff are rare. On the contrary, many of the fault-block ranges are carved into the most striking forms, and all the mountains display a great amount of naked rock. The little Tucson Range, for instance, which lies just to the west of Tucson, and is composed largely of andesite and other eruptives, presents one of the most jagged sky-lines to be found anywhere in America, a striking sight against the clear sunset sky. The Sawtooth Moun¬ tains, a few miles to the west, are of the same structure, and are, if anything, still more jagged. The granitic mountains, on the other hand, are not characterized by prominent peaks. From a distance they present the appearance of great solid masses, but near at hand are seen to be full of splendid deep canyons, often with precipitous walls of naked rock. The rockiness of the mountains speaks strongly of arid climatic conditions. Mountains in a similar stage of dissection in a moist climate would be covered with soil and would present graded slopes for the most part. In Arizona the slopes are largely washed bare of soil because lack of moisture restricts the growth of plants and prevents the accumulation of roots and fallen leaves which would hold the soil in place when heavy showers tend to wash it down. The truth of this statement is apparent from the fact that the low mountains, under 5,000 feet or so in height, are more rocky and on the whole more rugged than those which rise higher. The high mountains, such as the Catalinas, which, as we have seen, rise to a height of 9,000 feet, enjoy a much greater rainfall than the lower portions of the country, at least twice as much apparently. They are also cooler, so that evaporation is far less active than in the hot regions of lower elevation. Accordingly the supply of moisture available for plants is far in excess of that below, and the mountains above 5,000 feet are covered with forests. At the lower levels oaks and bushy trees of the smooth- barked manzanita and its allies prevail, while, higher up, the mountains are densely clothed with splendid forests of juniper and pine. In the mountains of moist lands the amount of soil commonly decreases from the bottom upward. In southern Arizona the case is different; from the base of the hills, at an elevation of approximately 3,000 feet, the amount of soil decreases in the normal fashion at first, but after 1,000 or 2,000 feet it begins to increase, and at a height of 6,000 or 7,000 it is much greater than at the base. Such con¬ ditions can occur only in an arid climate among mountains rising high enough to receive a considerable rainfall. (2) The Bahadas, or Piedmont Gravel Deposits .—The second element in the landscape in the basin region is the vast accumulation of gravel, sand, and silt which flanks the mountains on every side. This accumulation of detrital material slopes gently away, mile after mile, becoming flatter and flatter, until many of the slopes merge into level playas. The name "bajada” has been applied to such slopes by Tollman.* The Span¬ iards use the word “bajada” to designate any sort of descent, including the process of descending, but in the absence of any other appropriate term in English, I feel constrained to adopt it. The word is pronounced “bahadtha,” the sound of the d being neither d nor th exactly. The a’s have the French sound and the accent is on the second syllable. In defiance of all rules I venture to write the word with an h instead of a j, because otherwise it is sure to be mispronounced. Genetically it belongs to the same class as mesa, hutte, arroyo, playa, and others in conunon use. * C. F. Tollman: Erosion and Deposition in Southern Arizona Bolson Region. Jour. Geol., vol. xvii, 1909, p. 142. THE TOPOGRAPHIC INFLUENCE OF ARIDITY. 19 The bahadas consist primarily of innumerable detrital fans deposited by the streams at the point where they issue from the mountains. In moist countries such fans can not attain large dimensions, for they are soon washed away by the steady flow of the streams. In dry regions, on the contrary, they tend constantly to increase in size. None but the largest streams are permanent; for the great majority come to an end soon after leaving the constricted valleys of the mountains. Emerging from the uplands, their speed is checked so that they deposit their load of waste and are divided into many distributaries. Thus fans are formed in whose thirsty gravel most of the water is lost, while the remainder runs on a few miles farther with constantly diminishing volume until it finally spreads out into thin sheets, forming playas which soon evaporate. Except in the case of occasional floods which reach the main streams and run through to the sea, every bit of material that most of the streams bring down from the mountains is deposited in the lowlands. Thus year by year and century by century the fans grow in size, and finally coalesce into what appears to be a single great slope, a vast apron or glacis surrounding all the mountains, and ever rising higher as the mountains themselves are worn lower. In time the waste from the higher mountains may bury the lower ones, cutting them off at first and forming the gravelly passes which make it so easy to cross the minor ranges at frequent intervals. As time goes on, many small mountains are so buried that they merely stick up as little pointed buttes in the midst of a rising sea of gravel and silt. Doubtless in past ages many hills have disappeared entirely, for the deposits washed down from the mountains to the lowlands have a depth of over 1,000 feet not far from Tucson, as shown by the records of wells dug by the Southern Pacific Railroad.* Close to the mountains the bahadas consist of coarse material in the form of subangular boulders with a matrix of cobbles and sand. Farther out, as the slope decreases, the boulders disappear, although in some cases they are washed to a distance of 5 miles or more. Then the cobbles diminish in size and finally vanish, leaving only gravel, and that in turn gradually gives place to the fine sand and silt which alone are found in the playas where the slope is reduced almost to zero and the waters come to rest. The bahadas, playas, and half-buried mountains of the southwestern part of the United States reproduce exactly the topographic forms of other deserts in distant regions, such as Syria, Persia, and western China. In aU parts of the world these great piedmont deposits preserve full records of the cHmatic vicissitudes to which they have been subject. Manifestly the nature of the materials laid down under various conditions of climate is bound to vary, even though a certain degree of aridity may have prevailed at all times. If the mountains were at some time denuded of trees by excessive drought, a great amount of soil must have been washed down in ensuing years. If the amount of vegetation became greater than now, and the streams became more constant by reason of greater rainfall, deposition at the immediate base of the mountains must have diminished, while farther away it must have increased. Thus the depths of the bahadas must preserve a record of all manner of changes. In the present volume this subject will not be taken up, because it does not bear upon our immediate problem of recent climatic changes, but evidently any com¬ prehensive study of the climatic conditions of the geologic past demands a careful examina¬ tion of complete sections from the bahada slopes not only of America, but of all parts of the world. (3) The Terraces .—The bahadas by no means always merge into playas, nor do they universally coalesce with one another. In fact, they usually fail to do so. Once all the bahadas coalesced smoothly and merged into playas or flat valley bottoms, but now their smooth slopes come to an end in terraces and are constantly interrupted by small valleys and gullies of recent origin. These valleys may be just wide enough for a small torrential stream, or several miles wide. Their depth may be a few feet or hundreds. Their sides * Cam. Inst. Wash. Pub. 99, 1908. 20 THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. may show an unbroken slope, gentle or steep as the case may be, or may be broken into four or five terraces. Practically every waterway, large or small, is bordered by one or more terraces. They form the third of the persistent elements of the landscape. Not so noticeable as the rough mountains, not furnishing a home and land for tillage to the ancient inhabitants like the bahadas, they are in some ways quite as important. Their interpreta¬ tion, unlike that of the other features, is by no means a matter of general agreement. Therefore, when we have briefly discussed the vegetation of the country, I shall devote a chapter to a consideration of the two opposing theories, climatic and tectonic, which have been advanced in explanation of the terraces. ' w^' #■ >■ r f* \ m 'n ■ * ' '<^‘ ti^: •S^^'r f' i ^*'4 '(■9 V5r>" ., ^ Iff. -.^,' ,.; ^- fcs'' ■ ’^3 ^JfcvV! ♦ • t -- m r*.i .V - L-S-':T^. If. I'i - ^ - 1},-' ^ ., ’Si . t'.* fr • ■' ■'-■ 'v;^* ' ,7.sA* r^ Lu\ » rV r% »-j r»v Fv. iO»'.f* r «•€ . A..* ^ p(^S!^i5^iEk^■;,,/.*-