LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class ELEMENTARY METEOROLOGY FOR HIGH SCHOOLS AND COLLEGES BY FRANK WALDO, PH.D. LATK JUNIOR PROFESSOR IN THE UNITED STATES SIGNAL SERVICE, MEMBER OF THE AUSTRIAN AND GERMAN METEOROLOGICAL SOCIETIES, AUTHOR OF "MODERN METEOROLOGY," ETC. NEW YORK-:- CINCINNATI-:. CHIC AGO AMERICAN BOOK COMPANY COPYRIGHT, 1896, BY AMERICAN BOOK COMPANY WALDO'S METEOR. PREFACE. METEOROLOGY, which treats of our atmosphere, is a department of the physics of the globe, or " geo-physics," as it is called. It is only within recent years that meteorology has been elevated to the position of an independent science. The more apparent atmospheric conditions have been the sub- ject of observation and comment for many hundreds of years, but only within the past two or three centuries have accurate observa- tions and trustworthy records been accumulated. These observa- tions, which have greatly increased in number and accuracy during the present century, and especially during the past twenty years, are the groundwork which serves^ as a basis for the inductive development of the science of meteorology ; and they also serve as a criterion for the testing^ o'f : the truths evolved by the deduc- tive method which has been developed by mathematical physicists almost entirely within the last quarter of a century. It has been the custom to offer to English-reading students of meteorology the development of the subject from the inductive standpoint ; and in this elementary treatise the same form has been in the main adopted. In the deductive treatment it would be necessary to start with initial conditions, density, temperature, moisture content, and motion of the air, and then follow out the regular and the irreg- ular changes of possible occurrence, keeping in mind the relation between all of these conditions and the effect of one on another. Such a treatment, while certainly more logical, and on this account more desirable, is vastly more difficult, than the method adopted, and should be offered to mature minds only, which can grasp very complex processes. It was therefore decided to give merely the facts and their probable explanations, in treating the subject of the atmospheric conditions ; furthermore, the elements have as far as possible been mentioned separately, in order to avoid the confusion of a more complex treatment, and the better to isolate the obscure and uncertain parts of the subject. In treating of the atmospheric movements, the author has, how- ever, combined the two methods of presentation. 3 .(. PREFACE. The best example that we have of a non-mathematical deductive treatment of meteorology is Ferrel's " Popular Treatise on the Winds," in which that author has given the most complete account of the atmospheric circulation and allied phenomena to be found in any language, and all teachers and advanced students of meteor- ology are earnestly recommended to read Ferrel's book. In order to keep this book within proper limits, the author has been obliged to omit much matter which should find place in a complete treatment of the subject. Thus but the barest mention could be made of the simpler forms of meteorological apparatus ; and the application of meteorology to the sciences and arts, such as hygiene, agriculture, engineering, and to manufacturing and commerce, has been entirely excluded. The scope of this book must not, therefore, be misinterpreted by meteorologists on the one hand, or by teachers on the other. It is intended to serve as a text-book of the elements of the science for general students, and must not be considered as a manual for practicing meteorologists. Quantitative results have been given as far as possible in round numbers. The English systems of measurements are given, as necessary in a book intended for immediate and widespread use in our educational institutions. In the preparation of this book the author has drawn on his experience as a former teacher of meteorology in the school of training for officers and observers of the Weather Bureau of the United States Signal Service ; and he has tested much of the sub- ject-matter herein presented (although in somewhat greater detail) in a course of lectures delivered at Evelyn College for Young Women, at Princeton. The author is much indebted to Lieutenant Everett Hayden, [J.S.N., for suggestions and assistance in reading the final proof sheets. Many of the diagrams in the book have been redrawn from other printed sources, both foreign and American : among the latter must be mentioned in particular Ferrel's " Popular Treatise on the Winds " and Greely's "American Weather." FKANK WALDO. PKINCKTON, X.J. CONTENTS. [. THE EARTH'S ATMOSPHERE II. TEMPERATURE 19 Heat and Solar Radiation . . . . . .19 Thennometry . . 31 Observed Air Temperatures ...... 34 Distribution of Air Temperatures over the Earth . . 50 Temperatures below the Earth's Surface .... 70 III. AIR PRESSURE 73 Barometry .......... 75 Observed Air Pressures ....... 80 Distribution of Air Pressures over the Earth ... 88 IV. WINDS ioi Classification of Winds ....... 102 Observations of Wind Direction 103 Observations of Wind Velocity ...... 106 V. MOISTURE: VAPOR, CLOUD 118 Atmospheric Moisture. . . . . . . .118 Atmospheric Moisture as Vapor . . . . .122 Observed Atmospheric Humidity . . . . .124 Atmospheric Moisture as Cloud and Fog . . . .129 Clouds. . . . . . . . . . -130 Observations of Cloudiness . . . . . . . 137 5 6 CONTENTS. CHAPTER PAGE VI. MOISTURE: PRECIPITATION 142 Condensation . . . 143 Observations of Rainfall 145 Distribution of Rainfall over the Earth .... 148 Hail and Snow . . .159 Evaporation -163 VII. ATMOSPHERIC OPTICS AND ELECTRICITY 166 Atmospheric Optics 166 Atmospheric Electricity 175 VIII. GENERAL CIRCULATION OF THE ATMOSPHERE .... 180 General Air Motions . 180 Primary Circulation of the Atmosphere .... 187 IX. SECONDARY CIRCULATION OK THE ATMOSPHERE . . . 213 Cyclones . . . . . . . . . .216 Anticyclones .......... 234 X. LOCAL AND MISCELLANEOUS 'WINDS 241 Tornadoes .......... 241 Thunderstorms ......... 249 Spouts ........ . . . 259 Periodic Local Winds 262 Miscellaneous Winds ........ 263 XI. WEATHER AND WEAIIIER PREDICTIONS 269 Weather Conditions . 269 Weather Predictions -274 XII. CLIMATE 293 Climatic Conditions ........ 293 Climates of the Continents. ...... 307 XIII. CLIMATE OF THE UNITED STATES ...... 313 Climatic Subdivisions of the United States . . . . 317 Distribution of the Climatological Elements over the United States INDEX ELEMENTARY METEOROLOGY FOR HIGH SCHOOLS AND COLLEGES BY FRANK WALDO, PH.D. LATE JUNIOR PROFESSOR IN THE UNITED STATES SIGNAL SERVICE, MEMBER or THE AUSTRIAN AND GERMAN METEOROLOGICAL SOCIETIES, AUTHOR OF " MODERN METEOROLOGY," BTC. OF THE UNIVERSITY OF NEW YORK- :-CINCINN ATI- : CHICAGO AMERICAN BOOK COMPANY COPYRIGHT, 1896, BY AMERICAN BOOK COMPANY WALDO'S METEOR. PREFACE. METEOROLOGY, which treats of our atmosphere, is a department of the physics of the globe, or " geo-physics," as it is called. It is only within recent years that meteorology has been elevated to the position of an independent science. The more apparent atmospheric conditions have been the sub- ject of observation and comment for many hundreds of years, but only within the past two or three centuries have accurate observa- tions and trustworthy records been accumulated. These observa- tions, which have greatly increased in number and accuracy during the present century, and especially during the past twenty years, are the groundwork which serves as a basis for the inductive development of the science of meteorology ; and they also serve as a criterion for the testing of the truths evolved by the deduc- tive method which has been developed by mathematical physicists almost entirely within the last quarter of a century. It has been the custom to offer to English-reading students of meteorology the development of the subject from the inductive standpoint ; and in this elementary treatise the same form has been in the main adopted. In the deductive treatment it would be necessary to start with initial conditions, density, temperature, moisture content, and motion of the air, and then follow out the regular and the irreg- ular changes of possible occurrence, keeping in mind the relation between all of these conditions and the effect of one on another. Such a treatment, while certainly more logical, and on this account more desirable, is vastly more difficult, than the method adopted, and should be offered to mature minds only, which can grasp very complex processes. It was therefore decided to give merely the facts and their probable explanations, in treating the subject of the atmospheric conditions ; furthermore, the elements have as far as possible been mentioned separately, in order to avoid the confusion of a more complex treatment, and the better to isolate the obscure and uncertain parts of the subject. In treating of the atmospheric movements, the author has, how- ever, combined the two methods of presentation. 3 4 PREFACE. The best example that we have of a non-mathematical deductive treatment of meteorology is Ferrel's " Popular Treatise on the Winds," in which that author has given the most complete account of the atmospheric circulation and allied phenomena to be found in any language, and all teachers and advanced students of meteor- ology are earnestly recommended to read Ferrel's book. In order to keep this book within proper limits, the author has been obliged to omit much matter which should find place in a complete treatment of the subject. Thus but the barest mention could be made of the simpler forms of meteorological apparatus ; and the application of meteorology to the sciences and arts, such as hygiene, agriculture, engineering, and to manufacturing and commerce, has been entirely excluded. The scope of this book must not, therefore, be misinterpreted by meteorologists on the one hand, or by teachers on the other. It is intended to serve as a text-book of the elements of the science for general students, and must not be considered as a manual for practicing meteorologists. Quantitative results have been given as far as possible in round numbers. The English systems of measurements are given, as necessary in a book intended for immediate and widespread use in our educational institutions. In the preparation of this book the author has drawn on his experience as a former teacher of meteorology in the school of training for officers and observers of the meteorological depart- ment of the United States Signal Service ; and he has tested much of the subject-matter herein presented (although in somewhat greater detail) in a course of lectures delivered at Evelyn College for Young Women, at Princeton. The author is much indebted to Lieutenant Everett Hayden, U.S.N., for suggestions and assistance in reading the final proof sheets. Many of the diagrams in the book have been redrawn from other printed sources, both foreign and American : among the latter must be mentioned in particular Ferrel's " Popular Treatise on the Winds " and Greely's "American Weather." FRANK WALDO. PRINCETON, N.J. CONTENTS. CHAPTER PAGE I. THE EARTH'S ATMOSPHERE . . . .. . -7 II. TEMPERATURE . . .19 Heat and Solar Radiation 19 Thermometry .31 Observed Air Temperatures . . . . -34 Distribution of Air Temperatures over the Earth . . 50 Temperatures below the Earth's Surface . . . -7 III. AIR PRESSURE -73 Barometry . -75 Observed Air Pressures . 80 Distribution of Air Pressures over the Earth ... 88 IV. WINDS 101 Classification of Winds 102 Observations of Wind Direction 103 Observations of Wind Velocity . . . . . . 106 V. MOISTURE: VAPOR, CLOUD 118 Atmospheric Moisture 118 Atmospheric Moisture as Vapor 122 Observed Atmospheric Humidity 124 Atmospheric Moisture as Cloud and Fog . . . .129 Clouds. ...... ... 130 Observations of Cloudiness . . . . . . . 137 5 6 CONTENTS. CHAPTER PAGE VI. MOISTURE: PRECIPITATION ....,, 142 Condensation .. ... ... 143 Observations of Rainfall . . . . , .145 Distribution of Rainfall over the Earth . 148 Hail and Snow 159 Evaporation . . . . . .163 VII. ATMOSPHERIC OPTICS AND ELECTRICITY 166 Atmospheric Optics 166 Atmospheric Electricity . . . . . . -175 VIII. GENERAL CIRCULATION OF THE ATMOSPHERE . . . .180 General Air Motions .180 Primary Circulation of the Atmosphere . . . .187 IX. SECONDARY CIRCULATION OF THE ATMOSPHERE . . 213 Cyclones' 216 Anticyclones 234 X. LOCAL AND MISCELLANEOUS WINDS 241 Tornadoes 241 Thunderstorms 249 Spouts .... 259 Periodic Local Winds 262 Miscellaneous Winds ........ 263 XI. WEATHER AND WEATHER PREDICTIONS 269 Weather Conditions ........ 269 Weather Predictions ........ 274 XII. CLIMATE .... 293 Climatic Conditions . . . . 293 Climates of the Continents . 307 XIII. CLIMATE OF THE UNITED STATES 313 Climatic Subdivisions of the United States . . . -317 Distribution of the Climatological Elements over the United States 321 INDEX . 365 \ \ ELEMENTARY METEOROLOGY, CHAPTER I. THE EARTH'S ATMOSPHERE. Meteorology, strictly speaking, is the science which treats of the condition of the atmosphere, its changes in condition, and the causes which give rise to these condi- tions and changes. The science is as yet but partially developed, and much that is at present accepted as fact will be modified by future investigations. Air. The gaseous envelope which immediately sur- rounds the earth, and is called the atmosphere, is com- posed chiefly of air, just as the ocean is composed of water. Air is a mechanical mixture of oxygen and nitro- gen gases and a small amount of carbon dioxide, together with argon and traces of various other chemical substances which have little to do with the science of meteorology as now studied. Such air is called dry air. Atmospheric air always contains more or less vapor of water, which varies in amount from a very small quantity to about 5% of the amount of the dry air. Dry Air at the earth's surface consists essentially of nearly 21% of oxygen, about 78% of nitrogen, about i% of argon, and 0.03 % of carbon dioxide, by volume. 7 8 ELEMENTARY METEOROLOGY. Oxygen. There is a slight decrease in the proportion of oxygen at distances of several miles above the sea level ; and even near sea level the amount of oxygen is not absolutely invariable, but the variations are very slight. Oxygen is the most important element of the air. Nitrogen, which forms the greater part of the air, seems to have no special function except to dilute the oxygen. It is slightly lighter than oxygen. Argon enters into such close combination with pure nitrogen, that it has always, until its recent discovery, been included with the atmospheric nitrogen. Its peculiar function has not yet been ascertained. It is the most dense of the gaseous constituents of the air. Carbon Dioxide, which is a compound of oxygen and carbon, varies slightly in amount, being about 3 parts in 10,000 of air on the average ; but there seems to be a little more in cloudy than in clear weather, and a little more at night than by day. Carbon dioxide, though it forms such a small proportion of the air, is a very important factor in the life on the earth, since it forms the basis of food. All animal organisms inhale oxygen, and exhale carbon dioxide. Carbon dioxide is frequently emitted as a gas from volcanoes, in such quantities as to render the air into which it passes too impure for breathing purposes, espe- cially in low-lying adjacent places into which the gas sinks on account of its relatively great density. Carbon dioxide is of no direct use to animals after it has been exhaled in the breathing process. It is, however, of much importance to plant life ; for the carbon is the chief food of plants. The carbon dioxide of the air is taken up by certain cells of the plant, and is there decomposed by the sunlight into carbon and oxygen. The oxygen escapes THE EARTH'S ATMOSPHERE. 9 into the air, but the carbon is retained to build up the tis- sues of the plant. Microscopical Impurities in the Atmospheric Air, Minute solid particles of matter are found in the air, and these may be divided into two classes, inorganic and organic. The inorganic particles are the dust particles, which we see floating in a ray of sunlight as it crosses a darkened room ; and the smoke particles, which in the aggregate are visible as smoke. The dust and smoke particles are of great importance in the changing of the water vapor in the atmosphere into drops of water. The number of dust particles in the air varies enormously at different times and places. The number is least and most constant in clear weather, and greatest and most variable in cloudy weather ; it is least over the ocean and at great distances above the surface of the earth ; and the size of the particles in general decreases with the height above this surface. On the top of a high mountain, about a mile above sea level, it was found that there were 11,000 dust particles per cubic inch when the air was clear ; but the number increased to 50,000 or 65,000 during the temporary fog produced by the passage of an isolated cloud over the mountain top. The organic particles in the air are the minute germs Called microbes. These are of two classes, the bacteria and the molds. The bacteria are the minute forms of animal life, such as the disease germs ; and they are most frequent in the air of hospitals, and least frequent over the oceans and on mountain tops. The molds are minute forms of vegetable life (fungi) which occasion fermenta- tion and the decomposition of organic matter. They ar^e most numerous in foul damp air, such as that found in sewers. Organic particles vary in number with the sea- son of the year and the hours in the day. 10 ELEMENTARY METEOROLOGY. Density and Volume of Gases. Gases are elastic, and they may be compressed into a small space ; but if pres- sure is removed, they expand, and occupy larger space. With a constant temperature, the volume varies inversely as the pressure : that is, if the pressure is doubled, the volume is reduced one half. This relation is maintained for the air, within the limits of meteorological study, but for very great and very small pressures it does not hold good. The pressure on any point in a quiet gas will be the same in all directions, upwards, downwards, and towards either side. The air of the atmosphere is acted on by the gravity which draws it towards the center of the earth, and the amount of this downward force is the weight of the air. When we consider the outside pressures on a quantity _P r of free air, such as may be represented by the cube shown in Fig. i, then the pressure P t (from without) will just equal the pressure P r ; but the upward pressure P b will exceed the downward pressure P a by the weight of the air, that is, by the downward pull g exerted by gravity on the cubic mass of air. Then the pressure P b = P a + g. Likewise P a = P b g. The side pressure P t = ihe side pressure P r \ but these side pressures at the upper edge of the cube are each equal to P a , and they increase, by the pressure corresponding to the weight of g, to P b at the lower edge of the cube. Pressure and Weight of the Air. The weight of the at- mosphere, or its downward pressure on any surface, is the result of the aggregate weight of all the air above. It is evident, then, that as we go upward from the earth's sur- face towards the outer limit of the atmosphere, the weight of the air, and consequently its downward pressure, will decrease ; and this decrease will be in proportion to the weight of the air left below in the ascent. If the air were incompressible, then the decrease in weight and density THE EARTH'S ATMOSPHERE. II would be the same, for the same distance left behind in ascending, at any height in the atmosphere. The air, however, is compressible; and the greater the weight of the air above, the more the air below is compressed, and the denser it becomes. Thus, if we start from the outer limit of the atmosphere, where there is supposedly no air above, the first layer of air will be attracted by the force of gravity, but there will be no additional weight of air press- ing down from above ; but when we descend to the second layer of air, then gravity exerts its power on this second layer just as it did on the first, and in addition the weight of the upper layer presses down on the second layer, and compresses it, making it denser. The air of a third layer will be still more compressed, because it will have the weight of the two upper layers pressing down on it in addition to the attraction of gravity. Thus the air in- creases in density all the way down to the earth's surface by the compression due to the weight of the air above. The Elastic Force of a Gas in a given quiet condition, or its expansive force, is equal to the surrounding pressure exerted on it. The different gases of the atmosphere have different elastic forces for the same quantity of matter; but for any gas the elastic force is increased if the quan- tity and density of the gas are increased, provided the temperature remains unchanged. The elasticity of a gas increases as the gas grows warmer, and is measured by its pressure in the given or existing con- dition. For measurements of the atmospheric pressure and of the separate gases which are in the air, the pressure or weight of a column of mercury under given conditions is used; the pressure being measured by the length of the mercury column, in inches, which is necessary to counter- poise the expansive force of the gas. The instrument used 12 ELEMENTARY METEOROLOGY. for such measurements, and called the barometer, is ex- plained in the proper connection farther along. Arrangement of the Gases forming the Air. Each of the gases forming the air has a different degree of elas- ticity, and the same quantity of each occupies different volumes under similar outside pressures : so that, the greater the elastic force (pressing outward), the less will the gas be compressed for a given pressure from the out- side ; and consequently its density will be the less, since a smaller quantity or mass will suffice to neutralize the outside pressure. In the case of our atmosphere, the pressure at any height above sea level is due to the weight of the mass of the gas above : consequently the elasticity or pressure, and the resulting density, decrease with the height, because, the higher up we go, the less gas there is to press down from above. If the air were replaced by a single gas, then the greater the elasticity of this gas, the higher it would extend above the surface of the earth. Each of the various gases forming the air has the same distribution and arrangement of its parts which it would have if it alone composed the atmosphere, so that the gases with the least density extend upward the farthest. Of the gases nitrogen, oxygen, and carbon dioxide, nitro- gen is the least dense, and consequently extends up far- ther than oxygen, and much farther than carbon dioxide and argon, which are most dense. It has been calculated, from our knowledge of the density of the various gases at the surface of the earth or sea level, and the rapidity with which the density of each decreases with ascent above this surface, that the carbon dioxide practically disappears from the atmosphere at a height THE EARTH'S ATMOSPHERE. 13 of about 10 miles above the sea level, the water vapor practically disappears at a height of 12 miles, the oxygen gas disappears at a height of 30 miles, and the nitrogen gas disappears at a height of 35 miles. Observations indicate, however, that the atmosphere ex- tends much higher than this, which shows that our knowl- edge of the laws of atmospheric gases is still incomplete. Extreme Limit of the Air. It is probable that there is no definite outer limit to the atmosphere, where we can say that the air ceases and there exists only the ether which is supposed to occupy all space. We have seen that theoretically there should be no air at an elevation of 40 miles ; but meteors have been observed at a height of loo miles or more, and these are supposed to be rendered luminous by the friction of their passage through the air. Peculiar luminous clouds have also occasionally been observed at altitudes of upwards of 40 or 50 miles ; and if, as is supposed, they are due to crystallization of water vapor, then moisture also exists at much higher elevations than the present theories will allow. Weight of the Air. The actual weight of a quantity of air is variable according to its density. The greater the density, that is, the more the air is compressed, the greater is its weight for any specified volume. At sea level, water weighs about 840 times as much as the same bulk of air, a cubic foot of air weighing about i| ounces. With in- crease of height above the sea level, a cubic foot of air weighs less and less. The Atmospheric Air Mass Relative to the Earth. The diameter of the earth is about 7,900, and the area of its surface 197,000,000 square miles. Since the atmosphere does not, according to our most delicate tests, extend above IOO miles from the earth's surface, then its entire thick- 14 ELEMENTARY METEOROLOGY. ness is only about -fa of the radius of the earth ; but the lower part of the atmosphere is so much denser than the upper part, that one half the total weight of the atmos- phere is limited to the first three miles above the sea level, while the other half comprises the remainder of the atmos- phere to its outer limit. The thickness of the lower layer, containing half the total amount of air, is, then, less than ToVo f tne ra dius of the earth. If the earth were repre- sented by a globe 100 feet in diameter, this lower half of the air covering its surface would have a thickness of less than J of an inch. The atmosphere, then, must be re- garded as a thin gaseous layer spread out over a very large spheroid. Distribution of the Air over the Earth. The atmos- phere has not only length and breadth, covering the whole earth's surface, but it has also thickness, extending from the earth's surface up to some unknown height. The geographical terms latitude and longitude can be used for expressing the horizontal extent of the atmosphere, or for locating a mass of air at any point on the earth's surface ; but when some point above the earth's surface is to be con- sidered, then it is necessary also to express distances in the direction of the thickness of the atmosphere, which is done by altitudes. Altitude. By the altitude of a place, or of a point in the atmosphere, is meant its height above the level of the ocean. Altitudes are expressed in feet, yards, or miles in the English system of measurements, and by meters or kilometers in the metric system. Altitudes are also sometimes expressed in direction merely, without ref- erence to the absolute elevation of the point or object to be located. This is accomplished by means of angular measure, in which the altkude is the angle between a line THE EARTH'S ATMOSPHERE. 15 drawn from the observer to the object, and another line drawn from the observer to the unobstructed horizon. The angular altitude of the horizon is o ; and of the zenith, or point directly overhead, 90. Importance of Altitude in Meteorology. With increase or decrease of height above sea level, that is, with vari- ation of altitude, the atmospheric air undergoes rapid and marked changes in its condition. At places on the ocean the altitudes at which man lives are approximately the same ; but on the land this is not the case, for there we find him inhabiting regions at all altitudes from the low lands up to mountain tops reaching 15,000 feet above the sea level. Altitude thus becomes a very important matter in meteor- ology. Latitude. The heat from the sun is the chief cause of variations in atmospheric conditions : and since the aver- age position of the sun is the plane of the equator, where the average angular altitude of the' sun is greatest, and the point on the earth where the suri has the least average angular altitude is at the poles, the effect of the sun is felt most at the equator, and least at the poles ; and the various gradations between these extremes are measured by any scale marking the distance between the equator and the poles. Such a measure we have in the degrees of latitude, which divide the distance into 90 equal parts. But since the degrees of latitude are counted as increas- ing from the equator towards the pole, the latitude of the equator being o, and that of the pole being 90, then, as the latitude increases, the influence of the sun on the atmosphere decreases. Longitude. In the direction transverse to latitude, that is, in the direction in which we use longitude in geograph- ical measurements, there is no such absolute permanent 1 6 ELEMENTARY METEOROLOGY. variation in the atmospheric conditions brought about by the solar heat ; and so the degrees of longitude cannot be used in expressing a law of physical change such as has just been indicated for the degrees of latitude. But still, longi- tude has its usual geographical significance in locating points on the earth's surface, the meteorological conditions of which it may be desired to consider. Longitude, however, is used in another way, and that is, as representing time. We know that the rotation of the earth on its axis causes the sun to appear to travel around the earth in twenty-four hours. Now, this distance around the earth is 360 of longitude, so that the apparent motion of the sun is through 1 5 of longitude in one hour. The sun, then, in pursuing its apparent course in the heavens, gets back to the same position every twenty-four hours. For any one place there is thus a return to practically the same conditions, so far as the sun is concerned, at the expiration of each twenty-four hours, no matter what changes may have been experienced in the intervening hours. Such a circuit is called a cycle. Meteorological Elements. The different items by which the total meteorological condition of the atmosphere (at any place) is represented, are called the meteorological elements. These are, The temperature of the air, or its degree of heat. The pressure of the air,, or its amount or quantity, and density. The humidity, or amount of water contained in the air. The precipitation, or amount of water which the air loses as rain or snow. The evaporation, or amount of water which the air takes up from the earth. The wind y or the movement of the air. THE EARTH'S ATMOSPHERE. I/ The clouds, or the obscuration of the sky. The electrical and optical conditions of the air. Meteorological Conditions. Observations of the meteor- ological elements show that they do not long remain in the same condition, but are continually undergoing changes. These changes are of two classes, regular or periodic, and irregular or accidental. A periodic change is one in which the element returns to substantially the same condition after the lapse of periods of time of approximately uniform length. The principal periods are the diurnal or daily, depending on the rotation of the earth upon its axis ; and the annual or yearly, de- pending on the revolution of the earth about the sun. An accidental change is one which occurs at irregular intervals, and whose time of occurrence cannot be foretold. The average condition of any meteorological element dur- ing a given time is obtained by finding the sum of a number of single observations of its condition made during this time (usually at equal short intervals of time), and dividing this sum by the number of observations. In addition to the average condition, it is also of interest to know the extreme fluctuations which may occur in the meteorological elements during both periodic and acci- dental changes. The maximum amount is that which the phenomenon attains when it reaches its greatest value. The minimum amount is that which it attains when it reaches its least value. The amplitude is the difference between the maximum and the minimum amounts. Special conditions are called phases. The various phases of meteorological conditions are investigated as regards the time of occurrence as well as the amount or quantity. Meteorological Instruments. In order to observe with the required accuracy the meteorological conditions, it is WALDO METEOR. 2 1 8 ELEMENTARY METEOROLOGY. necessary to have instruments properly adapted for mak- ing measurements of the different meteorological elements. These instruments are of two classes. The first class is used by an observer for the direct observation of 'condi- tions. The second class is used for obtaining an automatic registration of conditions by mechanical or photographic means : these instruments are usually of very complicated construction. The Distribution of Meteorological Elements is the^ condi- tion of the meteorological elements at different localities during the same phases or periods of time. It is usually presented by means of geographical charts on which are entered the conditions, each at its proper locality on the map. Statical Meteorology treats of the conditions of the me- teorological elements without considering the changes in the air due to its motion oi translation. It therefore em- braces the whole matter of the constitution of the atmos- phere, and such conditions as have the question of space, but not of time, enter/ing into them ; as, f or^ instance, -the average values of the meteorological elements for any period, or their condition at any chosen instant. c Dynamical Meteorology treats of / the motions of the atmospheric air, their causes, and the conditions arising therefrom, and also of the modification which these motions cause in the. statical conditions. CHAPTER II. TEMPERATURE. Heat is due to very rapid motions of the minute particles, called molecules, of which bodies are composed. Heat is not a substance, but is a form of energy. It may be trans- ferred from one body to another. Heat of various degrees or intensities may exist, and these intensities are measurable. Temperature is a general term applied to express the intensity or degree of heat. We thus say the temperature of the body is high or low, according as the body is hot or cold. The instrument used for making accurate measure- ments of the temperature of bodies is called a thermometer, or heat measurer. Diffusion of Heat. The process by which heat is trans- ferred from one body to another, or from one part to another part of the same body, is called the diffusion of heat. The diffusion of heat always takes place by the transference of heat from a hotter to a colder body, or from a hotter to a colder part of the same body. The dif- fusion of heat is accomplished by conduction, convection, radiation, or reflection, or by some combination of these. Conduction is the flow of heat from the hotter to the colder places in an unequally heated body or in adjacent bodies. The thermal conductivity of a .body is its capacity for the conduction of heat through it. The conduction of heat does not take place with the same facility or rapidity in all bodies. The metals are good conductors, silver espe- 19 2O ELEMENTARY METEOROLOGY. dally, as may be realized by putting one end of a silver spoon into hot water, and holding the other end in the hand. Air, stone, water, ice, snow, wood, and wool are poor conductors. The reason why woolen clothing is so desirable in winter is that it is a poor conductor, and so the heat of the body is retained. Convection is the transference of the heat by the circula- tion of the hot body itself, by which it is brought succes- sively in contact with colder bodies. In this process the final transference of heat from the one body to the other usually takes place by conduction. Thermal convection takes place in fluids. Radiation is the transference of heat from a hot body through a medium which does not itself become much heated in the process, but which must be colder than the hot body. The diathermancy of a body is its capacity for the transmission of radiant heat without itself becom- ing heated. In the passage of heat through a body, some of the heat is always retained by the body, and this retention is called thermal absorption. Reflected Heat is the heat immediately thrown off from the surface of a body without entering it, when it is receiv- ing heat by radiation from another body. Radiant Energy. Energy is radiated from the sun, and is transmitted to our earth by a vibratory process, through the ether which fills space. The slower vibrations are rendered sensible to us as heat, and the more rapid ones as light. Heat of the Atmosphere. In meteorology we have to deal principally with the temperatures at and near the earth's surface. There heat is received from two principal sources, viz., the sun, and the interior of the earth itself. The heat from the sun we call the solar heat, and that from the TEMPERATURE. 2 1 earth terrestrial heat. The combined action of the solar and terrestrial heat is what we measure when we obtain the temperature of the air with which we are usually brought into actual contact. If we descend considerably below the surface of the earth, as into a mine, the influence of the heat of the earth itself is more strongly felt ; while at and above the earth's surface the heat from the sun is more noticeable. The solar heat is of most importance in meteorology, and it is the solar heat which maintains the animal and vegetable life on the earth. Solar Heat and Solar Radiation. The sun is a very large body, about 880,000 miles in diameter, and it is in- tensely hot. The space around the sun is very cold, and, according to the law of the transmission or diffusion of heat, the sun must lose a portion of its heat in the endeavor to equalize its own heat and that of the surrounding space. Heat is transmitted radially (that is, straight out, like the spokes from the hub of a wheel) from the sun through this space by radiation ; and, unless some matter interposes, these solar rays as they are called will keep on indefinitely. Our earth, in its revolution around the sun, intercepts less than one half of a billionth of the whole amount of heat given off by the sun, yet the amount received is amply sufficient for the purpose of sustaining animal and vegetable life on the earth. Although the sun is steadily parting with such vast quantities of heat, the most careful observations fail to show that it is be- coming appreciably cooler. Revolution of the Earth around the Sun. The earth moves around the sun in one year in a slightly elliptical path or orbit, the sun being not in the center, but in one of the foci of the orbit. The earth is therefore at times slightly nearer to the sun than at other times. The 22 ELEMENTARY METEOROLOGY. average distance of the earth from the sun is about 92,890,000 miles. On Jan. i, when the earth is nearest the sun, it is 91,300,000 miles from it; and on July i, when it is farthest away, it is 94,450)000 miles from it. If the earth moved around the sun in a circular path, it would receive a constant total amount of heat and light during every day of the year; but, as it is, the amount received on Jan. i is 7 % greater than that received on July i. It requires 365.2422 days, or nearly 365^ days, for the earth to make a complete revolution around the sun. Its path lies in a plane called the plane of the ecliptic, Inclination of the Earth's Axis of Rotation. The earth has a rotation once in 24 hours around an axis passing through its center. This axis passes through the north and south poles of the earth, and is inclined at an angle of about 66| to the plane of the ecliptic, or plane of revolution of the earth around the sun ; so that, when the earth revolves around the sun, this axis does not stand perpendicular to this plane of revolution, but is inclined 23^ to this perpendicular. This is a most important mat- ter in the formation of meteorological conditions, because it controls or marks out, to a great extent, the manner and places in which the solar rays reach the earth's surface. Effects of the Inclination of the Earth's Axis on the Dis- tribution of Solar Rays on the Earth's Surface. If the earth's axis were perpendicular to the plane of the ecliptic, the length of the days and nights would be equal, and the plane of the equator would coincide with that of the eclip- tic ; so that during the entire year the apparent path of the sun would lie along the equator. But with the con- stant inclination of the axis the following conditions result during a revolution of the earth around the sun : TEMPERATURE. 23 FIG. 2. PORTIONS OF THE EARTH RECEIVING THE SoLAR RAYS> The sun shines over a whole hemisphere at all times ; but the part of the earth receiving the solar rays varies with the position of the earth in its orbit. The earth has the position B (Fig. 2) on Dec. 21, C on March 21, D on June 21, and A on Sept. 22. We can consider that when the sun is over the earth's equator, and con- sequently the earth's po- lar axis is turned 90 from the sun, it is in its normal or average posi- tion. On Dec. 21 the earth stands so that the , , i ,. north pole is turned far- ther away from the sun by the full inclination of the pole to the perpendicular to the ecliptic, or at an angle of 23^, and the sun shines on the south pole and 23^ beyond it (B). On June 21 the opposite extreme is reached ; and the north pole is turned towards the sun, and the south pole away from it, by this same angle, and the sun shines on the north pole and 23^ beyond it (D). Halfway between these dates the condi- tions become as shown at A and C, where the region receiving the solar rays reaches from pole to pole. In the daily rotation of the earth when in the positions A and C, the sun appears to move through the sky in the plane of the equator. In the position D it appears to move in the plane of the Tropic of Cancer (23^ north latitude). In the position B it appears to move in the plane of the Tropic of Capricorn (23|- south latitude). The change from the position A to B, thence to C, and thence to D, and back again to A, takes place gradually, little by little each day, as the earth moves around the sun. ELEMENTARY METEOROLOGY. Periods of Presence and Absence of Solar Radiation at Places on the Earth's Surface. Only half of the earth can receive the solar rays at any time, and, as we have just seen, the region receiving these rays varies constantly as the earth moves forward in its orbit. When the solar rays reach from pole to pole, at the points A and T(Fig. 2) the days and nights are of equal length (12 hours long) all over the world. During the time when the earth is passing from A to C through B, the solar rays reach beyond the south pole, but fail to reach the north pole by a like amount, and the daytime is lengthened, and the nighttime shortened, in the southern hemisphere ; and for a varying region, reaching a maximum of 23^ at B, there is uninterrupted day around the south pole, and uninterrupted night around the north pole. Similarly, when the earth is passing from C to A through D, the daytime is longer and the nighttime shorter in the northern hemisphere ; and in a variable region, reaching a maximum at D, there is continual day around the north pole, and continual night around the south pole. The days are longer than the nights in the polar hemi- sphere which is receiving the most solar rays ; but at the equator the days and nights are always the same. The following table shows the greatest possible length of the con- tinued or uninterrupted visibility of the sun at various latitudes on the earth : Latitude Hours of visibility 12* I 7 i3 h 41 '5 h 49 i6 h 63 20 h 665' 24" 672i' i mo. 6 9 5 i 2 mo. 78! i' 4 mo. 90 6 mo. The twilight, which is the transition period between daylight and darkness, also increases in length with the distance from the equator. Thus, of the 8,766 hours which make up a year, there are, TEMPERATURE. 25 AT THE EQUATOR. AT THE POLES. 4,407 hours day. 445 hours day. 864 hours twilight. 2,403 hours twilight. 3,495 hours night. 1,913 hours night. The Apparent Motion of the Sun around the Earth. The various phenomena just described as due to the rotation of the earth on its axis, the inclination of the axis to the ecliptic, and the revolution of the earth around the sun, are best; realized by considering the matter as it appears to us. The sun appears to revolve around the earth once in 24 hours, and on March 21 this revolution takes place in the plane of the equator ; but the sun has a slow apparent movement towards the north at the same time that it moves around the earth ; and the result of these two motions is a spiral movement, which carries the sun across the successive parallels, so that by June 21 it revolves around the earth on the parallel of 23 north latitude. During this time the sun has apparently revolved around the earth 91 times. After this date there is a spiral motion southward, which gradually carries the sun back to the plane of the equator on Sept. 22. The sun then crosses the equator into the southern hemisphere, and pursues its spiral course until it reaches the plane of 237 south latitude on Dec. 21, when it recedes again towards the equator, reaching it on March 21. As the sun thus revolves around the earth (very nearly in the planes of successive par- allels), it always rises in the east, reaching the zenith at noon, and sets in the west, of places along the parallel of latitude on which it happens to be. Thus, on March 21 and Sept. 22, the sun is directly in the zenith of the equator at noon ; at noon on Dec. 21 it is directly in the zenith at latitude 23^ south, and on June 21 at latitude 23^ north. Amount of Solar Heat which would reach the Earth's Surface in the Absence of the Atmosphere. The sun's heat which would reach the earth's surface, were there no atmosphere, would vary in intensity at different places with the angle at which the rays struck the earth's sur- face. When the sun's rays lay parallel with the surface, which would be the case for any special locality on the atmosphereless earth when the sun was at the horizon, the 26 ELEMENTARY METEOROLOGY. amount of solar heat received at that locality would be zero. When the sun's rays struck the surface squarely, which would be the case when the sun reached the zenith, the intensity would reach its greatest value. In Fig. 3, the rays received between c and d are greatest, and those between a and b or e and /are least, in intensity. At the time of the equinoxes, when FIG. 3 . RELATIVE IN- the sun is directly over the equator, TENSITY OP SOLAR RAVS. surface at the equator with their maximum strength ; but at the poles they would be parallel to the surface, and therefore their effect there would be zero : and the heat- ing power of the sun's rays would decrease, then, from the equator to the poles, or, as it is usually expressed, would decrease inversely as the latitude. When the sun is over the equator, if the amount of solar heat at the equator were put at i.oo, then at the pole it would be o.oo. But since the sun departs from the equator towards the south 23^ at one season of the year, and a like amount towards the north at another season of the year, and since the distance of the earth from the sun does not re- main a constant quantity, it becomes a very complex matter to calculate the relative amounts of heat actually received at points on the earth's surface during any 24 hours. With the increase of latitude there is a lengthening of the number of hours a day during which the sun shines in summer, and this makes the total amount of heat received in higher latitudes during the daytime much greater than one would suppose. Thus, when the sun reaches its greatest altitude at the poles, and shines there continually, the amount of heat received there during 24 hours is over 20% more than that received during 24 hours at the equator, when the sun is above the horizon for but half the time. If the solar radiation during the average day at the equator, which we will call the thermal day, were taken as a unit, then 365.24 would repre- sent the annual solar radiation at the equator ; and the annual solar radia- c TEMPERATURE. 2J tion at other latitudes expressed in these same thermal days would be as follows : Latitude . . . Thermal days . 365-2 10 360.2 20 34S- 2 30 321.0 40 288.5 50 249.7 60 207.8 70 173.0 80 156.6 90 151.6 It must be distinctly remembered that these are only relative values. The absolute quantity of solar heat which the earth would receive dur- ing the year has been calculated to be one and one half quadrillion heat units, 1 or enough to melt an ice layer over 141 feet thick, and to evaporate a layer of water nearly 18 feet deep, covering the whole earth's surface. Solar Constant. It has been found by 'experiments in recent years, that, unaffected by the atmospheric absorption and radiation, the intensity of the solar rays falling vertically is such as to heat 3 grams of water i C. (1.8 F.) per minute over each square centi- meter of exposed surface. This number 3 is called the solar constant. Solar Heat reaching the Earth's Surface in the Presence of the Atmosphere. When the influence of the atmos- phere on the solar heat is considered, the computation of the amount of heat reaching the earth's surface becomes very complicated on account of the absorption of part of the heat by the air. The amount of absorption depends principally on the length of the path of the rays through the air, the amount of moisture present, and the degree of cloudiness. For a perfectly clear or cloudless sky when the sun is in the zenith of any point, about 75% of the solar radiation will reach the earth's surface at that point, and 25% will be absorbed by the atmosphere. But when the sun is not in the zenith, that is, when the altitude of the sun is less than 90, then the thickness of the air layer through which the solar rays must pass will be increased, and consequently the loss of heat will be greater ; while, when the sun is just 1 The unit of heat is the amount necessary to raise the temperature of one cubic centimeter of water one degree Centigrade. 28 ELEMENTARY METEOROLOGY. above the horizon, nearly all of the heat will be absorbed by the atmosphere, and almost none will reach the earth's surface at the point of observation. The following table shows the proportional thickness of the air layers through which the solar rays pass at different angular altitudes of the sun, taking the vertical thickness of the atmosphere as a unit or i.oo; and it also shows in per cent the amount of the solar radiation which reaches the earth's surface, taking the amount received from the zenith on the outside of our atmosphere as i.oo. Altitude of the sun . . . '. c; 10 20 q o C 70 00 Thickness of the atmosphere in units qt t IO 2 c e6 2 no I QQ I 31 I 06 Amount of solar radiation reach- ing the earth O OO O OCC O 2O The solar rays possess different qualities, and are ab- sorbed in different proportions in their passage through the air, according to their color, or position in the spec- trum. The ultra-violet and violet rays are absorbed the most ; next come the indigo, blue, green, yellow, orange, red, and ultra-red in the order named. The sun, therefore, has a more bluish appearance, the less the thickness of the atmosphere through which the solar rays pass. The heat rays are much more readily absorbed than the light rays ; so that, when the sun is near the horizon, we may receive some of the light rays, but almost none of the heat rays. The intensity of the sunlight is therefore not a direct measure of the total amount of solar radiation. In the middle latitudes in which we live, in wholly clear weather, about half of the daily solar radiation is absorbed by the air; that is, we receive only half the amount of heat that we should receive if there were no air, or if the air had no effect on the solar heat. TEMPERATURE. 29 The solar rays which are stopped by the air are not wholly lost to us, but we receive a portion of them by radi- ation from the air itself, and from the particles of matter in solid and fluid form which the air contains. But the great- est thermal use of the atmosphere is to prevent the rapid radiation of heat from the earth's surface off into space. It has been calculated, that, if there were no atmosphere to check the radiation of heat into space, the temperature of the earth's surface would be about 325 F. below zero. Solar Radiation on the Northern and Southern Hemi- spheres. In the northern hemisphere, during the period when the sun is north of the equator, the intensity of the solar radiation is slightly less than that in the southern hemi- sphere when the sun is south of the equator, because in the former case the earth is at a greater distance from the sun, and the intensity of the solar rays decreases rapidly as this distance increases. The summers on the southern hemi- sphere are consequently slightly warmer, and the winters slightly colder, than those of the northern hemisphere, so far as the intensity of the solar heat enters into the matter. In other words, the solar climate, as the climatic influence of the sun is called, undergoes slightly more extreme changes in the southern than in the northern hemisphere. For the whole year, however, the northern and southern hemi- spheres receive a like amount of solar heat. Distribution and Transference of Solar Heat on the Earth's Surface. The solar rays which reach the sur- face of the earth are partly absorbed by the substance on which they fall, and partly thrown off again by reflection into the air. The effect is different for the two sub- stances, land and water. Both of them are warmed by the addition of heat ; but the rigidity of the land and the 30 ELEMENTARY METEOROLOGY. mobility of the water, and the greater heat capacity of the latter, make the separate mention of each necessary. The Heat received by the Land warms it very much at the surface. Part of the heat is slowly conducted by the ground to the next layers of earth below ; but it does not penetrate to a depth of many feet, because the land is such a poor conductor of heat. Another part of the heat is communicated to the air layer nearest the ground, and by means of convective air currents it is transported to other localities ; the change being accomplished either by the transfer of the heated air as a whole, or else by the heated air mingling with colder air and imparting heat to it. Another portion of the heat of the surface ground is lost through outward radiation into space. Another part of the heat is lost from the land by reflection. A small portion of the heat passes upward from layer to layer of air by means of conduction alone ; but this is an exceedingly slow process, because air is such a poor con- ductor of heat. The Heat received by the Water Surface warms it but slightly ; for convection currents at once arise where the water has not everywhere the same warmth, and the heat is transferred by means of these currents to cooler por- tions of the water. Heat is also imparted by the water surface to the air, and is radiated outward into space as from a land surface, although more feebly, because the surface of the water is cooler; but the amount of heat returned to the air by reflection from the water surface is a very important part of the whole amount communicated to the air. Some of the heat received at the surface passes through the water for a considerable distance by continued direct radiation, and is gradually absorbed by the lower layers of the water. TEMPERATURE. A considerable quantity of heat is also used up in the process of evaporation from the water surface. Principle of the Thermometer. In general, the addi- tion of heat to a body causes it to expand, and the loss of heat causes it to contract. Equal increments of heat cause an equal expansion, and equal losses of heat cause equal contractions, in the same body. The amount of heat added or lost can, then, be measured by the expansion or contraction which some adopted substance undergoes when subjected to these changes. For various practical reasons, mercury is the substance chosen for the expansive or thermometric substance. It does not congeal, or become solid, in the coldest weather in most inhabited countries, and does not vaporize in the hottest weather ; and, moreover, its expansion and contrac- tion, due to small changes in the amount of heat to which it is subjected, are relatively great as compared with those of other metals, and can therefore easily be seen. FIG. 4. THERMOMETER For convenience in readily determining BuLB AND STEM ' the amount the mercury has expanded or contracted, it is placed in a small hollow glass bulb with a connecting small bored tube as an outlet (Fig. 4). The bulb is filled with mercury at some rather low temperature, and any increase of heat will cause the mercury to flow out into the capillary tube. The larger the bulb and the smaller the bore of the tube, the greater will be the amount of rise of the mercury in the tube for a given increase in the amount of heat. Thermometer Scales. In order to determine just how 32 ELEMENTARY METEOROLOGY. far the mercury moves in the thermometer tube when it is subjected to changes of temperature, it is necessary to have some kind of a scale for such measurements. If all thermometers had the same dimensions, we could use a scale of inches ; but this cannot ordinarily be used, because thermometers are of various sizes. There are several kinds of thermometer scales in use, and the divisions of the scale are called degrees, which are of the same length in each individual instrument. The scale divisions are marked either on the glass tube or stem of the thermometer or on a metal or glass strip placed adjacent to it. Graduation of a Thermometer. There are two temper- atures, which, because of the comparative ease with which they may be determined, have been adopted as the basis in graduating thermometer scales. These are the temper- ature of freezing and that of boiling water. These tem- peratures are called the fiducial points of a thermometer. In marking the scale of a thermometer, the position of the mercury at these two points is ascertained by holding the instrument first in freezing, and then in the steam from boiling water (under normal atmospheric pressure at sea level). The graduation is then completed by marking off the distance between these points in equal subdivisions. There are three systems of thermometer graduation in use in different part's of the world, the Fahrenheit, the Centigrade or Celsius, and the Reaumur. In the Fahrenheit thermometer there are 180 divisions or degrees between the fiducial points ; the freezing point is marked 32, and the boiling point 212. In the Centigrade thermometer there are 100 divisions or degrees between the fiducial points ; the freezing point is marked o, and the boiling point 100. TEMPERATURE. 33 In the Reaumur thermometer there are but 80 divisions or degrees between the fiducial points ; the freezing point is marked o, and the boiling point 80. The graduation of standard thermometers is carried below the freezing and above the boiling points to some distance, to permit of the determination of extreme temperatures. Temperatures below the o point are marked with a minus sign; those above are considered as plus, but this sign is usually omitted. The Reaumur scale is seldom used outside of Germany at present, and is therefore not further described. Centigrade degrees are converted into Fahrenheit de- grees by the following formula : - Cx i.8 + 32 = F. Fahrenheit degrees are converted into Centigrade degrees as follows : F- 3 2 1.8 Fig. 5 represents a mercurial thermom- eter with an attached brass scale show- ing degrees Fahrenheit. The temperature reading is 70. Effect of Change of Temperature on the Air. Starting with the condition of the air at the temperature of freezing water (which, as we have seen, is constant, and is 32 on Fahrenheit's thermometer scale, and on the Centigrade scale), the in- FIG. 5 .- MERCURIAL THERMOMETER. crease in volume within the limits of nat- ural temperatures is ^| T of itself for each degree of tem- perature on the Fahrenheit scale, and ^^ for each degree on the Centigrade scale. WALDO METEOR. 3 34 ELEMENTARY METEOROLOGY. Thus, if a column of air 491 inches long at a temperature of 32 F. has its temperature raised i F., the air will expand to 492 inches in length ; and if the temperature is lowered i F., it will contract to 490 inches in length. And following this law, if the temperature were lowered 491 F., then the length of the column of air would be reduced to nothing ; according to which, we could not have a temperature lower than 491 F. below the freezing point of water, and this would be nearly 460 F. below zero of the Fahrenheit scale. This assumes that the air would retain its gaseous form down to this low temperature. As a matter of fact, all of the constituents of the air would liquefy and solidify before such temperatures were reached, and so would cease to obey the law of expansion and contraction of gases, on which these results are based. Temperatures of about 375 F. have been produced by artificial means, and air has been liquefied and solidified at a somewhat higher temperature when it has first been made very dense by compression. Of the constituents of the air, oxygen is most easily, and nitrogen least easily, liquefied and solidified. Observations of Temperature. Temperature observa- tions of the air should be made in the shade, and at about six to ten feet above grass-covered earth ; but in the cities the thermometers are frequently exposed from upper windows or on the roofs of buildings. Observations of temperature of the air such as these, have been made at intervals, during longer or shorter periods, at thousands of places scattered over the earth's surface. Those on the land have generally been made at fixed localities, while those on the ocean have been made mostly on moving vessels. The observed air temperatures over the globe present a wide variety of phases, which vary with the latitude, altitude, and environment of the various localities considered. Temperatures also vary with the time ; that is, they do not remain constant at any one place. Fluctuations of Temperature. The sun in its daily course, due to the rotation of the earth, gives rise to ditirnal fluctuations of the temperature ; and in its yearly TEMPERATURE. 35 course, due to the revolution of the earth combined with the inclination of the earth's axis, it gives rise to the gradual changes which make up the seasonal fluctuations of the temperature. There are also long-period fluctuations in the temperature, which extend over many years. Seasonal changes in temperatures have been quite care- fully studied at nearly all parts of the surface of the earth ; but the changes during the day have been carefully studied at a few places only. Long-period fluctuations have been only roughly investigated. In addition to these regular fluctuations of temperature, there are irregular or accidental fluctuations, due to causes which will be mentioned later. The temperature at any place at the earth's surface de- pends principally on the length of time during which that point is exposed to direct solar radiation, and on the angle at which the sun's rays reach the place. Anything, therefore, which tends to interrupt the duration of solar radiation, or change the angle of inclination of the solar rays, must cause changes in the temperature of the point reached by those rays. Temperature Changes during One Rotation of the Earth. The changes of temperature during the diurnal rotation of the earth are of two classes, the regular or periodic changes, due to this rotation, which vary with the different conditions under which the solar radiation is received in different parts of the earth's orbit ; and the irregular or accidental changes, due to the oceanic or continental loca- tion, the varying degrees of cloudiness, and the movement of air masses by the winds. These irregular changes of temperature are least in the equatorial regions, but increase towards the poles, and in middle and higher latitudes are frequently of greater 36 ELEMENTARY METEOROLOGY. magnitude than the regular periodic changes. Their causes are mentioned later (p. 45 and elsewhere). The Regular Diurnal Change of the Temperature is as follows : Beginning at sunrise, there is a gradual increase in the temperature with the increasing altitude of the sun, until the highest or maximum temperature is reached shortly after the time when the sun reaches its highest altitude in the sky ; then, as the sun's altitude decreases, the tem- perature decreases, until (about) the time when the sun rises again to repeat the course of the previous day. There is thus a single highest and a single lowest tempera- ture during the 24 hours. The maximum or highest temperature of the 24 hours is reached at the time, after the sun has attained its greatest altitude, when the amount of heat lost from the earth by radiation just equals the amount received from the sun. This occurs at various times, ranging from 13 hours ( I P.M.) for the ocean exposures to 15 or 16 hours (3 or 4 P.M.) over the continents. The air is radiating heat into space, and is receiving heat from the sun, during the day. When the temperature of the air is increasing, it shows that the rate at which the heat is received from the sun is greater than that at which it is lost by radiation ; and it will keep on growing warmer as long as this continues, which is until a few hours after noon. But when the heat received and that given off by radiation become equal, the temperature remains stationary ; and when the loss of heat is more rapid than the gain, there is a fall in the temperature. The maximum temperature attained over the land is greater than that on the ocean ; and the hour when the amount of heat received and lost is equal becomes later in proportion to this temperature, which TEMPERATURE. 37 accounts tor the earlier occurrence of the maximum over the ocean. The hour of occurrence does not vary much with the season of the year. The degree of cloudiness and especially during the hours about noon greatly influences the maximum temperature. With an increase in the cloudiness, there is a decrease in the maximum temperature, because the solar rays cannot reach the earth's surface ; and there is a retardation of the time of maximum, because the radiation of heat from the earth is retarded. The minimum or lowest temperature of the 24 hours occurs at, or just a few minutes before, sunrise over the continents, and a little earlier than this over the ocean. The time of occurrence is, then, at about 6 hours (6 A.M.) at the equator ; and with increase of latitude, it becomes earlier in the summer half year, and later in the winter half year. In latitude 80 the min- imum occurs at about 2 A.M. during the summer half year. The minimum temperatures are lower over continents than on the oceans, because the land loses its heat more rapidly than the water. Under like conditions of surroundings, the minimum temperatures get absolutely lower with increasing distance from the equator towards the pole ; but they do not descend lower below the average temper- ature of the place considered. This last descent is greatest at the interior of deserts in the equatorial regions. The effect of cloud is to prevent the minimum temperatures from becoming as low as they would under a clear sky, for the radiation of heat from the earth is hindered. The amplitude of regular oscillation of the diurnal tem- perature (or the difference between the extreme maximum and minimum during the 24 hours) is in general greatest at the equatorial regions, and decreases towards the poles, for the same exposure ; that is, over the land or over the water. The amplitude is greatest over continents, and least over oceans, varying, in extreme cases, from about 32 F. for continental exposure in equatorial regions, to 2 or 3 F. in polar regions. ELEMENTARY METEOROLOGY. The amounts of amplitudes follow quite closely the variation in the meridian altitude of the sun and the relative length of daytime to night- time ; and consequently the amplitudes are greatest in summer, and least in winter. . The amplitudes decrease with increase of altitude above the ground, and approach in magnitude those obtaining for an oceanic exposure. The immediate surroundings of a locality strongly affect the amount of the amplitude of daily temperature changes. If the amplitude for a plain or level surface is taken as the normal condition, then, in general, on a convex surface (a hill or mountain) the amplitude is diminished, and on a not too great concave surface (a valley) the amplitude is increased. In the case of the hill, the heat rays are dispersed, while in the valley they are collected. The daily amplitude is greatest on clear days, and least on cloudy days, because the maximum tem- perature is higher, and the minimum lower, on the clear day. Representation of Diurnal Temperatures. The daily course or march of the tempera- tures (and of the other elements) from hour to hour is represented by meteorologists in sev- eral ways: (i) merely by numbers, which may be only the ac- tual temperatures, or may be the average temperature for the day, minus the individual hourly tem- peratures ; (2) by a graphical process (Fig. 6); and (3) by means of a mathematical formula, called Bessel's Formula, which expresses this curve mathematically. _- . i >i g a , s" *" ,~~ ^ 1 -. Lk ec in / V r ""^ -^ /J / ~~~* -^ S K ' '/ I ', [3 ^0 J'o / s^ / ^ / F a / 5 \ 1 \ 1 s^ X.. F 1 40 ^ 1 6 t. PC U s / *. *s 1 2 A 3* 0*1 1*1 2*1 3*1 4*1 6*1 8*1 0*2 9*2 I 2 2*23*24 FIG. 6. DIURNAL CHANGE OF AIR TEMPERATURE. TEMPERATURE. 39 The graphical process is the simplest and best method for illustrating clearly the rise and fall of the temperature indicated by the thermometer readings. A piece of paper is divided into small, equal squares by vertical and horizontal lines. The consecutive vertical lines represent consecutive hours, commencing either at o h or midnight, or at i h ; the hours are marked in a horizontal row below the diagram, at the foot of the vertical lines ; and the consecutive horizontal lines represent con- secutive degrees of temperature, commencing at some degree below the lowest temperature which is to be represented, and the degrees are marked in a vertical column to the left of the diagram. On the first ver- tical line put a dot at the proper point for showing the temperature at o h or i h , whichever is chosen to begin with, according to the scale of de- grees marked on the left of the diagram ; do this likewise on the next vertical line for the temperature at the next hour ; and so continue for all the 24 hours. Then join these points by a succession of short lines, and the resulting curved line will show the gradual rise and fall, or increase and decrease, of the temperature for the day. The diurnal changes in the temperature, taking the average for the year, on the island of Key West, Fla. (ocean coast), at St. Paul, Minn, (continental), at Fort Conger in the Arctic region, and also over the tropical oceans, are shown by the following hourly temperatures in degrees Fahrenheit: ih 2 h 3 h 4 h S h 6h 7 h 8fc 9 h IQh nh I 2 h Tropical Ocean, Key West, Fla. St. Paul, Minn. Fort Conger, 77.0 74.1 40.3 i.i 76.9 74.0 39-7 '3 76.8 73-8 39-1 1.2 7 6.7 73-6 38.5 i.i 76.7 73-5 37-9 1.0 76.8 73-4 37-6 0.7 77-3 74.0 38.1 0.4 77-9 75-3 38.9 0.2 78.5 76.4 40.6 O.I 79.0 77-3 42.5 0.6 79-3 78.1 44-5 1.0 79-5 78.3 46.2 1.2 I3 h I 4 h i5 h I6h I7 h i8h igh 20^ 2lh 22h 23 h 24. h Tropical Ocean, 79-5 79-3 79.2 79.0 78.5 78.2 78.0 77-8 77-6 77-5 77.4 77-3 Key West, Fla. 78.5 78.7 78.7 78.2 77.6 76.7 75-7 75-3 75-0 74.8 74.6 74-4 St. Paul, Minn. 47-4 484 48.8 48.9 48.5 47-6 46.3 45. 43-9 42.8 41.9 41.0 Fort Conger, 1.2 i-3 1.2 1.0 0.8 0.6 03 0.0 0.2 -0.5 0.8 0.8 The heavy-faced type shows the maximum, and the Italics the mini- mum, hourly temperature. The same data, except for Fort Conger, are shown by the graphical method in Fig. 6. 40 ELEMENTARY METEOROLOGY. The Average Daily Temperature. This is obtained ac- curately by taking the sum of the temperatures observed at .each of the 24 hours, and dividing it by 24. Gener- ally, however, observations are made but two or three times a day; and it is customary to choose such hours of the day for times of observation as will best allow the aver- age temperature to be computed from the observations made at them. Sometimes two separate hours are chosen, but more frequently three ; and, in addition, sometimes the readings of maximum and minimum recording thermome- ters are also available. If the sum of the temperature observations made at 7 A.M., 2 P.M., and 10 P.M., is divided by 3, very nearly the true average daily temperature will be obtained ; and special combinations of other hours will give nearly as accurate results. The average of the observations at 8 A.M. and 8 P.M., as observed by the United States Weather Bureau, gives, within a fraction of a degree, the average tempera- ture for the day. There must be two times during the daily march of the temperature when the average temperature of the day is reached, one during the time of increase of heat, and the other during the time of decrease, because this average lies between the maximum and minimum tem- peratures. The average temperature for the day occurring during the morning increase of heat is reached at about 8 h over a water surface, and at about 9 h over a land surface ; and during the evening decrease of heat, it is reached often as late as 4 hours after sunset in winter, and at about sunset in summer, over the land surface, while over the water surface it occurs at about an hour earlier than over the land. Temperature Changes during the Revolution of the Earth. - The inclination of the earth's axis causes a difference in the length of the days and nights, and a variation in the intensity of the solar rays at points on the surface, when the earth is in different parts of its orbit. TEMPERATURE. 41 The meteorological seasons (winter, spring, summer, and autumn), while each embraces three months in our temperate latitudes, are quite arbitrarily fixed when we consider the earth's surface as a whole ; for they neither occur during the same months in the two hemispheres, nor do they have the same duration in different latitudes. With us the coldest month (January) is taken as the mid- winter month, and the warmest month (July) as the mid- summer month ; and the mid-spring month (April) and the mid-autumn month (October) come midway between these. It is only in middle latitudes that the four seasons are all well marked and of approximately equal length. In lower latitudes the change from winter to summer is but slight, and the transition occurs so gradually as to be almost imperceptible so far as the temperature is concerned. In the very high latitudes, on the contrary, the change from summer to winter and from winter to summer is very abrupt, and there is really only a winter and a summer season. The seasons, except for local purposes, are to a great extent losing their former importance in meteorological re- ports, because of late years nearly all data are given for the separate months and for the calendar year, in order that the observations made all over the globe may be readily compared. For agricultural purposes the average meteor- ological conditions during the seasons are important. The times of the seasons are reversed in the northern and southern hemispheres. When it is winter in the northern hemisphere, it is summer in the southern hemi- sphere ; and when it is spring in the northern hemisphere, it is autumn in the southern hemisphere. In middle lati- tudes, where three months are given to each season, they are as follows : 42 ELEMENTARY METEOROLOGY. Months. Dec. Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. Nov. Seasons in the northern hemi- [ Winter. Spring. Summer. Fall, sphere. Seasons in the southern hemi- j- Summer. Fall. Winter. Spring, sphere. In Fig. 2 these ) periods B. C. D. A. are shown at J Gradual Change of Seasons. The change in the angle of inclination of the solar rays, and in the length of days, from day to day during the progress of seasons, is very slight ; and except near the poles the resulting change in the temperatures of places on the earth's surface is cor- respondingly small. We have, then, slow temperature changes during each season, and but a gradual passing from one season to the next. In following out the changes in temperature due to the revolution of the earth around the sun, we have, then, naturally to turn to the series of average temperatures for each day for successive days during the year. Accidental causes which give rise to variable temperature condi- tions from day to day, so disguise this slow gradual change, that we can realize its occurrence only after the lapse of a number of days. The Yearly Course or March of the Temperature. This is represented by the series of average temperatures for equal fractions of a year, usually for the months ; but in some cases five-day averages are used. Single daily aver- ages are not used, because even a century of observations would not give the average daily temperature with suffi- cient accuracy to remove all accidental irregularities which occur in them. There exists during the year (away from the equatorial TEMPERATURE. 43 region) but a single maximum and a single minimum, as was^seen to be the case for the daily temperature. In the northern hemisphere the warmest month is July, and the coldest is January, which is about a month later than the times when the sun reaches the highest and lowest altitudes respectively. On and near the oceans the greatest retardation takes place. In the southern hemisphere the months of maximum and minimum temperatures just -stated are reversed. It thus occurs that in the region about the equator, the maximum temperatures occur at the time when higher latitudes are enjoying medium temperatures (spring and fall). Two maxima and two minima thus exist for the equatorial region, but the amplitudes are compara- tively slight. The reason that the highest and lowest monthly temperatures occur later than the times of highest and lowest meridian altitude of the sun, is similar to that for the retardation in these phases in the diurnal changes of temperature. When the whole amount of heat received from the sun during the 24 hours of the day just equals the amount lost from the earth by radiation, then the average daily temperature remains the same. But as the sun gets higher in the heavens in the spring, the amount of heat received from it exceeds more and more the amount lost by radiation, and the days grow warmer. When the sun recedes from the tropics in summer, the amount of heat received still exceeds for another month or so the amount lost : consequently the average temperature still increases during this time, and continues to until the sun has moved so far equatorward that the amounts of heat received and lost are equal. The average temperature also con- tinues to decrease during some weeks after the sun has reached its lowest meridian altitude, for the amount of heat lost by radiation still exceeds the amount received from the sun until the sun's altitude has con- siderably increased. In our latitude the two become equal in January. The amplitude of the yearly temperature period is usually understood to be the difference between the 44 ELEMENTARY METEOROLOGY. highest (maximum) and the lowest (minimum) monthly averages. This amplitude, for the same surroundings, increases from the equator poleward. The land and water distribution very powerfully influences the yearly ampli- tudes, which are least over the oceans ; so that a very dry continental location near the equator may have as great an amplitude as a moist oceanic location far to the poleward. For instance, the yearly amplitude at Biskra in the Sahara (conti- nental) is 39.8 F., and at Madeira (oceanic) it is but 10.8 F. At Vardo (oceanic), northwestern Europe, the amplitude is 25.9 F., and at Werchojansk (continental), Siberia, it is 1 17.9 F. At this last place there has been observed a temperature of 90 F. in winter, and of + 86 F. in summer, thus making an absolute extreme amplitude of about 1 80 F. in the year. The temperature amplitudes for the year in our middle latitudes increase along the same parallel from the western coast towards the middle and eastern parts of the conti- nents, and then decrease again towards the eastern coasts. On the eastern coast the amplitudes are greater than on the western coast, on account of the great influences of the prevailing winds. With increase of altitude, the yearly amplitude decreases, and more nearly approaches that for a marine exposure at the same latitude. Below are given, in degrees Fahrenheit, average monthly temperatures for Key West, Fla., St. Paul, Minn., and Fort Conger in the Arctic region. Jan. Feb. Mar. April May June July Aug. Sept. Oct. Nov. Dec. Key West (Seashore) 69.7 71.4 73- 2 76.6 79-8 83.0 84.1 84.3 82.2 78.8 74-3 70.0 St. Paul (Continental) 77.7 17-5 28.2 44-9 58.8 67.2 71.9 6y.2 58-7 47.0 31.0 18.1 Fort Conger . . . -38.2 -40.1 -28.1 -13-6 14.1 32.6 37-i 33-8 15-8 -8.9 -23.6 -28.1 The same, except for Fort Conger, are shown graphically in Fig. 7. TEMPERATURE. 45 Excessive Range of Monthly Temperatures. Excessive amplitudes of temperature are due to the extremely low minimum temperatures, and not to unusually high maxi- mum temperatures ; the minima going lower than usual, and not the maxima going higher. This is realized by considering the fact that we should call a day warm when the thermom- eter reaches 85 F., but very hot if it reaches 100 F., the difference being but 15 F. While 30 F. would be considered quite cold, yet a fall of the temperature to 20 F. below zero is not an unusual experience in the northern part of the United States ; and this would mean a lowering of the temperature by 50, which is over three times as great as the increase of 15 F. just mentioned. The highest average monthly temperature ever observed is that of 102 F. for July, at Death Valley, Cal. ; and the lowest is -66 F. for January, at Werchojansk, Siberia. Proba- bly the temperatures in some parts of the Desert of Sahara are greater than those for California, but we have little accurate knowl- edge of the temperatures of that African re- gion. The average temperature for the year is obtained by taking the sum of the average temperatures for all the months, and dividing it by the number of months. Since monthly temperatures vary, so the yearly average temper- ature will vary, but not so much as the monthly temper- atures ; and these last, in turn, are less variable than the daily temperatures. Irregular Changes of Temperature occur, which depend on the degree of cloudiness, effects of radiation, and on the V JFMAMJJAX X D FIG. 7. ANNUAL CHANGE OF AIR TEMPERATURES. 4 6 ELEMENTARY METEOROLOGY. movement of air masses. These temperature changes are local, but may be of wide extent. When irregular rapid fluctuations of the temperature take place, the distribution of the air temperature near the earth's surface is usually as follows : There is a central, more or less limited, area of high or low temperature, from which the temperature irregularly but gradually decreases or increases, as the case may be, in all directions. 500 Miles FIG. 8. AREAS OF Low AND HIGH TEMPERATURE. 1000 Miles. These areas are shown in a general way in the accompanying dia- grams (Fig. 8), in which the curved lines inclose limited areas on the earth's surface. In the figure on the left, the air in the central area has a temperature of 20 F. or less ; around this is a zone in which the temperature is from 20 to 10 F., and around this are others in which the temperatures are from 10 to o and from o to -f 10 F. respectively. In the figure on the right, the central area has a tempera- ture of 80 F. or over, and within the successive surrounding rings the temperature is from -f8o to +70 F., from +70 to +60 F., etc. Many times these areas do not appear closed on all sides, and this regularity of increase or decrease of temperature does not take place on the open side within the limits of the region considered. These TEMPERATURE. 47 areas of abnormally cold or warm air move over the earth's surface in regular tracks, and this causes the temperature irregularities for particular places to have quite constant average values, depending on the location of these places with reference to the tracks. Sometimes these areas extend over many thousands of square miles, the inner region of greatest or least temperatures sometimes covering more than one of our States. The Temperature Anomaly. If the average tempera- ture for the same month during many successive years is obtained, and the difference between this average and the monthly temperature for each year is taken, the mean of these differences will be the mean temperatitre anomaly for that month. Temperature anomalies are the least in lower latitudes, and are greater at the interior of continents than on the oceans. They are the greatest in winter, and least in midsummer. The mean anomalies are relatively quite constant for adjacent places. Variability of Temperature. The variability of temper- ature from day to day is the difference between the aver- age temperatures for successive days. The average of such temperature differences is usually computed for each month. Temperature variability is least in the tropics, and increases with the latitude up to about latitude 50, whence it decreases again toward the poles. It increases from the sea towards the interior of continents, and also with the alti- tude on the average for the year ; and up to a certain limit it decreases in summer and increases in winter with the altitude. It is greater on the east than on the west coast in our latitudes, and is greatest in winter and least in summer. Concerning the extreme variability of mean daily tem- perature from day to day, it may be said that changes of 4 F. occur in all parts of the world, of 7 F. in most parts, and of 10 F. in but few parts. In the interior of 48 ELEMENTARY METEOROLOGY. continents, as, for instance, North America and Asia, individual changes of even 45 F. occur, but at rare intervals. The amount and frequency of the individual temperature changes increase towards the interior of a continent, but in our middle latitudes are greater on the eastern coast of the continent than on the western. Average Temperatures for Adjacent Regions. The difference in the temperature for adjacent places is a much more constant quantity than are the absolute temperatures themselves. For instance, suppose that the temperature has been observed at two neighboring places, at the one for the last 100 years, and at the other for only the last 10 years, then the best way of obtaining the true average temperature of the second place is to find the average temperature at both places during the same 10 years, and take the difference between the two ; and this difference applied to the average for the one place for 100 years will give the average temperature for the second place much more accurately than it has been determined from the 10 years 1 observations alone. Long-period Temperature Oscillations are shown by direct observation of thermometers, and by several indirect meth- ods, such as the dates of the harvests, the opening and closing of navigation by ice, the relative severity of winters, and other methods. Observations (mostly European) show a periodic fluctuation of temperature during a period of about 35 years. During the past century the years 1791 to 1805 were relatively warm; 1806 to 1820, cold; 1821 to 1835, warm; 1836 to 1850, cold; 1851 to 1870, warm; and 1871 to 1885, cold. The average fluctuation amounts to about 2 F. during these periods. Change of Temperature with the Altitude above the Sur- face of the Earth. The normal condition of the tempera- ture of the air is a decrease with the altitude above the earth's surface. This decrease is not the same in amount TEMPERATURE. 49 at all places on the earth, nor is it constant at any one place. In fact, it frequently occurs that for a short time, and for a short distance upward, there is an increase of temperature with the altitude. The observed decrease of temperature with the altitude is, on the average, about i F. for 330 feetjj but it is more rapid in summer than in winter, and it is more rapid in the lower air layers than in the higher. In the winter and in the nighttime, in clear weather, the air in the valleys is cooler than that on the low hilltops ; and in summer and in the daytime, the air in the low valleys is warmer. Observations have been made on mountains and in balloons with the view of determining accurately the law of the decrease in tempera- ture with the altitude ; but such a great variety of results have been obtained, that the matter is at present by no means definitely settled. The laws of thermodynamics and adiabatic 1 cooling show that for dry air there will be a decrease of temperature of about i F. for 183 feet of increase in altitude, and a corresponding increase of temperature for a like decrease in altitude. The usually observed decrease of tem- perature with the altitude, as has been stated, falls short of this amount. The outer limit of the decrease of temperature in the atmosphere, or the 1 By adiabatic heating or cooling of a mass of air is meant a change in its temperature brought about by a change in its density, and without the addition of the heat from or the loss to outside sources. Dry air will cool adiabatically about i F. if its density is decreased by a change in pressure equal to that involved in a change in altitude of 183 feet. Hence, as a result of the de- crease in the pressure and density of dry air with altitude, there will be a corresponding decrease in the temperature of i F. for each 183 feet of altitude. When the pressure upon a body of air is decreased, the air expands. To effect this expansion, part of the heat energy of the air is utilized, and becomes insensible as heat; hence the air becomes cooler, or its temperature falls, as a result of its expansion. On the other hand, when the pressure on a body of air increases, the air is compressed into less space; and part of the energy which was employed in maintaining its former bulk, being relieved of this duty, appears as sensible heat; wherefore the air becomes warmer, or its temperature rises, as it grows denser under compression. WALDO METEOR. 4 5O ELEMENTARY METEOROLOGY. temperature at points above the earth's surface where the air is very rare, is not known with any degree of certainty ; but it has been thought to be about 45.5 F., judging from the observed decrease of temperature up as high as ten or twelve thousand feet. A temperature of 54 F. has, however, been observed in a balloon at an altitude of 31,500 feet. This temperature would be the same above the pole and above the equator. It is seen, then, that in Siberia, with an average tempera- ture for January of 58 F., there must be an increase in the temperature with the altitude for a relatively long time ; but it is probable that this increase in such cases extends upwards for a short distance, and that it reaches a limit at a not great altitude, from which point there is a decrease farther upwards ; that is, there is a relatively warmer mass of air somewhere between the very cold air layer at the earth's surface and the very cold outer limit of the atmosphere. The rate of decrease varies for different months of the year, and the amount of this variation is much greater for inland mountain regions than for those near the ocean. The maximum rate of decrease is about i F. for 200 feet in most cases ; and the great variation which occurs for individual cases is due to inequalities in the minimum rates, which appear to increase with the latitude. The average rate of decrease for the year, while varying from i F, per 370 feet to i F. per 220 feet, is probably about i F. in 330 feet altitude. The rate of temperature decrease with the altitude also varies from hour to hour during the day, and moreover is greatly influenced by the degree of cloudiness. Near the ground at midday in clear weather the decrease of tem- perature with altitude is much more rapid than in cloudy weather ; but at evening the decrease in cloudy weather is greatest. At an altitude of several hundred feet above the ground the midday decrease in clear weather is not much greater than for cloudy weather, and may even become equal to it ; but in the evening the decrease is less during clear weather than for cloudy weather. The decrease of temperature in cloudy weather is, then, quite constant for both day and night above the very lowest air layers ; but in clear weather it varies greatly. The cases of inversions of temperatures, or increase of tem- perature with altitude, occur in clear weather. Geographical Distribution of Air Temperatures over the Earth's Surface. If the earth had a homogeneous surface TEMPERATURE. 5 1 of either land or water, the average temperature for the year would decrease from a maximum at the equator to a minimum at the poles, and the average temperature along any parallel could be considered as dependent on the lati- tude alone. If the surface were all land, it would have a higher temperature than if it were all water, in the equatorial regions, because land is more easily warmed than water; but in the polar regions the water surface would have the higher temperature, because the land loses its heat by radiation faster than the water. At some middle latitude the temperatures for the two surfaces would be equal. If we compare the temperatures for an ideal land-cov- ered with that of a water-covered earth, then at the equa- tor the temperature of the land would be 115 F., and of the water nearly 72 F. ; and the difference between the equatorial and polar temperatures would be, for the land about 135 F., and for the water about 54 F. In the actual case of the unequal and irregular distribu- tion of land and water on the surface of the earth, however, the average temperature along any parallel is dependent on this distribution as well as on the latitude. There are two other causes which are also effective, but to a much less degree. These are 'the transference of heat toward the poles, and cold toward the equator, by means of ocean currents; and the transference of heated or cooled air by the prevailing winds. The effects of these, however, are practically neutralized over the entire circumference of the earth, because where one part is rendered cooler by a cold current, another is made warmer by a warm current. Isothermal Charts. An isothermal line, or isotherm, is a line every point of which has the same temperature. An isothermal surface is a surface every point of which TEMPERATURE. 53 has the same temperature. When the average tempera- tures for the year at different places are written down on a map at those places, and the isothermal lines are drawn on the map, then we have a chart of the isothermal lines for the year for the region covered. These isothermal lines show where the isothermal surfaces of the air, which are inclined downwards from the equatorial towards the polar regions, intersect the level of the surface of the earth. Charts of this kind have been made for the whole earth, except for the polar regions, for which we have few observations. Similar charts have been constructed, show- ing the average temperatures for each month of the year. Such charts for the year (Fig. 9), for the midwinter month, January (Fig. 10), and for the midsummer month, July ( Fig. 1 1 ), are given here. Courses of the Isotherms. As was to have been ex- pected from what has just been said, the isothermal lines for the year (Fig. 9) do not follow the parallels of latitude around the globe. These irregularities are due to three main causes : , 1. The unequal distribution of the land and water sur- faces, and the excessive heating of the land in summer and cooling in winter, as compared with a water surface. The irregularities are much more marked in the northern than in the southern hemisphere, because in the former the ratio of land to water is the greater. The excessive cooling of the interior of the con- tinents in winter causes the isothermal lines to make there a bend equatorward at that season, while in summer the excessive heating of the land causes the isotherms to make a poleward bend. 2. The transference of the equatorial heat poleward, and the polar cold equatorward, by means of the oceanic currents ; and this heat and cold are communicated to the air. 54 ELEMENTARY METEOROLOGY. The great oceanic currents give a greater warmth to the waters in the western parts than to those in the eastern parts of the oceans up to about latitude 40, where the warm current coming from the equator, along the eastern coast of the continent, crosses the ocean ; and, divid- ing as it approaches the western continental coast, part of it flows pole- ward, taking abnormally warm water with it, while part of it flows equatorward, taking with it colder water than was to be found at the same latitude on the western side of the ocean. On the eastern conti- nental coast to the poleward of latitude 40, where the warm current from the equator crosses the ocean, there is a cold current coming from the polar regions, which makes the coast waters abnormally cold. 3. The transference of heated and cooled air by the prevailing winds. In the middle latitudes of the two hemispheres the winds from the west blow in winter the colder air from the continents on to the west- ern parts of the ocean ; and the isothermal lines crossing the oceans from west to east are inclined poleward (that is, in a northeasterly direction in the northern hemisphere), and separate farther and farther apart in their progress. The isothermal lines crossing the continents are inclined toward the equator (that is, in a southeasterly direction in the northern hemisphere), and converge more and more in their progress, because the prevailing winds are from the west, and in winter blow the warmer oceanic air on to the western coasts of the continents, and the cooler continental air on to the eastern coast regions. In the summer the winds from the west blow the cooler oceanic air on to the western coasts of the continents, while the same winds blow the heated continental air on to the eastern coasts. Thus the western coasts are kept cooler, and the eastern become warmer. In the equatorial zone the winds blow mostly from the east, and the conditions depending on the general direction of the wind are in gen- eral the reverse of those in middle latitudes. A more detailed statement of these reasons is deferred to the chapter on climates. Normal Average Temperatures. The average temper- atures for the different degrees of latitude have been derived (55) ELEMENTARY METEOROLOGY. by taking the average of the temperatures at various points along individual parallels. The following table shows these average temperatures for the year, for Jan- uary, and for July, along each 10 of latitude, both north and south of the equator : LATITUDE. JANUARY. JULY. YEAR. North 80 F. 3O Q F. -4- 1.2 1 F. 4- r 6 : North 70 North 60 -15.7 + 7 Q r/ + 44 .I t;6o 13-7 29 8 North 50 . . . 10 Q 64.6 4.2 C North 40 42 Q 7S 4 **o C7 1 North 30 North 20 59-5 71.7 />* 8 1.0 82.4." J/' 1 68.4 76 Q' North 10 ... 78. S 807 80 8 : Equator o 800 78 2 7Q Q South 10 800 74 Q 782 South 20 77.6 66.0 73 Q South 30 South 40 70.0 i 57.1 460 64.0 S40 South 50 South 60 47-4 + 34-9 +36.8 0^- w + 41-5 Up to about 45 latitude the northern hemisphere is warmer than the southern, but beyond that latitude the southern hemisphere is the warmer. The warmest par- allel for the year, or the thermal equator, is found to be about 10 north of the equator. The rate of poleward decrease in the temperature is irregular, since the relative amounts of land and water vary. The differences between the midwinter and midsummer temperatures for the two hemispheres are found to be very unequal ; but the average for the January and July temperatures gives nearly the annual temperature in the respective hemispheres. (5?; 58 ELEMENTARY METEOROLOGY. The average surface air temperature of both the northern and southern hemispheres is about 59 F. For the extreme months January and July the average temperatures have been computed as follows : TANTAKY. JULY. F. 46.4 F. 72 C Southern hemisphere 63 * HA 1 Whole earth o J^O 6l 1 jj'^ U O'J It has also been found, that, on account of the greater amount of land in the eastern hemisphere (counting this from 80 west longitude to 100 east longitude from Green- wich), the eastern hemisphere is about 2 F. warmer than the western. Abnormal Average Temperatures. The irregular distri- bution of temperatures over the earth is best shown by considering the difference between the average tempera- ture of each parallel and the actual temperature at places along the same parallel. These differences are called the abnormal temperatures of those places. The abnormal temperatures are charted, and the lines of equal magni- tudes drawn on the charts are called is-abnormals. For the whole year and for the winter the temperatures are abnormally highest over the eastern parts of the oceans, and abnormally lowest over the centers of the continents in northern latitudes. For the summer the temperatures are abnormally high- est over the centers of the continents, and abnormally lowest over the eastern parts of the oceans in northern latitudes. Such charts are shown, for January in Fig. 12, and for July in Fig. 13. (59) 60 ELEMENTARY METEOROLOGY. Abnormal Temperatures for the Year. The regions where the local temperatures depart most below the normal are in north-central North America (-I2F.) and northeastern Asia (-i8F.). The regions where the local temperatures depart most above the normal are off the coast of Norway ( + 20 F.), around the eastern end of the Mediter- ranean Sea ( + 10 F.), and on the southern Alaskan coast ( + 8 3 F.). Abnormal Temperatures in January. In the chart (Fig. 12), the minimum negative abnormal temperatures are 30 F. in northeastern Asia, between 20 and 30^ F. in north-central North America, and 10 F. to the west of the middle latitudes of South America and the southern latitudes of Africa. The maximum positive abnormal tem- peratures are +40 F. off the western coast of Norway, +20 F. off the southern coast of Alaska, and + 10 F. in central Australia, southern South America, and southern Africa. Abnormal Temperatures in July. In the chart (Fig. 13), the mini- mum negative abnormal temperatures are 10 F. on the northeastern part of the Pacific Ocean, on the North Atlantic between southern Greenland and the North American Continent, and to the west of South America and Africa a few degrees south of the equator. The maximum positive abnormal temperatures are -f 20 F. in the north- western United States (inland), + 10 F. in central Asia, and + 10 F. over the desert of northern Africa. In the southern hemisphere there is a slight negative abnormal temperature in central Australia. Over the greater parts of South America and Africa there is a slight positive abnormal temperature both in winter and summer, while over the ocean to the west of these the air is abnormally cool. Annual Range of Average Monthly Temperatures. The difference in the average temperatures for the coldest and warmest month is called the range in the monthly tem- peratures for the year. It increases with the latitude and towards the interior of continents, but decreases with the altitude. These temperature differences for many localities have been deduced and entered on a chart, and the lines of equal range of temperature have been drawn by connect- ing adjacent places having the equal differences of tem- TEMPERATURE. 63 perature. Such a chart is given in Fig. 14; and it shows that along the equator the annual range of temperature is 5 F. ; and there is an increase with the distance from the equator, slowly over the oceans, but rapidly over the interiors of the continents. In the southern hemisphere the temperature range reaches 30 F. in Africa, South America, and Australia. In the northern hemisphere the temperature range increases, at first slowly, and then at the extreme north rapidly, to about 40 F. in the Pacific and Atlantic oceans ; but the increase, is to 80 F. in north-central North America, and to 120 F, in northeastern Asia. The 5 F. line, extending nearly across the South Pacific Ocean, shows a remarkable instance of uniformity in the temperature range over a water surface ; and the 5 F. line in Africa, and the 40 F. and the 70 F. line in southern and central Asia, show for the interior of continents a like parallelism with the circles of latitude. Charts of Average Highest and Lowest Temperatures for the Year. The accompanying charts (Figs. 15 and 16) show the average maximum and average minimum air temperatures for the year over the whole earth. The lines of equal maximum and minimum temperatures are drawn for each 9 F. It must be understood that the charts represent the averages, for a number of years, of the extreme highest and lowest temperatures occurring in each year. The average highest temperature for the year decreases with increase of latitude and altitude, but increases towards the interior of continents. The average lowest temperature for the year becomes lower with increase of latitude and towards the interior of continents. % Chart of Average Extreme Maximum Temperatures for the Year (Fig. 15). This chart shows that the maximum temperatures have a quite regular distribution on the oceans, where, for wide zones extending both sides of the equator, there is a maximum of about 86 F. ; and in no TEMPERATURE. 65 case does it reach 95 F. From these zones poleward there is a rela- tively rapid decrease with the latitude ; and the temperature of 68 F. is reached near 60 latitude in the northern, and 50 latitude in the southern hemisphere. The continental distribution is quite different, however, for there an increase of the maximum temperatures takes place with the progress inland. In the interior of northern Africa, Persia, northern India, Aus- tralia, and southern North America, temperatures of 113 F. are to be found ; and in the southwestern United States, and perhaps in the Sahara desert, they even reach 122 F. At high altitudes the extreme ranges of the maxima are less than on the lowlands, and are thus more like those for marine localities. Chart of Average Extreme Minimum Temperatures for the Year. This chart (Fig. 16) shows very strongly marked regional character- istics. On the Pacific, Atlantic, and Indian oceans, in the equatorial region, there are extensive zones stretching from west to east, in which the minimum temperatures do not go below 68' F. To the north of these there is a rapid decrease in the minimum temperatures, but to the south it is not so rapid. There is rapid decrease towards the interior of the continents, and especially where there exist mountain ranges to cut off the access of the sea air to the interior. In the northern hemisphere there are three centers of extreme minimum temperatures, one in the eastern part of Siberia, another in the northern part of North America, and a third in the interior of Greenland. The 32 F. line is very interesting as showing the limits of the region in which the freezing point of water is reached. In the northern hemi- sphere, commencing at the Yellow Sea, we can follow its easterly course with a northern bend on the Pacific Ocean, cutting North America at about latitude 30. On the Atlantic Ocean, following the Gulf Stream for a distance in a northeasterly direction, it reaches almost to Ireland, when it takes a southeasterly trend, and passes along the coast of the Spanish peninsula. Continuing through the Mediterranean Sea, it passes through southern Asia to the place from which we started in tracing it. In the southern hemisphere it makes almost a loop within the continent of Australia, and passes from its southeastern corner in an easterly direction through northern New Zealand, and thence onward until the southern end of South America is almost reached ; but, stop- ping short of it, the line suddenly makes a northward curve up along TEMPERATURE. 67 the coast to about latitude 20 south, where it crosses the continent, and then runs southward again along the eastern coast to latitude 45 south, and thence takes an easterly course to the southern end of Australia. Variations in altitude above sea level have even a less influence on the minimum temperatures than on the maxi- mum temperatures, as frequently in the severest cold weather there is an inversion of the temperatures with the altitude. Average Annual Oscillation of Temperature. The aver- age extreme oscillation or amplitude of the temperature during the year, increases with the latitude and towards the interior of continents, but decreases with the altitude. The lines of equal amounts of the extreme temperature oscillation during the year are drawn on the chart (Fig. 17) at intervals of 9 F. , These curves bring out most clearly the difference between the con- tinental and marine climates. On the ocean near the equator the absolute temperature amplitude becomes less than 18 F. It increases, however, towards the poles and towards the continents, and especially towards the interior of the continents. In eastern Asia an average amplitude of 171 F. (in single cases over 180 F.), and in North America 153 F., is reached ; but at the interior of Australia the ampli- tude is only 90 F., and in South America only 81 F. A further description of these charts cannot be given here, but they will well repay a careful study. In many respects they are of more interest than those of the average monthly or annual temperatures. The average time of the lowest temperature for the year occurs ear- lier for a continental climate than for a marine climate. The warmest and coldest days of the year do not occur especially frequently on the average dates of extreme temperature, or on any other particular dates- The Snow Line as Dependent on Temperature. The snow line, or line of eternal snow, is that line along which, in the course of a year, just as much snow falls as can be TEMPERATURE. 69 melted by the sun in the same time. Below this the snow would disappear ; and above it, it would remain perma- nently present. This snow line is not fixed, but varies in altitude in different regions, and from year to year in the same region. The region of the snow line, then, is an alti- tudinal zone within which this line oscillates up and down with the varying temperature of the seasons. The limits of this zone, and even the average height of this line, may be determined not only by means of many years' observations of the snow phenomenon direct, but also by means of the study of per- manent glaciers. These last must extend up into the regions of eternal snow, and the upper limit of the snow line is lower than the peak sur- mounting the glacier. By an examination of the surrounding lower peaks, which are favorably situated for glacier formations, but do not have them, a lower limit for the glacier can be established ; for the snow limit must be above these glacierless peaks, and below the peaks which have glaciers. The temperature at the limit at which the snow disap- pears depends a good deal on the amount of snow which has accumulated. The mean annual temperature at the lower limit of the snow line varies from 37 to 3 F. for different regions for which we have observations ; and it may be said, that, the greater the differences between the winter and summer temperatures, the less will be the average annual temperature at which snow is always found. Observations covering a number of years are necessary for determining the average position of the snow limit, which is more constant in lower than in higher latitudes. The average annual temperature at which the snow line is found decreases towards the pole, as is seen from the table given on the next page. The snow line is found nearer the sea level as the cold polar regions are approached. ELEMENTARY METEOROLOGY. There has been a great deal of data, of more or less accuracy, accu- mulated concerning the snow limit in various parts of the world, but the following table will suffice for reproduction here : MOUNTAINS. GEOGRAPHICAL LATITUDE. ALTITUDE OF SNOW LIMIT ABOVE SEA LEVEL. AVERAGE ANNUAL TEM- PERATURE. Feet. F. about Andes (near Quito) . . Equator. i55o 34 East African Mts. . . . Near Equator. I5.5 00 Himalaya (south side) . 27-34 N. l6,OOO 3i Himalaya (north side) . 27-34 N. 18,500 27 Middle and West Alps . 46 N. 9,000 27 Tyrolean Alps (central) . 47 N. 9,000 + 25 Nova Zembla .... 73 N. 2,000 12 Spitzbergen .... 77 N. I, COO 14 / / *o T" Temperature of the Ground. The heat of the ground at or near the earth's surface is received from two sources, the interior of the earth, and the sun. The heat of the earth increases towards the interior, and there is a flow of heat by conduction from the interior outward to supply the heat lost by radiation from the surface of the earth. The heat from the sun raises the temperature of the earth's surface higher than that due to the heat received only from within. According to the law for the conduc- tion of heat, there must, then, be a transference of heat from the earth's surface towards the interior of the earth, and there is a decrease in the temperature of the suc- cessive layers of earth as far as the heat from the sur- face penetrates ; and from this point inwards towards the center of the earth there is an increase of temperature. The diurnal changes of ground temperature^ following those of the air above, are greatest at the equator, and TEMPERATURE. 71 decrease towards the pole, and are noticeable to a depth of about 40 inches at the equator. With descent below the surface, the times of maximum and minimum temper- ature are retarded. The annual changes of ground temperature follow after those of the air above to a depth of about 80 feet; but the times of maximum and minimum temperature are retarded as much as several months at such depths. The average ground temperature at a depth of about 3 feet below the surface is about 2 F. higher than the air temperature above. The normal increase in the earth temperature at greater depths is about i F. for each 60 feet descent ; but very different rates of increase have been observed in various parts of the world. The Region of Frozen Earth, where, at some depth below the earth's surface, there is a continuous temperature of 32 F. or a lower temperature, occurs in the polar regions and some very elevated mountains on other parts of the earth. The southern limit in the northern hemisphere at which the ground is perpetually frozen, is along the annual isotherm of about 28.5 F. for air temperatures at sea level. Temperature of the Ocean. Observations of the sur- face temperatures of the ocean water exhibit a daily change somewhat similar to that of the air above it; but the amplitude or range amounts to only a fraction of a degree during the 24 hours. The maximum temperature occurs shortly after midday, and the minimum just after sunrise. The annual change of the ocean surface temperatures follows that of the air; but the amplitude is much less, and increases with the latitude up to high middle latitudes, where it decreases again. The times of maximum and 72 ELEMENTARY METEOROLOGY. minimum temperatures are retarded about a month later than those for the air temperatures, and occur in August and February respectively, in the northern hemisphere. In the warmest parts of the tropical zone the surface temperature changes very little during the year, and is about 82-84 F. In middle latitudes it varies from about 50 F. in winter to 68 F. in summer. In the polar zone the temperature approaches or goes below the freezing point of fresh water, and varies but little with the season. Temperatures of Small Isolated Bodies of Water. The distribution of temperature in the lakes is regulated by the currents which arise from the alternate heating and cool- ing of the surface water. The warm surface water in summer imparts little of its heat to the water below, which consequently does not get much above 39 F. for deep lakes. There is, however, a condition of stable equilib- rium, because the temperature decreases with the depth, and the coldest and heaviest water is at the bottom. In the fall and early winter the surface cools to about 39 F., at which temperature water is heaviest ; and then, when the whole body of water reaches this temperature, the water is in indifferent equilibrium, and there are no ver- tical currents. When the surface water becomes still colder, it becomes lighter than the water below, and does not sink. CHAPTER III. AIR PRESSURE. Nature of Air Pressure. The atmospheric air, obeying the law of gases, exerts a pressure in all directions, and the amount of this pressure varies according to the den- sity of the air. The air at the sea level, weighted down by the air above it, exerts a pressure of nearly 1 5 pounds per square inch of surface against which it presses. The air pressure decreases with increase of altitude, because there is less air above, the higher the ascent. We do not know just how far from the earth's surface the air pressure is a measurable quantity, but at an altitude of a few miles it becomes very small. Where the free air has access below the earth's surface, there is a continued increase of air pressure with the descent. Also at various points on the earth's surface the air pressures are not quite the same ; and for any one point the pressure undergoes variations of the nature of oscillations occurring in shorter or longer periods of time. Distribution of Air Pressure. The pressure of the air called the barometric pressure, from the instrument used for measuring it would be symmetrical, and would be practically the same at all places having a common alti- tude above the level of the sea, if it were not for the dis- turbing influence of the solar heat and of the movements 73 74 ELEMENTARY METEOROLOGY. of the air caused by its unequal heating at different places on the earth. If, under the condition of constant temperature, it is necessary to ascend a certain distance above the earth's surface in order to leave sa y T X O f tne atmosphere below and have the other T 9 9 24 8,200 22 10,000 20 13,200 18 16,000 . 16 The Barometer. Galileo discovered, over two hundred and fifty years ago, that the air exerts its elastic pressure in all directions. Shortly afterwards (in 1643) Torricelli gave us an easy method of measuring this air pressure, or barometric pressure as we shall usually call it, when he invented the barometer. The simplest form of barometer is obtained by filling WALDO METEOR. 5 ELEMENTARY METEOROLOGY. with mercury a glass tube of a length of, say, three feet- and one third of an inch in diameter, and closed at one end. Hold the closed end uppermost and insert the mouth of the tube in a cup also containing mercury (Fig. 19). On making the tube vertical, some of the mercury will flow downward out of the tube into the cup, until the weight of the mercury in the tube is counterpoised by the hydro- static pressure of the air on the surface of mercury in the open cup. The height of the surface of the mercury in the tube above the sur- face of the mercury in the cup is meas- ured by means of a scale of length ; and the amount, expressed in inches, is called the barometer keig/tt, or the barometric pressure , or merely the barom- eter reading. In making accurate determinations of the baro- metric pressure, several sources of error must be allowed for. If proper care is taken in the manip- ulation of the tube, the space above the mercury in the barometer tube will be practically a vacuum, and consequently there will be no pressure on the top surface of the mercury in the tube. Fig. 19 shows on the left a barometer tube situated as just described. The figure on the right shows the position the mercury would take if the tube FIG. 19. SIMPLE FORMS ; OF BAROMETERS. had a U bend at the bottom, in which case the mercury cup would be unnecessary. The distance AB is about 30 inches under ordinary air pressures at sea level. Fig. 20 shows an actual barometer where the glass tube is incased in a metal tube which serves not only to protect it, but also for a measuring sca!e. The cup or cistern of mercury is at A. By means of the screw C the mercury is forced up until the mercury surface in A is brought AIR PRESSURE. 77 FIG 20 - COMPLETE fixed altitude. MERCURIAL BAROM- to the little pointer shown at B, which is the zero or beginning of the barometer scale. The height of the mercury in the tube is then read off from the scale at C. D is an attached thermometer. Fig. 21 shows an aneroid barometer. This is an instrument shaped something like a small clock, and the barometric pressure is read by means of a hand or indicator, on a scale of inches placed on the face like the minute divisions of a clock. A little flat, circular, metal box within the case is partly exhausted of air, and tightly sealed ; and when the outer air pressure increases, it forces in the side of the box, and this mo- tion is communicated to the hand, making it move around toward the right. When the pressure is lessened, the hand moves to the left. Another hand, FIG. 21. -ANEROID BAROMETER. which is usually placed on the case, remains station- ary unless moved by some person. The Reduction of Barometric Pressure to Sea Level. This is one of the most un- satisfactory problems connected with prac- tical meteorology. The observations of air pressure are usually made at places having various (known) altitudes ; and, since the air pressure decreases with alti- tude, then, in order to compare the results from these places, the pressures must be reduced to some common plane having a For convenience the sea 78 ELEMENTARY METEOROLOGY. level has been chosen as a level for reduction, but the observations could be reduced to any other desired level, and sometimes other levels are chosen. Closely connected with this matter is the determination of the difference in the altitude of two places by means of the observations of the air pressure at both places. We can suppose observations to be made on various parts of the mountains shown in Fig. 18. If the barometric pres- sure is observed in a balloon at B, and we wish to reduce this to the corresponding pressure at sea level, we should add it to the weight of the air column BC (expressed in barometric pressure); and similarly for reducing baro- metric pressure observed at A to sea level at E, we must add to it the weight of an assumed column of air, AE. The problem of the reduction of the barometric pressure to sea level is too complicated to explain in full here, but it amounts to this, that we have the air pressure given for a certain altitude, and it is required to find the weight of a column of air which would extend from this alti- tude down to sea level. In the case of an observation in a balloon over the ocean, there would be a real column of air to sea level ; but in the case of an observation on a mid-continental mountain, the column of air would be purely fictitious. In order to determine the weight of the intervening column of air, it is necessary to know its average temperature. This might be determined in the case of the balloon ; but for the mountain the local temperature of the region must be taken as the upper temperature, and from this the temperature at the assumed sea level below must be calculated, in order to obtain the average tem- perature of the fictitious air column, so that its weight can be com- puted. The laws of the decrease of air temperature with altitude are so different for different regions, that there is room for considerable error in obtaining the average temperature of the air column by this means ; and in proportion as this is in error, just so much is the com- puted weight of the air column in error. When the weight of the air column is determined on the barometric scale, this amount is added to the corrected observed barometric pressure, and the sum is the barometric pressure reduced to sea level. Tables for facilitating the AIR PRESSURE. 79 reduction of barometric observations to sea level are numerous. Those in the " Smithsonian Meteorological Tables " and Hazen's " Meteoro- logical Tables " are very complete. The Determination of the Difference in Altitude of two neighboring places by means of simultaneous barometer observations at the two points, is the reverse of this reduc- tion from one known level to another known level. In this case the barometric pressure and air temperature are given for the two points ; and it is required to determine the vertical length of a column of air which has the pres- sure and temperature of the higher place at its top, and those of the lower place at its base. If we wish to deter- mine the difference in altitude between A and D (Fig. 18), we should observe simultaneously the barometric pressure at D and -at A (which would be the same as at F\ and the difference in the pressures will be the weight (in pres- sure) of the column of air FD. We can determine the vertical length of the air column for each inch of difference in these barometric pressures, and, by combining their whole and fractional parts, obtain the total distance correspond- ing to the whole difference of pressures. Tables are also used to facilitate these computations ; and since the difference in the altitudes of the two places seldom extends from high elevations to the sea level, and since the temperatures of the lower station are known, these reductions and computations are usually capable of being carried out with greater accuracy than the reductions to the sea level, which have been described just above. As it may be useful to make a rough computation of the reduction of barometric pressure to sea level when the altitude, and consequently the vertical length, of the air column, is known, or to determine the difference in altitude when the approximate height of the column of air (equal to a difference in barometer readings) is given, the following table is offered for such use. It gives the vertical length, in feet, of a column of air corresponding to o.i of an inch of the barometric pres- sure at various air pressures and temperatures. 8o ELEMENTARY METEOROLOGY. VERTICAL LENGTH OF AN AIR COLUMN CORRESPONDING TO o.i OF AN INCH BAROMETRIC PRESSURE. AVERAGE TEMPERATURE IN DEGREES FAHRENHEIT. 20 30 4-? . ' III I 14. 116 I IQ 124. 124 126 *A 1 06 IOQ III 114. 116 121 121 2C IO2 IOC IO7 IOQ I 12 114. 116 *J 26 Q 8 IW 5 IOI IO"? iwy IOC IO7 I IO 1 12 2? 28 yu 94 QI 97 Q-J 1W J 99 QC IV O IOI Q8 103 IOO 1 06 IO'' 1 08 I O4. 2Q 88 QO Q2 Q4. q6 Q 8 IOO ^V JO 8c 87 8q QI yvj QI V" QC Q7 "3 uy V 1 yj y.> y/ Thus, when the barometer reads 28 inches and the temperature is 50 F., then at a point 98 feet lower in altitude the barometer will read o.i of an inch more than 28 inches, or 28.1 inches; or at a point 98 feet higher it will read 27.9 inches. Results of Observations of Atmospheric Pressure. The observations of barometric air pressure are not so fre- quent as those of air temperature, on account of the cost, complexity, and fragile construction of barometers. Still, enough observations have been made to allow the con- ditions and changes of the atmospheric pressure to be noted in most of the accessible parts of the surface of the globe. The observed conditions of atmospheric pressure have been studied locally, that is, for individual places ; and also in their geographical distribution, that is, in their connection with those of other places. Where but a single place is concerned, it is usual to consider the observed pressures, correcting them for instrumental errors only ; but where the geographical distribution is to be considered, it is necessary to reduce the observations to a common level. AIR PRESSURE. 81 Observed barometric pressures at individual places show the existence of both diurnal and annual periodic changes, somewhat similar to those found for temperatures, but in a less degree. These periodic changes are, however, masked by the far greater accidental changes which must be elimi- nated (by taking the average of many observations) before the periodic changes are visible. The Diurnal Change or March of the Air Pressure does not culminate, like the temperature, in a single maximum with a corresponding minimum, but there are two maxima and two minima. In general, the minima occur at about 4 h. and 16 h. (4 A.M. and 4 P.M.), and the maxima at about 10 h. and 22 h. (10 A.M. and 10 P.M.). The amplitudes of oscillation vary in different parts of the world, but are always small, amounting to only about 0.15 of an inch in the regions of greatest, and to o.oi of an inch in the regions of least oscillation. The amplitude is in general greatest near the equatorial regions, and diminishes towards the poles. The hourly pressures for a few stations are given in the following table : TABLE SHOWING DAILY MARCH OF THE BAROMETRIC PRESSURE ON THE AVERAGE FOR THE YEAR (hourly barometric pressures in Inches and hun- dredths} . I" 2 h 3 h 4 h 5 h 6h 7" 8* 9 h ioh Ilh ia> i3 h Key West, Fla. . 30.07 .06 05 S 05 .06 .07 .08 .09 .10 .09 .08 30.07 St. Paul, Minn. . 30-13 -13 !3 r 3 .14 .14 .14 15 15 15 15 .14 30.12 Fort Conger . . 29.844 .846 .847 .849 .849 .848 .847 .846 .844 .842 839 .838 29.839 I 4 h I5 h I6h I7 h i8h I 9 h 2Qh 2lh 22h 23 h 24*1 Av. Key West, Fla. . 30.05 .04 03 03 .04 05 .06 .07 .08 .08 30.08 30.06 St. Paul, Minn. . 30.11 .11 .11 .11 .11 .11 .12 .12 13 13 30.13 30-13 Fort Conger . . 29.841 .844 .846 .846 .846 .845 .845 845 .844 .842 29.843 29.844 82 ELEMENTARY METEOROLOGY, tnchesh It is to be noticed that for the Arctic regions it is necessary to give the pressures in thousandths of an inch in order to show the diurnal change. In the accompanying figure (Fig. 22) the daily march of the baro- metric pressure is given for Key West, St. Paul, Fort Conger in the Arctic region, and for Calcutta, India. In this diagram the average daily barometric pressure is entered as the horizontal line marked o.oo inches ; and the departure or the amount above or below this at the different hours is shown on the different lines, and the rise and fall by the course of. the curves. The principal maximum barometric pressure occurs be- fore noon, and the principal minimum after noon. During the winter season they approach noontime, but in summer they recede from it. The amplitudes of the diurnal oscillations are greater during the warmer than during the cooler part of the day, and for the dry than for the moist continental localities. They di- minish from the equatorial region of greatest solar heat towards the poles. The cause of these oscil- FIG. 22"! DIURNAL CHANGES OF AIR lations is not at present perfectly PRESSURE. understood \ but they may be due to waves in the atmosphere which depend in some way on the intensity and distribution of the solar heat. The Annual Change or March of the Air Pressure, that is, the change in the average pressure from month to month, presents a variety of characteristics for different regions of the earth. On the continents at the lower altitudes there is a maximum air pressure in winter, and a minimum in summer. On the ocean, on the contrary, the air pres- sure is highest in summer, and lowest in winter ; and this is also true of the air pressure at high altitudes. 0.05 West vtta 0.00 nger 005 ; \ / 1 \ L \ '/ \ \ /C \v. 5* 4 vv __ _ J V >- V / 2 "" \ /' VI \ (/ s / ; i , \ 0.10 2 AIR PRESSURE. This matter cannot be fully understood until the motions of the atmosphere are studied ; but it may be said in gen- eral, that over the continents in the summer the air becomes very much heated, and the surfaces of equal air pressure more elevated, which causes the air to flow out- ward (above) towards the oceans, while the air must flow inward (below) from the oceans towards the continents to compensate for the outflow above. Thus the warmer continental air causes a deficiency in the amount of air, and consequently a diminished air pressure, over the continent. The greatest and most regularly occurring variations in the mean air pressure take place on the conti- nents, and the least on the oceans and their coasts. There are, however, many local deviations from this law. *J~F M AM J J A fey West St.Paul 30.00 Calcutta Atlantic Ocean Ft.Conger Jjj^ 5 * s < ^- ^ -^ s- H n* -j \ s ~ \ y f *^ \ ' x / \ 7 FIG. 23. ANNUAL CHANGES OF AIR PRESSURE. In the interior of Asia the amplitude of oscillations amounts to over 0.70 of an inch in a year ; while over the Atlantic Ocean near the equator it is only about o.io of an inch. The following table shows the average monthly air pressures at a few points on the earth's surface : ANNUAL MARCH OF AIR PRESSURE {monthly barometer pressure in inches and hundredths}. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year Calcutta . . . 30 01 29-95 29.86 29.76 29.70 29-58 29-57 29.62 29.71 29.84 29-95 30.01 29.80 Key West . . . 30.14 30.12 30.10 30.04 30.01 30.04 30.06 30.01 29.99 29.99 30.06 30.12 30-05 St. Paul .... 30.13130.10 30.04 29.95 29.91 2qqo 29-93 29-95 29-97 29.99 30.04 30.09 29.99 Fort Conger . 29.80 29.67 30.07 29.88 29.79 29-83 29-77 29.9029.86 29.92 29.88 Atlantic Ocean 29-93 29-94 29.9229.93 29.94 29.98 30.04 30.02 30.01 29.97 29.94 29.92 29.96 The data in this table are shown graphically in Fig. 23. The Irregular or Non-periodic Oscillations in the Air Pressure are those changes which occur more or less 8 4 ELEMENTARY METEOROLOGY. gradually from time to time, and which do not belong to the regular diurnal or annual periodic changes. They are due to the movement, over the earth's surface, of areas of abnormally high or low barometric pressure. The monthly barometer oscillations are the difference between the highest and lowest barometer reading during the month. The follow- ing table shows the average barometric oscillations in the northern hemispheres, over the sea and over the land, for the winter and the summer seasons. AVERAGE MONTHLY BAROMETRIC OSCILLATIONS (inches and hundredths) . IN WINTER. IN SUMMER. NORTH LATITUDE. OVER THB OVER THE OVER THE OVER THE OCEAN. CONTINENTS. OCEAN. CONTINENTS. O.2O O.24 0.20 0.24 10 O.24 0.32 O.2O 0.24 20 0.32 o-43 O.24 0.32 30 0.63 0.51 0-35 0-43 40 .14 0.71 0.63 0.47 50 5 0.98 0.98 o-55 60 77 1.22 I.IO o-75 70 57 I.I 4 0.98 . 0.71 80 34 0.71 -~ The place of greatest oscillation is in the North Atlantic Ocean, be- tween Newfoundland and the British Islands, where it amounts to nearly 2 inches in winter. From this locality there is a decrease in the amplitude of oscillation, quite gradual on the north, east, and west sides, but very rapid towards the south ; and in the neighborhood of the equator it becomes almost identical with the periodic daily oscillations. These barometer oscillations are much greater over the sea than over the land ; and greater in winter than in summer, except near the equator, where they are quite uniform, as may be seen from the table just given. AIR PRESSURE. 85 Graphical Representations of the Distribution of Air Pres- sures. An isobaric line, or an isobar as it is usually called by meteorologists, is a line drawn through adjacent places which have the same barometric pressure. In order to draw isobaric lines on a chart, the barometric pressures reduced to sea level are written down on the map at the places where the observations are made. Then the iso- baric lines are drawn on the map usually for each even tenth of an inch of barometric pressure. An isobaric surface is a surface in the air all points of which have the same barometric pressure. When we wish to show the position or distribution of isobaric lines and isobaric surfaces in different parts of the atmosphere by means of diagrams, it is usual to take some plane or level surface, such as the sea level or some im- aginary surface parallel to it, and designate where the isobaric surfaces intersect this plane, and the positions which they assume with reference to it. The lines where the isobaric surfaces cut this plane are isobaric lines, or isobars. There are two ways in which the level of isobaric sur- faces is disturbed : - i. Air may be added to a region or taken from it by means of air currents. In case air is added, the pres- sure is increased, and the isobaric surfaces are forced upward from the earth's surface, and lie closer together than in adjacent regions where the amount of air is dimin- ished or unaltered. In case air is abstracted, the pressure is diminished, and the isobaric surfaces are forced down- ward, and lie farther apart than in adjacent regions. Such changes in the isobaric surfaces are made manifest all the way down to the earth's surface, and at the ground there is an increase or decrease of the air pressure. 86 ELEMENTARY METEOROLOGY. 2. The quantity of air over any given region may re- main unchanged, but the temperature may be varied ; in which case there is no increase or diminution of the air pressure at the ground, and the isobaric surfaces there remain unchanged, but at various elevations above the ground the positions of the isobaric surfaces will depend on the distribution of the temperature. This last condition has been chosen for showing in more detail the nature of isobaric surfaces and isobaric lines. Let us suppose that we are up in the air at some distance above the ground, and that we take the plane or level of this page held horizontally as a chosen level in the air, having a barometer reading of 29 inches, for example. It becomes necessary, then, to imagine the leaf of the book to be so perforated that the air can pass freely through it, and that no re- sistance is offered by it to the air moving either upwards or downwards. Suppose that the temperatures are uniform on this level, and there are also common temperatures at other levels above and below. Then for this region, inclosed by the circle in Fig. 24 (i), the barometric pressures are 29 inches. Now, let the temperature of the air become higher in the center at A, and for some distance above and below the plane, and colder on the outer edge at Z?, but so that the present condition of temperature and pressure is maintained at the points on a circle, C, surrounding A, and between A and B, Fig. 24 (2). Then, since warmer air is expanded and colder air is contracted, the warmer air at A on the level of the page will in Fig. 24 (-2) be forced upwards above the page towards the reader's eye, while the colder air at B will be contracted below or beneath the page ; but at C it will remain on the level of the page. When more air is forced above the plane of the page' at the center A, the former pressure of 29 inches on the page is increased by the amount of air pushed up : while on the outside, BB, the air pressure on the page is diminished by the amount of air drawn below the level of the page. But around the circle C, where the temperature remains the same, the air pressure remains 29 inches. We will suppose that the pressure on the level of the page at A is increased to 29.5 inches, and that on the outskirts BB it is reduced to 28.5 inches, then we shall have the pressures on the level or plane of the page at A, , and C, as shown in Fig. 24 (3). AIR PRESSURE. The pressure on the level of the page will then be greatest at the center A, and will gradually decrease to the outer limits B. We can then also draw on the page the locations where an,y of the air pressures between 29.5 inches and 28.5 inches will occur by taking the propor- tional distances. The air pressure 29.4 will occur on the circle drawn around A at I of the distance towards C; the air pressure 3 4 FIG. 24. ILLUSTRATION OF LOCAL INCREASE AND DECREASE OF AIR PRESSURE. 29.3, at another | of this distance, etc. : so that, if we wish to show the location of the pressures for each -^ of an inch, we should have a dia- gram like Fig. 24 (4). The numerals on these circles show the air pressure on this plane ; and they, with the position of the circles, show us the relative loca- 88 ELEMENTARY METEOROLOGY. tions, above and below this plane, of the isobaric surfaces which inter- sect it at these circles ; and this is the best graphical presentation that we can give of them, because it is so difficult to represent, on a sheet of paper having only length and breadth, the additional feature of thickness. It is by a similar graphical process that the air pressures over the earth are shown on charts. We trace out by isobaric lines the inter- section of the isobaric surfaces with some chosen level or plane, that of the sea level being usually selected ; and by so locating these isobaric surfaces as shown by observations, we can determine the regions of greatest and least air pressure. Geographical Distribution of Air Pressure. In compar- ing the average air pressures at various places, the observed pressures, as we have seen, must be reduced to a common level. It is usual to make the reduction to the sea level ; but any level may be chosen, and, as we shall see, other levels, are also sometimes used as a plane of reduction. Charts have been prepared, showing the distribution, over the surface of the earth, of the average air pressure (reduced to sea level) for each month of the year. The isobaric lines or isobars drawn on these charts show very clearly the regions of relatively high and low air pressures, and the intermediate conditions. Areas of relatively high air pressure are called barometric maxima, or sometimes simply highs; and areas of rela- tively low pressure, barometric minima, or simply lows. Only selected charts are given here for the year, and for the months of January and July, on which are indicated the isobaric lines which show where the isobaric surfaces cut the plane of the sea level. The order of the distribution of the air or barometric pressure depends on two great causes, primarily on the air temperatures, as shown by their distribution over the surface of the earth ; and secondarily on the movement of the air itself from one part of the earth's surface to another. The discussion of the relation of these causes to their 90 ELEMENTARY METEOROLOGY. effects must be deferred until the movements of the atmosphere are presented in a later chapter. The mere facts of the distribution of the air pressure are all that can be given at present. Distribution of the Average Air Pressure for the Year. This chart of isobaric lines at sea level for the year (Fig., 25) shows, in the northern hemisphere, an area of high air pressure of more than 30.1 inches over eastern Asia; another with a pressure of more than 30.2 inches over the eastern North Pacific Ocean, in the same latitude ; and another of more than 30 inches over North America. In the southern hemisphere the regions of high pressure are, over the Atlantic Ocean, near latitude 25, more than 30.1 inches ; over the Indian Ocean, west of Australia, more than 30.1 inches; over the eastern Pacific Ocean, between latitude 30 and 40, more than 30.2 inches. Areas of low air pressure are found, for the northern hemisphere, over the North Atlantic Ocean, near Iceland, with pressure below 29.7 inches; over the extreme north- ern Pacific Ocean, with pressure below 29.7 inches; and to the southeast and southwest and over the southern part of Asia, with pressure below 29.8 inches. In the southern hemisphere the regions of low air pressure are over north- ern Australia and over the Antarctic Ocean, where there is a quite uniform decrease from 30 inches just north of latitude 40, to 29.3 inches in latitude 60. Distribution of the Average Air Pressure for January. The chart of the isobaric lines at sea level, for the month of January (Fig. 26), shows that in the northern hemisphere there is an area of high barometric pressure over central Asia of more than 30.5 inches, and over central North America of more than 30.2 inches; in both cases lying mainly north of the 4Oth parallel. In the southern hemisphere the areas of maximum pressure of 92 ELEMENTARY METEOROLOGY. 30.1 inches are on the equatorial side of the 4Oth parallel over the oceans to the westward of Australia, Africa, and South America. Areas of lowest pressure in the northern hemisphere are, over the North Atlantic, near Iceland, less than 29.5 inches ; and over the central North Pacific, less than 29.6 inches. In the southern hemisphere the areas of lowest pressure are, over northern Australia, less than 29.7 inches ; the Indian Ocean, less than 29.8 inches ; and southern Africa, 29.8 inches. South of the 4Oth parallel there is a rapid but regular decrease from 29.9 inches, to about 29 inches in the Antarctic latitudes. The distribution of the average air pressure for January reduced to planes above the sea level (as indicated by isobaric lines drawn for other elevations than that of the sea level as just given) shows that for the January average air pressures the above-mentioned maximum at the earth's surface over Asia exists only for the lower air layers ; and at an altitude of about 5,000 feet, the North American and European maxi- mum air pressures disappear. At an altitude of about 10,000 feet, the maximum air pressure lies over the equatorial region, and it decreases thence towards the poles. The distribution of air pressure for January, at an altitude of 13.000 feet, is about as follows : Over the continents in the northern hemisphere there is a region of low pressure over northern North America (16.8 inches) and Asia (16.5 inches) ; but on the intervening oceans the pressure is higher, especially over the North Atlantic Ocean (17.3 inches). From these northern regions there is an increase with southward progress, more rapid over the land than over the sea, down to the region between the equator and the Tropic of Capricorn, where a maximum pressure of about 18.7 inches is reached. Within this maximum region the pressures vary somewhat, and are a little less over the Atlantic Ocean than elsewhere. From this region of high pressure there is a decrease towards the south, and a pressure of about 17.7 inches is reached at latitude 50 south. It is to be particularly noticed that the courses of the isobars are 94 ELEMENTARY METEOROLOGY. very irregular over the northern hemisphere, where the earth's surface is mainly land, and that they run regularly along the parallels of lati- tude in the southern hemisphere, where the surface is mostly water. Distribution of Air Pressure at Sea Level for July (Fig. 27). The areas of high pressure for July in the northern hemisphere lie north of the Tropic of Cancer (over the Atlantic Ocean, 30.2 inches; and over the Pacific Ocean, 30.3 inches). In the southern hemisphere the areas of high pressure lie over the oceans, south of the Tropic of Capricorn (in the eastern Pacific and mid-Atlantic, 30.2 inches; and in the western Indian Ocean, 30.3 inches). The areas of low pressure in the northern hemisphere lie over the continents, a little north of the Tropic of Cancer (in western Asia, 29.4 inches ; in western America, 29.8 inches). There is also an area of low pressure over the Atlantic near Iceland (29.8 inches). In the southern hemisphere, south of latitude 40, the pressure is below 30 inches, and decreases rapidly towards the Antarctic regions, being 29.4 inches in latitude 60, For July, at the higher planes above the sea level, the distribution of the air pressure differs from that just described. At about 5,000 feet altitude, the curves are much more regular than at sea level. At about lo.ooo feet altitude, the zone of the highest pressure approaches the equator, and at 13,000 feet it reaches the equatorial region. The distribution of air pressure for July, at an altitude of 13.000 feet, is as follows : The region of highest pressure lies mainly on the north side of the equator, and extends even beyond the Tropic of Cancer. This region of maximum pressure is somewhat broken ; and over the center of Africa, southern North America, the Middle Atlantic and Pacific oceans, the pressure reaches maximum values of 18.6 inches or 18.7 inches, which is slightly above that of the intervening regions along the Tropic of Cancer. To the northward the pressure decreases to 17.8 inches over northern North America, and 17.9 inches over the North Atlantic and northern Asia ; but the decrease is more gradual AIR PRESSURE. 95 over the western coasts than over the eastern coasts. To the south- ward of the equator the decrease is much more regular than towards the north, and at about latitude 55 south the pressure (17.6 inches) is without many irregularities with change of longitude ; which shows, that, even at this altitude, the difference between land and water sur- faces is noticeable. Average Air Pressure along a Meridian. The average change of air pressure with latitude, or the change along an average meridian, may be seen in the table on p. 96, which gives the average or normal pressure for each 5 of latitude for the year, January, and July, at sea level; and for the year at altitudes of about 6,500 feet and 13,000 feet. For the year, at the sea level, there is a maximum air pressure at about latitude 35 in the northern hemisphere, and 30 in the southern ; with a decrease on both sides towards the equator and towards the poles. This pole- ward decrease is very much greater in the southern hemi- sphere than in the northern, due to the effects of the uniform water surface in the former; and in fact in the northern hemisphere there is a slight increase again at the far north. The decrease toward the equator is very slight. At 6,500 feet altitude above the sea level, the place of maximum pressure in each hemisphere approaches the equator, and is quite uniform near it, and the decrease towards the poles is much more nearly equal in the two hemispheres. At 13,000 feet altitude, a single maximum occurs near the equator (a little to the south of it), and the diminution of the pressure towards the poles is still more symmetrical for the two hemispheres. The area of high pressure at the north pole at sea level seems to disappear at a slight altitude, just as it does in the more local region of eastern Siberia. 9 6 ELEMENTARY METEOROLOGY. AVERAGE AIR PRESSURES ALONG A MERIDIAN. AVERAGE AIR PRESSURE AT SEA LEVEL. AVERAGE AIR PRESSURE FOR THE YEAR AT ALTITUDE OF LATITUDE. YEAR. JAN. JULY. 6,562 FT. 13,^23 FT. Inches. Inches. Inches. Inches. Inches. North 80 ... 29.941 29-937 29-945 22.913 17.528 75 ... 29.921 29.929 29.874 70 ... 29.866 29.882 29.850 22.976 17.583 6 5 ... 29.850 29.874 29.827 60 ... 29.870 29.910 29.831 23-I34 17.791 55 ..- 29.910 29.961 29.858 50 . - - 29.949 30.004 29.894 23-347 17.992 45 ... 29.980 30.039 29.921 40 ... 30.000 30.063 29.937 23-543 18.252 35 -.. 30.016 30.083 29.949 30 ... 29.988 30.055 29.921 23-658 18.437 25" ... 29.937 30.000 29.874 20 ... 29.890 29.945 29.835 23-658 18.500 15 -.. 29.854 29.894 29.815 10 . . . 29.839 29.858 29.819 23-658 18.532 North 5 ... 29- 8 43 29.843 29.839 Equator o . 29.843 29.819 29.866 23-665 18.543 South 5 ... 29.854 29.807 29.902 10 . . . 29.886 29.819 29.953 23-685 18-547 15 29.929 29.850 30.008 20 ... 29.988 29.902 3 -075 23.728 18-547 25 ... 30.047 29-953 30.154 30 ... 30-059 29-973 30.146 23.709 18.476 35 30.016 29-945 30.087 40- ... 29.941 29.886 29.996 23-508 18.232 45 -.. 29.815 29.776 29.854 50 ... 29.654 29.634 29.673 23-150* 17.862 55 ... 29-457 29-457 29-457 60 ... 29.268 22.717 17.476 65 ... 29.122 South 70 . . . 29-055 AIR PRESSURE. 97 The accompanying figure (Fig. 28) shows the general slope of the isobaric surfaces along a meridian, at sea level and at various altitudes above it. The dotted line shows where the isobaric surface of 30 inches lies below the sea level, the pressure at that level being less than 30 inches. Attitude 6,500 feet____ 23.5 inches' Altitude 0,2,000 feet N.Latitude Sea Level Isobaric Surface 20' 20 40 *".-_ 80 FIG. 28. SLOPE OF ISOBARIC SURFACES ALONG A MERIDIAN AT VARIOUS ALTITUDES. The Variability of the Average Air Pressure increases with the latitude, and to a certain extent with the increase of the distance inland. With increase of altitude above the sea level, the variability decreases in winter, but in- creases in the summer for low altitudes. The variability is greater in winter than in summer. Long-period Oscillations of Air Pressure, extending over a period of a little more than 30 years, undoubtedly exist ; but the extreme average maximum pressure for a number of places is not more than 0.06 of an inch greater than the extreme average minimum. Slight as these variations are, they are sufficient to cause great climatic oscillations, owing to their effects on the atmospheric circulation. It is very probable that when there is an excess of air pressure over the continents, there is a deficiency over the oceans. Physiological Effects of Air Pressure. The outside sur- face of the body exposed to the air pressure, in the case of a grown person, is about 16 square feet, and this would WALDO METEOR. 6 98 ELEMENTARY METEOROLOGY. give a pressure of about 35,000 pounds. If it were not for the ease with which this air penetrates the body under this pressure, very slight changes in it would prove dis- astrous to life. The variations in this pressure may arise in two ways, first, through the changes in the air pressure at the same altitude (due to the passage of areas of high and low barometric pressure, in which the extreme cases would give a variation of not over 4 inches) from about 27.5 inches to 31.5 inches; second, through the change in altitude and increase in pressure by making a descent, or decrease in pressure by an ascent. In the former case of varying the pressure merely, a maximum change of O.8 of an inch air pressure in a day would be equivalent to changing the altitude by about 690 feet during that time ; or in the extreme case of an absolute change from 31.5 inches to 27.5 inches air pressure, it would be the same as coming from a depth of about 1,378 feet below the ground surface, and ascending to a height of about 2,133 ^ eet above this surface. In other words, the changes in air pressure which we may experience at the earth's surface are as great as those which would be encountered in an ascent of about 3,500 feet in altitude. There are, however, limits to the amount that the air pressure can be changed without harm. Just what these limits are, we do not know, and they would vary for different persons ; but one individual (Hum- boldt) was exposed to a barometric pressure of about 48 inches in a diving bell, and to 14.75 inches on a mountain top; others have been exposed to a still lower air pressure. It is probable that animals can stand a somewhat greater increase than decrease of pressure. In cases of such extreme variations of air pressure the changes must take place gradually. If they take place too rapidly, fainting, bursting of blood vessels, or even death, will result. If the change takes place slowly, then, as the limit of endurance is approached, certain physiological effects are made apparent. These AIR PRESSURE. 99 are the symptoms of the so-called mountain sickness, such as difficulty of breathing, headache, pressure on the eardrums, and a general apathy, which prevents exertion of any kind. These symptoms disappear, however, when lower altitudes are again reached. Permanent residence at very high altitudes does, however, impoverish the blood by decreas- ing the amount of oxygen inhaled in breathing. Classified Distribution of the Air Pressures. The whole lower air mass may be divided into regions of low and high air pressure, or areas of deficiency or excess, between which lie areas of more nearly normal pressure. Regions of Low Barometric Pressure may be divided into four classes or orders : 1. Depressions of the first order are those belonging to the air system of whole hemispheres. They include those permanent ones extending equatorward and poleward from about latitude 30. They are of permanent character. 2. Depressions of the second order are of great extent, covering areas many hundreds of miles in diameter, and occur mainly in the middle and lower latitudes. They are not permanent as regards either location or duration. 3. Depressions of the third order are minor depressions occurring within depressions of the second order. 4. Depressions of the fourth order are those arising from the local excess of temperature. They are fixed in posi- tion, but are not permanent. Such occur during the day- time over limited isolated land areas (as, for instance, the Spanish Peninsula). Regions of High Barometric Pressure may be divided into three classes or orders, as follows : I. Regions of high pressure of the first order are the permanent hemispherical ones to be found in the lower, middle, and north polar latitudes. They extend upwards to considerable altitudes. 100 ELEMENTARY METEOROLOGY. 2. Regions of high pressure of the second order are those resulting from the heaping-up of air between two depressions of the second order. They are therefore not permanent in either duration or location. 3. Regions of high pressure of the third order are caused by the local cooling of the air at the earth's surface. They are of local occurrence, and do not extend to great altitudes. Barometric Gradient. By gradient is usually meant the degree of steepness of a slope. In meteorology we speak of the steepness of the slope of isobaric surfaces as the barometric gradient. Instead of measuring the angle of the slope, it is customary to take the difference in baro- metric pressure along the same level. As a unit of distance in all directions, the length of a degree of the meridian is taken. The barometric gradient between two places is, then, the difference in the barometric pres- sures at the same level, divided by the number of merid- ional degrees between the two places. Thus, if at one place the barometric pressure is 30 inches, and at another place the pressure (reduced to the same level) is 30.50 inches, and the distance between the two places is five meridional degrees, then the barometric gradient between the two places would be (30. 5 30) -5- 5, or o. I, of an inch. Wherever a gradient exists, there is a gradient force acting, due to the efforts of gravity to render the isobaric surfaces level. The amount of this gradient force, in the case of isobaric surfaces, increases with increase of the magnitude of the gradient ; that is, the steeper the incli- nation of the isobaric surfaces, the greater the gradient force. This gradient force is effective, or operates, in the direction of the slope of the isobaric surfaces, and air flows down a sloping isobaric surface in a manner somewhat similar to the flowing of water down an incline. CHAPTER IV. WINDS Air Currents. When the air is at rest, relative to sur- rounding objects, we are scarcely aware of its existence; but when it is in motion, we notice that it exerts a pressure or force against these objects. Air in motion is called wind. When air is at rest, it is said to be calm. The air of the atmosphere would as a whole be continually calm, if it were not for the unequal distribution of temperature over the earth's surface and through the air itself. This unequal heating gives rise to air motion or air currents, the passage of which gives the phenomenon of wind. The air in moving must go from some one place to another ; and therefore wind has direction, as we see from the movement of light objects, such as leaves or clouds, which are carried along by the wind. There must also be a rate of motion, which we call wind velocity ; and we judge of this by the rapidity with which the air transports such light objects. Heavy objects lag behind the true motion of the air. Since air has density or mass, then when it is in motion it must exert a pressure or force, called wind force or wind pressure, against objects with which it comes in contact, as is realized by the way the wind bends trees, or blows against our bodies. Air currents moving approximately parallel to the sur- face of the ocean are called horizontal air currents. Air 101 102 ELEMENTARY METEOROLOGY. currents moving approximately perpendicular to the sur- face of the ocean are called vertical air currents. The possible movement of vertical currents is limited to a few miles between the earth's surface and the outer limit of the air, while horizontal air currents might make the entire circuit of the earth. The usual direction .of the wind is some combination of these two directions. The vertical movement of the air is usually so slight, as compared with the horizontal movement, that ordinarily the latter only is observed for general winds. The Winds of the Globe may be divided into the follow- ing classes : 1. A permanent and continual interchange of air be- tween the equatorial and polar regions, due to the differ- ences in temperature between those regions : this is called the general circulation of the atmosphere. The annual shifting of the thermal equator (which follows the latitude of the sun) causes an annual displacement and inequality of this system of winds. 2. An interchange of air between the masses lying over bodies of water and of land, due to the unequal heat- ing of the two surfaces. When the interchange takes place over great regions, as between continents and oceans, it has an annual period of occurrence : this is called the monsoon wind. When the interchange takes place between the coast lands and the coast waters only, it is diurnal in occurrence : this is called the land and sea breeze. 3. An interchange of air between mountains and val- leys, due to the unequal heating and cooling of the two localities : this is called the mountain and valley breeze, and is diurnal in its occurrence. 4. An interchange of air between extensive regions of WINDS. 103 higher and lower barometric air pressures : this gives rise to the cyclonic and anticyclonic winds, which are irregular in occurrence. 5. A local rush of air for the restoration of the normal condition (stable equilibrium) when the latter has been dis- turbed by local causes : to this class belong spouts, torna- does, and squalls, which are of occasional occurrence, 6. A rush of air which is produced in front of ava- lanches and land slides : this is called avalanche wind, and is of occasional occurrence. 7. The outrush of air which takes place with volcanic eruptions : this may be called volcanic wind, and is of occasional occurrence. 8. To these must be added another class of winds, which may be called eddy winds or whirlwinds. These occur as follows : When a current of air is flowing steadily along, there break out at intervals from the sides whirls or eddies such as may be seen in flowing water ; or they may be produced by the meeting of two air currents differ- ing in direction. These are irregular in occurrence, and vary in magnitude from large whirls produced by the gen- eral air currents (No. i) to the small dust whirls which we see in the streets. Direction of the Wind. The direction of the wind is named from the direction of approach. Thus, if a wind blows from the north, it is said to be a north wind ; if it blows from the east, it is said to be an east wind ; etc. In accurate work in meteorology, 16 points of compass are used, as follows : North, north northeast, northeast, east northeast, east, east south- east, southeast, south southeast, south, south southwest, southwest, west southwest, west, west northwest, northwest, and north northwest. These are indicated by their initial letters, as shown in Fig. 29. 104 ELEMENTARY METEOROLOGY. N. The Horizontal Direction of the Wind is usually observed by means of an instrument called the arrow wind vane (Fig. 30), the head of the arrow pointing in the direction of approach of the wind. The arrow is free to re- volve horizontally around a vertical axis placed near the head. The best vanes have a divided tail with a divergence of about 22. The Velocity of the Wind is the rapidity with which the air moves past some fixed point, or covers the known distance between some two points. Wind velocities are usually given in miles per hour, or feet per second, in English measure. The Anemometer. An instrument used to measure the velocity of the wind is called an anemometer. It is a small wind wheel FIG. 29. POINTS OF COMPASS FOR WIND DIRECTIONS. Vertical Sec. FIG. 30. ARROW WIND VANE. with an attachment by ~T^ Horizontal sec. means of which the num- ber of revolutions of the wheel is indicated on a little dial. When exposed to the wind, the rapidity of revolution of the wind wheel varies with the actual veloc- ity of the wind. The commonest form is the Robinson cup anemometer shown in Fig. 31. The four hemispherical cups revolve as a whole around a vertical axis, when exposed to the wind. It has been found that the distance traveled by any one of the cups, when multiplied by about 2.5, will give WINDS. 105 the velocity of the wind. The distance passed over in one revolution of the cup is found by multiplying the distance from the center of the vertical axis to the center of the cup by (2) x (3.1416). The dial on the vertical axis shows the number of miles of wind. The Force of the Wind is .the pressure which it exerts on a flat surface held perpendicular to its direction of mo- tion. It may be judged by an observer, or ex- pressed in actual pounds of pressure per square foot of exposed surface. Observation of Wind Direction. In most cases when wind direction is observed two or three times daily, as in ordi- nary observations, the results are tabulated for each month and for the year by giving the num- ber of times the wind blows from each of the eight principal points of the compass, and the number of calms during the time. Thus in 1891 the number of times the wind blew from various direc- tions at St. Paul during the year and the extreme months, as shown by two observations daily, was as follows : FIG. 31. ROBINSON CUP ANEMOMETER. N. N. E. E. S. E. S. S.W. w. N. W. Calm. January . . . July Year .... 2 3 4. O 4 -JQ 2 8 1:7 23 19 224 3 AT. 4 9 oo 7 7 84. 17- 9 1 1 1 4 2 c6 o u 106 ELEMENTARY METEOROLOGY. The direction from which the winds blow with greatest frequency is called the direction of prevailing winds. It varies for different regions. Tables have been published which give the relative frequency of the wind from the different points of compass, for the months and the year, for various regions of the earth ; and charts have been drawn which give the direction of the prevailing winds, for each month and the year, for portions of the continents and traversed seas of the globe. (See wind directions given on the charts, Figs. 26/27, showing the distribution of air pressure at sea level over the whole earth. The arrows fly with the wind.) Where the wind blows at different times in different directions, all of these motions combined are equivalent to a motion in a single direction, called a resultant direction. This may be explained graphically as follows : If we start at a point A, and draw a line in the proper direction equal to the amount of north wind, and from this end of the line draw another line in the proper direction equal to the amount of the -northeast wind, and from the end of this line lay off in the proper direction another line equal to the amount of the east wind, and so on for all the wind directions, then a line drawn from A to the end of the last line drawn, will represent the direction, and in length the amount, of the resultant wind. Observation of Wind Velocities. Wind velocities are subject to diurnal and annual periodic variations and to marked irregular changes. The diurnal change of wind velocities has in general but a single maximum and mini- mum. The maximum usually occurs at about two to three hours after noon for the average of the year on the land near the surface of the ground, and about the same time before noon on the ocean ; and the hour of minimum varies from two to eight hours after midnight on the land, but is nearer midnight (it may be before or after) on the ocean. The diurnal amplitude or difference between the maximum and minimum wind is small over the ocean, but over the WINDS. 107 land it is large. In the United States it varies from 20 to 1 20 per cent of the average velocity of the wind. The daily variations in the wind velocity at considerable altitudes above the ground are the reverse of those at the earth's surface as to time of occurrence of principal phases. From observations which have been made, it appears, that, at an altitude of perhaps 400 to 600 feet above the ground, the daily change in the wind velocity FIG. 32. HOUR OF MAXIMUM WIND, IN THE UNITED STATES, FOR JANUARY. practically disappears, while at greater altitudes occur phases the reverse of those at the surface of the ground, but with less amplitudes. In the warm season of the year the maximum and minimum phases are much more pro- nounced, and the time of maximum is later in the after- noon than in the colder season. During the night there is little change in the wind velocity ; but as soon as the sun's heat makes itself felt, there is a more or less rapid increase near the surface of the ground, and a slighter decrease at greater altitudes. io8 ELEMENTARY METEOROLOGY, The cause of this has been explained as follows : The wind veloci- ties increase with the altitude above the ground to a certain extent. When the ground is heated by the sun's rays, the warm air rises, and cooler air from above must descend to take its place. This warm air from the ground surface carries with it the lesser (horizontal) velocities at the surface, and mixing with the upper air retards its horizontal movement ; while- the descending upper air takes with it the greater horizontal velocities at the higher altitudes, and thus accelerates the horizontal movement of the lower air with which it mixes. Some of FIG. 33. HOUR OF MAXIMUM WIND, IN THE UNITED STATES, FOR JULY. the increased velocity of the lower air is directly due to the increased air movement caused by the warming of the air during the daytime, and would take place even if the velocities aloft were not greater than those below. The accompanying charts (Figs. 32, 33) show the continental dis- tribution of the hours of maximum wind for January and July. Fig. 34 shows the daily march of the wind velocities, in miles per hour, for the extreme months of January and July at a few selected stations in the United States. Great absolute diurnal variations in the wind velocity occur in some localities ; as, for instance, in July for San Francisco and Corpus Christi (coast stations), and Whipple Barracks (an inland station). WINDS. I0 9 On cloudy days the daily amplitude of wind velocity is relatively small, because the action of the sun is weak. I i s j, January 10 2 4 6 8 10 12 14 16 JQ 1 2 4 6 8 ?0 12 U 16 18 20 22 %T 15 / \ /^ ^ 12 s ^ 11 / S Key West 11. 10 9 11 10 Corpus Christi 9 Texas 7 St.Paul e 5 9 8 7 6 lilpple Barracksb Arizona 3 San Francisco 7 6 10 ^ S S^ 8 ,** "^, ~\ x ^-- 1 ^- . ^ 7 ^ ""^ j S. s ,X ^ 6 X, ^N x' '^-4 ^- ^ 14 ^ 13 12 \ -^* . " 11 ^ / \ 10 \ ^ ^ 9 y \ 8 \ s ~\ 7 \ f v^ 6 \ ^ ^^^ -N 5 \J f \ / 4 7 f ll \ *^, -" ^v_ 3 10 ^ ^ > 9 / \ / ^ 8 / I 7 / \ \ 6 \ V 5 K^ / S 4 f ITS \ "^ 3 ^ 1 16 \ < - 2 "\. y 1 15 \ 1 14 13 , ^^ 12 1 \ X K^ 11 i \ 1 ' f ^^ 10 j 9 ^ 8 \ / 7 \ 6 \^ ^ FIG. 34. DIURNAL CHANGES IN WIND VELOCITIES (MILES PER HOUR). The Annual March of the Wind Velocity has in general but a single maximum (usually during the cold season) and minimum (in the warm season). The maximum occurs about March, and the minimum about August, in most of the United States ; but in some portions of the 1 10 ELEMENTARY METEOROLOGY. western section the primary maximum occurs about April, and there is a slight secondary minimum about December. In most of Europe the maximum wind occurs in a winter month., J FM AM J J AS O X D St.Paul 7 San Francisco - / s \ / \ / \ / ,' N\ ^' ">, 1 .^^- N, , \ ' ~t <' N \ ) .^, _, .x* s. v. L- The Annual Amplitude of Oscil- lation of the average hourly wind velocities for the extreme months is variable for different regions, amounting to over 10 miles per hour in some cases; that is, the windiest month may have an aver- age of over 10 miles per hour more FIG.. 35- -ANNUAL CHANGES w i n d than the Calmest month. WIND VELOCITIES (MILES PERHOUR). In the United States the amplitude is about 4 or 5 miles per hour, but on the northwestern coast it is 1 1 miles per hour. On the central Great Plains there is relatively slight variation in the wind velocity for the different months of the year. The following table and diagram (Fig. 35) show the average monthly wind velocities, and consequently the annual march of the wind, at selected places in the United States : AVERAGE MONTHLY WIND VELOCITIES (miles per hour) . J. F. M. A. M. J. J. A. S. O. N. D. Year Block Island 16.3 16.0 15-3 134 12.2 u.6 II.O II.O 13.0 14.0 iS.5 16.9 13.8 Key West 10.4 10.2 10.6 9.8 8. 9 7.6 7-3 74 8.0 10.7 10.7 10.7 9-3 St. Paul 7-i 7-5 8-3 9.0 8.5 7-5 6-5 6.7 7-7 7-9 74 7.0 7-6 Dodge City, Kan. . . 10.3 10.5 u.6 13-3 12.9 I2.I II. I 10.4 II.2 10.7 9-5 9-7 ii. i Corpus Christi, Tex. 10.4 10.9 I2.I 13-7 13-3 12.3 10.7 10.7 10.6 10.5 1 0.0 IO.I "3 San Francisco .... 6-7 7-i 8.4 9-5 n.8 II.7 II. I 1 1.8 9.3 94 6.2 6-5 9.0 Tatoosh Island, Wash. 16.0 14.4 I2. 7 10.6 10.3 7-9 7.6 74 IO.I 11.7 14.8 iS-7 u.6 The Average Wind Velocity for the Whole Year has not been determined at enough places to permit the show- ing of lines of equal wind velocities for the whole earth. These have been drawn, however, for the United States WINDS. 1 1 1 (p. 355) and for the Russian Empire, and roughly indicated for some portions of the oceans. In the United States the wind velocities are greatest on the coast, and in general decrease towards the interior ; but on the treeless Great Plains near the center of the continent there is an increase again to nearly the same velocities as are found along the low shores of the ocean. In Russian Siberia there is a decrease towards the center of the continent ; but there is no central region of increased wind, as found in the United States. Whether this is due entirely to the difference in the physical features of the two regions, or is partly due to the higher lati- tudes of Central Siberia, has not been determined. The Increase of the Wind Velocity with Increase of Alti- tude is very rapid for the first hundred or two hundred feet ; but above that it is slow, and very variable not only for the yearly averages, but also for the different months of the year. It is quite probable that up to some unknown altitude the wind velocity increases, and at higher altitudes it de- creases again. The increase in wind velocity with altitude is very much less over the ocean than on the land, for the first few hundred feet; but above this altitude there is probably not much difference between the wind velocities over a land and those over a water surface. The move- ments of clouds show that wind velocities of perhaps 200 miles per hour sometimes occur at high altitudes. Obstacles to Air Motions. Air movement does not progress unhindered, for the moving air layers near the ground rub against it, and those up above rub against other air layers which have a different direction of motion, or no motion at all. This rubbing is called friction, and it retards the air currents. The friction of one current of air on another is exceed- 112 ELEMENTARY METEOROLOGY. ingly slight ; but the friction of the air against the earth is very great. The main hindrance to the air currents acquiring enor- mous velocities is the continual mixing or mingling of air layers having different directions of motion, and especially the breaking-up of the air into vortices, within which the air layers which had the initial velocities become so increased in numbers, and separated, and so broken up into strata which twist spirally around one another, that the friction, and especially the interference of such innumerable sur- faces having inequalities of motion, becomes very great, and the motion is thus equalized. Mountain ranges may also somewhat retard the velocities of the lower air currents. The Decrease of Wind Velocities through Friction with the Ground is well shown by comparing the wind velocities on the land, where the friction and direct obstacles are greatest, with those on the ocean, where they are least. It is found by observation that the wind velocities over the land at the height of low buildings (40 or 50 feet above ground) are about 25 %, at the height of the tallest buildings (looto 150 feet above ground) about 50%, and on the well-exposed seashore about 75 %, of the normal velocities existing on the open ocean (40 feet above the water) for the same geographical locality. The Variation of Lower Wind Velocity with the Latitude is not easy to determine in amount, on account of the different exposure of anemometers ; but it is thought that in the northern hemisphere, at least, there is an increase in wind with increase of latitude up to latitude 50- 60, and then a decrease farther poleward. The discussion of a few observations on the open Atlantic Ocean showed an increase, in the average wind for the whole year, of from about 8 miles per hour in latitude 22 D or 23 north, to 17 miles per hour WINDS. 113 at about latitude 50 north, from which latitude northward to 58 north there was again a slight decrease ; but how this decrease continues onward towards the pole has not been determined. The general region of maximum wind in middle and higher latitudes is shown for the Atlantic Ocean in Figs. 36 and 37, which give for January and July the wind relations of the Atlantic Ocean as regards both the direction, and roughly the force, of the wind. The heavier arrows denote the stronger wind, and the double arrows the strongest wind; short arrows denote variable, and long arrows steady wind; circles show region of prevailing calms. These show in a general way the increase in wind velocities from the tropics up to the higher middle latitudes. The Periodic Daily Change in the Wind's Direction is rela- tively slightly marked, but still it is sufficient to have been recognized. It stands in close connection with the daily variation in velocity, since the vertical air currents occur- ring about midday and in the early afternoon communi- cate not only part of their velocity, but also their direction, to the lower air currents. The air currents above have a tendency to move or deviate to the right of the lower air currents, and this motion causes in the northern hemi- sphere, over a level land surface, a tendency for the wind in the morning to turn in the direction of the hands of a watch or movement of the sun, and towards evening to turn in the opposite direction ; but in the higher atmosphere (as on high mountain peaks) this is reversed. In the southern hemisphere, on the level land surface, the morning turning is in the direction opposite to the motion of the hands of a watch, and towards evening with it ; and in the upper atmosphere this is reversed. At the equator the vertical air currents do not cause this diurnal turning of the wind's direction. On the ocean the tendency for the wind's direction to turn as described is very slight on account of the weak WALDO METEOR. 7 w~5i r ^* > >O i m , r~^^^ ir '^S's'ji > >VH ~~* y/Pvv-~fT5><5/lo / Pvl -4 u- L-r -^^Kx ^//k-^ c^-- -- Longitude (i() W,. s t 10 I'min :iO Greenwich Longitude 20 Kast FIG. 36. WINDS or THE ATLANTIC OCEAN, JANUARY (DEUTSCHE SEEWARTE). 100 80 ^Mo 80 Longitude 60 West 40 from 20 Greenwich Longitude 20 East FIG. 37. WINDS OF THE ATLANTIC OCEAN, JULY (DEUTSCHE SEEWARTE). (H5) ELEMENTARY METEOROLOGY. vertical currents which arise over the relatively cool water in the daytime. The local diurnal changes in the wind's direction (known as sea and land breezes, and mountain and valley winds) are described in treating of the atmospheric motions (p. 262). The Periodic Annual Changes in the Wind's Direction are due to the seasonal shifting of the temperature conditions which have already been mentioned. First, there is the change in the general circulation of the atmosphere, due to the shifting of the region of greatest heat, first to the north, and then to the south, of the equator. Secondly, there is an inflow of the lower air towards the center of the continents in summer, and an outflow from the continents in winter. These are called the monsoon winds ; and they cause reversals of the direction of the general flow of air in the regions affected. Thirdly, there is a seasonal variation in the general courses or tracks of barometric minima which move across the earth's surface, and this changes the general direction of the wind in the region lying in their paths. In the middle latitudes, where these annual changes are most marked, the direction of the wind is a most important factor in determining the climate of the region. The resultant wind directions at selected stations are, Jan. Feb. March April May June St. Paul . . Key West . S. 73 W. N. 59 E. s. 7 o- w. N. 66 E. N. 62- \V. N. 76 E. N. 10 W. N. 87 E. S. 75 E. N. 82 E. S. 4 E. S. 65 E. July Aug. Sept. Oct. Nov. Dec. Year St. Paul . . . Key West . . S. 31 W. S. 71 E. S. 4 E. S. 76 E. N. 88 E. S. 37 W. N. 60 E. S. *2 W. N. 5 4E. S.79-W. N. 5 3E. N. 78 E. WINDS. The resultant direction of the horizontal and vertical components of the wind has never been determined with any great accuracy, but it must on the average amount to only a few degrees from the horizontal. Formation of Air Waves. It is seen that when the wind blows over a water surface there results the well- known phenomenon of water waves. Similarly, if one cur- rent of air moves over another at rest or of a different velocity or direction, and the line of demarcation between the two is sharply drawn, there will be a favorable con- dition for the formation of air waves. Such conditions do occur very frequently in our atmosphere, and since the effects of friction between two air currents moving one over the other are so very slight, we can have very great differences in velocity existing between the two within a short distance of the boundary between them. Air waves have a magnitude 2,500 times that of water waves pro- duced by the same wind conditions. The summits or crests of air waves are made visible by some of the detached clouds high up in the sky ; and their troughs are made evident by the gusts of wind felt in their passage near the surface of the ground. The further discussion of the winds will be taken up in other chap ters on the subject of the movements of the atmosphere. CHAPTER V. MOISTURE: VAPOR, CLOUD. Moisture of the Air. There is a continually repeated interchange of water going on between the air and the surface of the earth, and in considering the circuit through which the water passes in these changes it is most con- venient to separate it into three sections ; viz., 1. The moisture as it exists in the atmosphere as invis- ible vapor and as cloud, or the hygrometry of the air. 2. The rate at which it condenses and descends to the earth as precipitation, and the amount of this precipita- tion ; or the rain, hail, and snow fall. 3. The rate at which the moisture is taken up again into the air from the earth's surface, or the evaporation. Moisture exists, either as a gas or vapor, in an invisible form ; or as a liquid or solid, in visible form. The change from one form to another depends mainly on the con- ditions of temperature. For a given temperature, only a certain maximum amount of moisture can exist as a vapor. When this temperature is reached by cooling, saturation occurs, and any further cooling condenses the moisture, and renders it visible. Evaporation is that process by which a liquid is con* verted into its vapor ; as, for instance, the conversion of water into water vapor. Some solids also give off vapor, ice among others. The molecules of both the liquid and 118 MOISTURE: VAPOR, CLOUD. 1 19 its vapor are in constant motion. The average velocities in the vapor are greater than in the liquid ; but the largest velocities in the liquid may be greater than the smallest in the vapor. If any of the molecules at the surface of the liquid have these abnormally large velocities, and if they are moving from the liquid, they will escape from those forces which retain in the liquid the molecules of lesser velocity, and will fly off from the liquid as vapor. The number of molecules which pass from the liquid to vapor depends mainly on the temperature, but partly also on the amount of vapor already existing above the liquid. In the process of evaporation there is a disappearance of a quantity of heat, and a depression of temperature occurs. This heat which dis- appears is called the latent heat of vaporization. Its amount varies according to the temperature at which evaporation takes place. For a further account of this, the reader must consult a text-book of physics. Condensation. If the molecules of vapor have their velocities diminished, by cooling, down to the velocities of the liquid, then condensation occurs, and the vapor is con- verted into a liquid ; or the molecules of vapor striking the liquid may become entangled among the molecules of the liquid, and thus become part of the liquid. The num- ber of molecules which pass from the vapor to the liquid depends on the density of the vapor as well as on the temperature. Saturation. If the temperature of the vapor and liquid is the same, evaporation will be in excess of condensation until the density of the vapor has become so great that as many molecules are condensed as are evaporated, then the vapor has attained its maximum density (for that tem- perature), and is said to be saturated. Evaporation and condensation are still going on, however, after saturation ; but they just balance each other. I2O ELEMENTARY METEOROLOGY. Conditions of the Atmosphere with Regard to Moisture. Four conditions of the atmosphere with respect to moisture are recognized, which may be termed stages, since they are brought about by a progressive succession of changes to which the moist air is subjected. There is first the dry stage, in which the air contains vapor; but it is not saturated, and consequently no con- densation or precipitation takes place. The amount of unsaturated or superheated vapor in the dry stage for the free air (during its changes) may be assumed to remain constant; that is, no vapor is added or abstracted. In this dry stage, if the air mixture has been cooled, then by the addition of heat, or by adiabatic compression, the original condition can be resumed, as no water has been lost. The rain stage is reached when the unsaturated moist air of the dry stage is cooled by the abstraction of heat, or subjected to adiabatic expansion until the condition of saturation is reached and passed, and water drops form. The water drops fall away from the air as rain, unless, by means of a strong upward wind, they are' carried along with the air in which the condensation occurred. The absolute amount of moisture in the air is decreased by the amount which falls away as rain, and the remaining vapor is in a saturated condition. In the rain stage the changes which occur are seldom reversible, that is, capable of returning the air to its original condition, as has been mentioned for the dry stage, because usually some of the moisture be- comes lost as raindrops. In the rain stage there are two extreme con- ditions which may exist : either all of the water drops which have been rormed may fall away as rain, or the water drops may be carried along with the air from which they were condensed, by means of the support offered by sufficiently powerful vertical air currents. In the former case the temperature will fall more rapidly than in the latter as the change goes on. MOISTURE: VAPOR, CLOUD. 121 The hail stage is reached when the temperature of the cooling moist air reaches the freezing point of water. The air mixture in this stage is made up of dry air, satu- rated vapor, any water drops which are present, and frozen water drops which exist as ice or hail. When the tempera- ture is just at the freezing point, water drops and hail pel- lets may exist side by side. At the beginning of the hail stage we had the rain stage, in which the air mixture contained water in the form of raindrops and saturated vapor. The amount of hail will depend on the amount of these water drops when the hail stage was entered upon, and also on the amount of change in the volume of the mixture by expansion. It must be carefully noted that the hail stage cannot be entered upon until there is an appreciable amount of water present as raindrops. The snow stage is reached by the gradual cooling of the ascending air to a temperature below freezing of water, but without the formation of raindrops ; or when the pre- cipitation of rain and hail is so slight that the latter stage, at least, is practically skipped over. In the snow stage the air mixture consists of dry air, saturated vapor, and snowflakes, and the mass is still less than for the hail stage. Since the amount of vapor which may be present depends on the temperature, then, the lower the temperature, the less the possible snow- fall ; which accounts for the fact of comparatively light snowfalls in extremely cold weather. By a reversion of the process by which these stages are reached that is, by a direct warming, or by adiabatic compression of the air mixture it is possible to carry it back into the dry stage again ; but the amount of vapor 122 ELEMENTARY METEOROLOGY. will be less than that it started with, by the amount of moisture which has fallen away in the form of rain, hail, or snow. ATMOSPHERIC MOISTURE AS VAPOR. The Amount of Moisture in the air existing as a vapor may vary from nothing up to saturation. These two extreme conditions serve as terminal points for a scale of measurement of the degree of moisture in the air. The temperature is the most important factor in determining the amount of moisture, and in general the possible maxi- mum amount of moisture diminishes with decrease of tem- perature ; or, the colder the air, the less the possible amount of moisture. There are three terms constantly used in speaking of atmospheric moisture : absolute humidity, relative humid- ity, and the dew-point. Atmospheric Humidity. The humidity of the air is the amount of moisture which it contains ; that is, the amount which is mixed with the dry air to form the atmospheric air. The Absolute Humidity is the weight of this moisture. If we consider it in its watery condition, it is then ex- pressed in grains per cubic foot of air. If we consider it in its invisible gaseous condition, then we measure the pressure or tension of the vapor, and it is then expressed in fractions of an inch of barometric pressure. The Relative Humidity is the relation of the amount of moisture present to the amount necessary for saturation under the existing condition, and it is expressed in per- centage of the latter. When it is said that the relative , humidity is 50 %, this means that half as much moisture is present as would be necessary for the saturation of the MOISTURE; VAPOR, CLOUD. 123 vapor under the existing conditions of temperature and barometric pressure. The Dew-point. Since the amount of vapor which can exist in the air depends chiefly on the temperature of the air, then, if we make any moist air colder, we shall increase its relative humidity ; and if it is cooled far enough, the relative humidity will become 100 %, and saturation will OCCUH The dew-point is that temperature of the air at which its invisible moisture begins to condense into visible water drops. The following table shows the weight, in grains Troy, of saturated vapor per cubic foot at the annexed dew-point temperatures : Dew-point, F . . . . 1.7 4 2.Q 60 5,8 80 II.O 100 IQ.8 If a little ice is put into a tin cup of water, and the water is stirred so that all of the water may cool at the same time, and if the bulb of a thermometer is put into the water and the temperature is noted when the moisture first begins to be deposited on the outside of the cup (like the "sweating" of a pitcher of cold water on a warm day), then the thermometer reading gives the dew-point of the air for its existing conditions. It is a little difficult to detect the exact instant at which the condensation of the moisture on the outside of the cup begins, and so it is usual to note as accurately as possible the tempera- ture when this condensation begins, and then to warm the water slightly and note its temperature when the condensed moisture disappears ; and the average of the two temperatures will give quite closely the desired dew-point. Methods of Measuring Atmospheric Moisture. There are several methods of determining the amount of moisture in the atmosphere, but those in chief use by meteorologists are: (i) by determining the temperature of evaporation (by using the psychrometer) ; (2) by observing the proportional saturation of certain animal and vegetable substances (by using the hair hygrometer) ; (3) by observing the various aspects of the sky. 124 ELEMENTARY METEOROLOGY. Determination of Humidity by the Temperature of Evapo- ration. The psyclirometer is used in this method. This instrument consists of two similar mercurial thermometers placed side by side ; the one being in its unaltered condi- tion, and the other having a wet thin muslin covering for the bulb which contains the mercury. The muslin cloth is usually made wet by connec- tion with a wicking dipped into water. The evaporation of the water from the cloth cools the thermometer bulb to the tem- perature of evaporation; and the drier the air, the lower will the temperature of the wet bulb sink below that of the dry bulb. A psychrometer is shown in Fig. 38. More accurate results than for still air are obtained by making an air current pass over the thermometer bulb, which can be done most easily by whirling the thermometers. The Hair Hygrometer (Fig. 39) FIG. 38. PSYCHROMETER; WET AND j s sometimes USed for determin- DRY BULB THERMOMETERS. . ... ... ... ing the relative humidity of the atmospheric air. This instrument is based on the fact that hair expands in length with increase of moisture, and con- tracts with decrease of moisture. The amount of this ex- pansion or contraction is measured on a scale which shows the percentage of total expansion for saturation. Actual Observation of Atmospheric Humidity, The atmospheric relative humidity most directly affects our MOISTURE: VAPOR, CLOUD. 125 sensibilities, and it is generally expressed in per cent of saturation. For many questions of physical investigation concerning the atmosphere, however, it is necessary to know the absolute amount of water in the atmospheric air, expressed either in vapor pressure or in weight of water. The Daily March of the Relative Humidity is very pronounced. In general it decreases with the diurnal increase of temperature. There is an early morning maximum and an afternoon minimum. During the day- time, when the temperature increase is rapid, just before and about noon, the relative humidity falls very rapidly, but during the night it does not vary by a great amount. Where there is a large daily amplitude for the oscillation of temperature, there will also be found a great ampli- tude for the relative humid- ity ; and where the amplitude of the temperature is small, that of the relative humidity will likewise be small. The diurnal amplitude of the relative humidity is small near the ocean coasts, but increases towards the interior of the continent. FIG. 39. HAIR HYGROMETER. On the northwest coast of Europe the amplitude is 7% in winter (December), and 17% in summer (August) ; while at Nukuss, in cen- tral Asia, the amplitude is 26% in winter (December), and over 50% in summer (August). The relative humidity has the greatest amplitude on clear days, and the least on cloudy days. The amplitude is several times as great in 126 ELEMENTARY METEOROLOGY. clear weather as in rainy weather. There is a tendency for the rela- tive humidity to increase on the average for the 24 hours of cloudy days, and to decrease during clear days. The Annual March of the Relative Humidity is somewhat irregular. The time of maximum is in midwinter, and of minimum in early summer. The annual amplitude is least at the coast stations, and greatest in the interior of con- tinents. The amplitude for the extreme months on the coast of the ocean amounts to perhaps 5 %, and in the interior of the continent to over 30 %. The average annual relative humidity is greatest for the marine climate, and decreases towards the interior of the continents. Up to a certain altitude (that of the clouds), the annual amount increases with the elevation, but be- yond that point it decreases. But it varies greatly with altitude, since the temperature and amount of vapor, upon which relative humidity depends, vary so much with alti- tude. Local influences, such as mountains, the direction of the winds with regard to the neighboring drier or moister regions, and other causes, affect the relative humidity of a place. The annual march of the relative humidity at a few stations is shown in percentage by the following average monthly relative humidities : J- F. M. A. M. J- J- A. S. O. N. D. Year Key West, Fla. . . 80 77 70 69 7i 7i 70 72 75 76 78 79 74 St. Paul, Minn. . . 72 7 1 69 60 60 68 70 72 7 69 72 74 69 Salt Lake City, Utah 60 57 58 45 40 32 30 3 1 3i 42 52 60 44 The Geographical Distribution of the Relative Humidity is not easily shown on charts, on account of its variability with altitude. On the open ocean, however, it is quite MOISTURE: VAPOR, CLOUD. 1 2? regular, and on the average varies between 82% at the equator and 92% in high latitudes. On the continents, owing to the high temperatures, there is usually a small relative humidity in summer, and a large one in winter ; but for the absolute humidity we have just the reverse of this. The Daily March of the Absolute Humidity varies greatly for different places. There are, however, two pronounced types, the maritime and the continental. The maritime type has a maximum in the warmest part of the day, with an early morning minimum. The conti- nental type has two maxima, one just before noon, and the other late in the afternoon, with a secondary minimum between ; the primary or lowest minimum occurring in the early morning. i On the mountains the amplitude is less than on the plains. The secondary minimum in the continental type is due to the ascending air currents which occur towards the hottest part of the day ; the surface air being carried upward more rapidly than the moisture is supplied to it by the evaporation which takes place from the earth's surface. Over the water surface these ascending currents are not so strong as over the land (especially on dry plains), and therefore the air is not so likely to rise upward faster than the lower layers are supplied with moisture by evaporation. The absolute humidity has the greatest amplitude on clear days, and the least on cloudy days, when precipitation also occurs. There is a tendency for the absolute humidity to decrease during the 24 hours of clear days, and to increase during cloudy days. The Annual March of the Absolute Humidity is very similar in character to that of the temperature. The minimum is in the winter months, and the maximum in the midsummer months. The amplitude for the year is least on the coasts, and greatest in the interior of con- 128 ELEMENTARY METEOROLOGY. tinents ; and it is less in the tropics than in the temperate zones. The Average Amount of Vapor decreases with the alti- tude above sea level, and it decreases more rapidly than the decrease of air pressure (and density) with altitude. The relative decrease of the- amount of water vapor with the altitude, and also the decrease of the air density, are shown by the following table, in which the air density and water vapor at the earth's surface are each placed at 1 .00 : ALTITUDE. WATER VAPOR. AIR DENSITY. Feet. O 1. 00 1. 00 13,000 + .24 .61 30,000- .04 32 It is seen how much more rapidly the water vapor decreases than the air density. Half of the water vapor lies below the altitude of about 6,500 feet, while half of the air lies below the altitude of about 18,000 feet. The low altitude of the greater portion of the moisture in the atmosphere accounts in a great measure for the powerful influence which even low mountains exert in the distribution of this moisture over the earth's surface. The Geographical Distribution of the Water Vapor over the different regions of the earth follows quite closely that of the temperature ; in general the greater the temperature, the greater the amount of water vapor. With increase of latitude the amount of vapor decreases. In January the region of greatest amount of vapor is the region extending from the equator to about latitude 20 south. In this region the vapor pressure is over 0.8 of an inch, but in equatorial Africa it reaches i inch, and on the northern coast of Aus- tralia 0.95 of an inch. In western Europe it varies from 0.2 to 0.4 of MOISTURE: VAPOR, CLOUD. 1 29 an inch, in eastern Europe from o.i to 0.2 of an inch, and in the cold regions of northeastern Asia (and likewise in north-central North Amer- ica) it decreases to 0.05 of an inch or lower. In July the region of highest vapor pressure lies to the north of the equator, where in India it reaches I inch. In the extreme northern part of the northern hemisphere the vapor pressure is reduced to 0.2 of an inch or less ; while in the southern hemisphere the pressure of 0.2 of an inch is reached even at latitude 40 south. It is thus seen that there is a movement of the maximum region backwards and forwards across the equator from one hemisphere to the other, following the sun. ATMOSPHERIC MOISTURE AS CLOUD AND FOG. Fog and Cloud Formation. When air is cooled just below the dew-point, then fog or cloud occurs, and, as may be seen by observing the sky, in a great variety of forms. While in a general way the process of cloud formation is understood, yet the exact manner of the building of clouds of the various shapes and appearances which occur has not been entirely studied out. Clouds are at present divided into classes, according to the forms which they pre- sent to the eye, and not according to their processes of formation. Composition of Clouds. Since the clouds contain more particles of dust than the surrounding clear air, it is con- cluded that the air within clouds has been drawn up from close to the earth's surface, where these dust particles abound. The dust particles within clouds usually number only a few thousand per cubic centimeter; but the water particles may number as high as 50,000 in the same space. WALDO METEOR. 8 I3O ELEMENTARY METEOROLOGY. The density of a cloud depends directly on the number of water particles present. When cloud forms, precipitation begins at once ; but the water particles are very small, and evaporation takes place when they fall into drier air below. The distance which they fall depends on their size and the dryness of the air beneath. The denser the cloud, the larger are the raindrops, and the faster they fall. Nomenclature of Cloud Forms. There are four distinct classes of clouds, according to their forms, cirrus, stratiis, cumulus, and nimbus. Cirrus clouds are those which are seen in striated forms grouped high up in the sky. Stratus clouds are those which present a stratified or bank-like form. They may be high or low lying clouds. Cumulus clouds are more or less isolated, have rounded tops, and are found at but moderate altitudes above the ground. Nimbus clouds are those from which rain or snow descends. Combinations of these forms add several varieties with recognized names. The principal ones are as follows : cir- rus, cirro-stratus, alto-stratus (strato-cirrus), stratus, cirro- cumulus, alto-cumulus (cumulo-cirrus), strato-cumulus, cumulus, cumulo-nimbus, nimbus. These various forms are represented in the accompanying illustration (Fig. 40). i. Cirrus clouds (No. i) are feathery in form and delicately fibered, usually of a white color, and well outlined against the sky background They lie arranged in a variety of fantastic forms. Nearly parallel groups of these clouds are sometimes seen stretching across the heavens in converging, meridian-like bands. Such bands as these are also some- times formed of cirro-stratus and cirro-cumulus clouds. The cirrus clouds have perhaps an average altitude of between 5 and 6 miles. MOISTURE: VAPOR, CLOUD. . 131 2. Cirro-stratus clouds (No. 2), consisting of fine, white, veil-like clouds, and the alto-stratus clouds (No. 5), thick, veil-like clouds of grayish or bluish color, look in some respects quite alike in form, but their altitudes are very different ; the altitude of the cirro-stratus being about 5> miles on the average, and that of the alto-stratus being only about half as great, perhaps about 3 miles. The cirro-stratus clouds usually precede bad weather, and they gradually give place to alto- stratus clouds. The cirro-stratus are the clouds which give rise to the phenomenon of rings around the sun and moon. The cirro-stratus formation is sometimes very widely diffused over the heavens, and then becomes a cirrus vapor ; but sometimes a distinct but intricate-fibered structure is visible, which may be called cirrus-felt. 3. Cirro-cumulus clouds (No. 3) are very small white, and alto- cumulus (No. 4) large whitish-gray, balls or fleecy clumps grouped in herds. Sometimes these fleecy clumps are arranged in rows, extending in one or two directions. The cirro-cumulus lie much the higher, being at an altitude of 3^ to 4^ miles, while the alto-cumulus are only about 2}^ miles up. 4. Strato-cumulus (No. 6) and nimbus (No. 7) clouds belong to the lower air layers, and form clumps or layers at an altitude of from ^ of a mile to iX miles above the ground. Strato-cumulus clouds occur in dry weather, and appear very frequently in winter, when they more or less cover the heavens, but with patches of blue sky between the clouds. The nimbus is the cloud of continued rain or snow. It has ragged edges, and above are always to be seen the alto-stratus clouds ; so that between the nimbus clouds a gray cover or background is visible, and not the blue sky such as is seen above the strato-curr.ulus. 5. Cumulus clouds (No. 8) are the thick, dense clouds with rounded, festoon-like tops and horizontal bases. The top usually comes to a point or peak higher than the rest of the cloud. These clouds are built up by ascending currents within ; and when these air currents cease, the clouds disappear gradually. Such clouds are characteristic of the summer sky, especially over the land. Sometimes over tropical islands the vertical air currents which arise cause more or less permanent cumu- lus clouds to overhang the islands. Cumulo-nimbus clouds (No. 9) are the thunder and shower clouds which roll up in such an imposing manner, and present a majestic ap- pearance of mountain-like character. The tops are of a light, fluffy ap- pearance, while the bases are of the dense nimbus character, from whose (132) FIG. 40. PRINCIPAL F s OF CLOUDS (AFTER KOPPEN). (133) 134 ELEMENTARY METEOROLOGY. center showers of local rains and hail descend. The upper edge is festooned like the cumulus, and towering cumulus peaks are formed. The edges are sometimes bordered by small, cirrus-like clouds, or these last are detached from the main cloud. 6. Stratus clouds (No. 10) consist of an elevated fog. Fog which lies at the ground is designated simply fog; but when it is at an alti- tude of, say, a thousand or more feet above the ground, it is called stratus cloud. Such is dry-weather fog. The broken, tattered clouds ^which appear at low altitudes in wei weather are called fracto-nimbus . The tattered clouds which occur with the true cumulus clouds are called fracto-cumulus . Processes of Cloud Formation. The following processes are those by which fog and cloud are produced : 1. Direct cooling of the moist air through contact with colder bodies or through loss of heat by radiation. 2. Adiabatic cooling of ascending moist air. 3. Mixture of moist air of different temperatures and humidities. These are mentioned in the order of their effectiveness for the production of cloud. Dissipation of Clouds. Conversely, fog or cloud may be dissipated in the following manner : 1. Direct warming of the clouded air through radiation or contact with warm bodies. 2. Adiabatic heating of descending clouded air. 3. Mixture of the clouded air with other air masses of proper temperature and humidity. Ground Fog. The condensation produced by direct con- tact with a cold body and by loss of heat through outward radiation is that of ground fog, which extends upwards to a moderate height above the ground. The outward radia- tion from the ground on clear nights cools the ground very rapidly ; and, as soon as the dew-point is passed, condensa- tion begins in the lowest air layers, and fog forms. The MOISTURE: VAPOR, CLOUD. 135 upper surface of this fog radiates heat, and cools the air layers just above it, so that condensation ensues in these layers. Thus the process of fog building proceeds, and the fog layer becomes thicker and thicker. When the solar radiation begins to make itself felt, the reverse takes place ; and the upper layers are dissipated first by the warming of their upper surface ; then the next layer is dissipated ; and so on until the ground is reached, when it too becomes warm, condensation ceases, and the fog entirely disappears. The reason that no excessive precipitation occurs in this process is that the formation of the fog cloud above the ground prevents the fur- ther excessive cooling of the latter by continued radiation. The fog growth on the upper limit of elevated clouds may also occur in the manner just described ; but, in order to produce condensation by direct radiation in the upper air layers, there must exist a cloudiness formed by some other process, or by means of a framework of such impurities as smoke particles. The Formation of Clouds through Adiabatic Expansion, and their dissipation through compression, occur where there exist ascending and descending air currents. The huge summer clouds with rounded tops and horizontal bases, the so-called thunderclouds, and the usual rain clouds, are formed by this process, when ascending air currents are present, and the air thus expands and cools; and their dissipation occurs for descending movements, whereby the air is compressed again and made warmer. The Formation of Cloud by Mixture of Air of Different Temperatures and Humidities is a much more complex matter than that just described ; and for cloud formation it is of great importance, but is comparatively unimportant for causing the precipitation of moisture to the ground. Condensation takes place more rapidly when a current 136 ELEMENTARY METEOROLOGY. of cool moist air penetrates a large mass of warm moist air, than when a current of warm moist air enters a mass of cool air. Condensation does not always proceed gradually, and it may not occur until the final stages of cooling by mixing are reached, when it will take place all at once. Likewise in the reverse process the dissipa- tion of the cloud may be retarded until the condition is such that the whole of the cloud will disappear quite suddenly . In the case of the breath leaving the nostrils and penetrating the cold air, the condensa- tion does not take place until actual mixture is effected ; and when this mixing process distributes the breath further through a larger mass of the drier colder air, the fog is dissipated. It has been shown, that, in the mixing of saturated cool air with larger quantities of saturated warm air, the warming of the former takes place at first rapidly, and then more gradually ; but when the cold air is greatest in quantity, then the cooling of the warm air becomes more rapid as the process con- tinues. The maximum amount of condensation will occur when the amount of cool air is in excess. By this process of mixing, the following kinds of clouds and fog are produced : 1 . The fog which arises over relatively moist warm surfaces, where cooler air is also present. An example of this is the surface ocean fog which so frequently arises during the cold season. 2. The clouds which are formed at the common boundary of two air currents (the air in which may differ as regards both temperature and moisture) moving with different velocities, by which means a regular succession of clouds is formed as a consequence of a wave motion which may exist ; and so great is the amplitude of these waves, that the adiabatic condensation arising during the upward surges must make itself visible. 3. The layers of stratus clouds which are formed at the juncture surface of two such air currents having different velocities, and which have at first a disconnected form, but which afterwards become joined. 4. The banner-like clouds which form (and dissolve) on some mountain peaks, and in mountain passes, when the contour is such that warm or cold masses of air are penetrated by currents of air having other temperatures. MOISTURE: VAPOR, CLOUD. 137 5. The loose and tattered clouds which are so noticeable in strong winds, and especially in thunderstorms, and which are constantly undergoing changes of form. It is very probable that fogs and clouds of various forms are also produced by combinations of any two or all three of the causes of condensation ; but we do not as yet know their method of forming well enough to specifically point them out. It may also be remarked, that, without a knowledge of the mode of formation of clouds, it is a very easy matter to misinterpret their movements as we observe them. Amount of Cloud. The degree of cloudiness is es- timated on a scale of from o when there are no clouds visible, to 10 for the whole sky overcast with clouds. It is therefore merely estimated how many tenths of the sky are covered with cloud. Almost all of such observations are made by the unaided eye. These estimations are re- duced to percentage of the total visible sky by multiplying them by 10. While there has been some slight attempt at estimating the amount of the various kinds of clouds, yet in nearly all cases the degree of cloudiness is estimated irrespective of the kind, density, or height of the clouds. Thus, if the degree of cloudiness is recorded as 7 (on a scale of o to 10), and cirrus and stratus clouds are present, no attempt is made to state what proportion of this 7 refers to the cirrus and what to the stratus clouds. The Daily March of Degree of Cloudiness is somewhat difficult to determine, since the amplitudes are small, and vary only about 10 % of the whole surface of the sky. In the morning hours fog and stratus clouds are the most frequent; but towards midday, and in the afternoon, the cumulus are in excess. In general there is a principal maximum of cloud about noon or a little after, and a prin- cipal minimum in the night. This varies, however, not only with change of locality, but also at different seasons of the year. 13* ELEMENTARY METEOROLOGY. The Annual March of Degree of Cloudiness shows in gen- eral a maximum amount of cloud in the late fall or early winter, and a minimum in the spring, but sometimes re- tarded until summer for relatively low lands ; but in the high lands (the Alps, for instance) the maximum occurs in the summer, and the minimum in the winter. The maximum cloudiness occurs mostly in early winter, because then the average temperature is decreasing rapidly, which causes an increase in the relative humidity. The minimum occurs mostly in the spring, because then the average temperatures are increasing rapidly, and this causes the relative humidity to decrease. The average monthly cloudiness depends, however, so much on the direction of the winds with regard to the supply of moisture, the relative humidity, and other conditions, that a great variety of phases are to be met with. The amount of cloud in Bombay varies from 18% in December to 91 % in July. In Cairo, Egypt, it varies from 33 % in December to 6 % in June. In Spitzbergen it varies from 51 % in December to 87% in Sep- tember. On the northern coast of Norway it varies from 68 % in January to 60 % in June. The following table shows the average monthly amount of cloud, and consequently the annual march of cloudiness, at a few places in the United States (in percentage of total cloudiness) : J- F. M. A. M. J- J- A. S. 0. N. D. Year Key West, Fla. . . . 42 35 29 3i 43 48 49 49 52 46 38 40 42 St. Paul, Minn. . . . 49 48 5i Si 52 49 42 45 48 5i 58 5i 50 Salt Lake City, Utah . 54 52 53 52 46 3i 3 3 26 40 47 56 43 Prescott, Ariz. . . . 28 26 29 26 17 13 32 33 *7 x6 17 2 5 23 The Variation of Average Cloudiness with Latitude is shown by the following table, in which the cloudiness is given in percentages of the whole sky visible at one time : MOISTURE: VAPOR, CLOUD. AVERAGE CLOUDINESS. 139 LATITUDE. NORTHERN HEMISPHERE. SOUTHERN HEMISPHEKK. JANUARY . JULY. JANUARY. JULY. 70 55 59 ^ - 65 56 61 60 62 62 70 70 55 59 61 70 6 7 50 57 57 70 62 45 52 50 60 * 54 40 50 44 53 56 35 46 41 48 52 30 44 42 47 47 25 37 45 48 46 20 37 50 & 43 15 40 53 53 48 10 45 59 58 5i 5 5 59 56 57 Equator o 5<> 58 - The following conclusions are drawn from charts made of the cloud distribution. The degree of cloudiness is arranged in zones parallel to the equator, The zones of minimum cloudiness are reached at 2O-25 north latitude, and about 25-3O south latitude, and the amounts are least in the northern hemisphere. The region of maximum cloudiness is reached at about latitude 60 in the northern hemisphere, and then there is a decrease toward the north pole; but in the southern hemisphere there is an increase up to 60 south latitude, which is as far as observations extend. These zones present a well-defined tendency to follow the sun in its variations in latitude for the year, lying at a higher latitude in summer than in winter. Fogs. Fog at the earth's surface shields the ground from, the solar rays in proportion to its density ; and at its upper limit it must acquire 140 ELEMENTARY METEOROLOGY. a temperature something like that which would occur at the earth's unclouded surface. The thickness of a fog is shown by the length of time necessary for it to be dissipated by the solar rays. Above the ground fog the air is clear and dry, and radiation takes place un- hindered. Thus it happens that bright clear weather is experienced after the dissipation of a fog, which at first seems to be a forerunner of rain. Ocean Fogs. When warm moist air is carried by the atmospheric circulation into a cooler region, the moisture is condensed, and fog forms. Such cases occur with marked frequency over the North Atlantic Ocean in the region to the south and east of Newfoundland, and con- sequently directly in the track of the transatlantic ocean steamers. In storms, when the wind blows from the east or southeast, the warm moist air from the Gulf Stream is blown over the colder waters of the Arctic current or over the ice floes which this current brings down from the north, and fog condensation takes place. The greatest frequency of fog occurs in those latitudes at about the 55th meridian .(west of Greenwich). The number of days with fog for each month in that region is as follows: J. F. M. A. M. J. J. A. S. O. N. D. Year. 5 ii 12 15 18 17 23 22 15 13 10 4 165 The sea air contains more moisture in the summer time than in win- ter, which, together with the great differences in temperature due to the large quantity of field ice and icebergs off Newfoundland, causes the relatively great frequency of fogs during the warm season. Maximum Amount of Water in the Air. The largest possible amount of moisture which can exist at any one time in the air (without the supporting power of upward air cur- rents) between the earth's surface and certain altitudes is best shown by the number of inches' depth of rain which it would make if all of the moisture were condensed, and fell as rain. In the most favorable case there would not be MOISTURE: VAPOR, CLOUD. 141 two inches of water between the earth's surface and an altitude of three miles above it. DEPTH OF WATER FOR THE FOLLOWING DEW POINTS AT HEIGHT OF COLUMN OF AIR THE EARTH'S SURFACE. 80 F. >JO F. 60 F. 50^. Feet. Inches. Inches. Inches. Inches. 6,OOO i-3 1.0 0.7 0.5 12,000 2.1 i-S I.I 0.8 l8,000 2. 5 1.8 i-3 0.9 24,000 2.7 2.0 1.4 I Jo 30,000 2.8 2.1 i<5 I.I The above table shows the amount of water which may exist in the air below certain altitudes for various temperatures. CHAPTER VI. MOISTURE : PRECIPITATION. Precipitation occurs when moist air is cooled below the dew-point, the moisture taking the form of rain, snow, hail, dew, or frost, depending on the conditions under which the condensation takes place and is maintained. The greater part of the precipitation occurs in the form of rain in the regions of the earth inhabited by man, and we shall therefore treat it more fully than the other forms of pre- cipitation, which will receive but brief mention. Precipitation is an intermittent phenomenon of a common occurrence, and it is therefore of importance to observe its duration and frequency as well as its amount. No other important meteorological element ex- hibits the variability of rainfall, and it is next to temperature in the importance of its effects on life on the earth. Rain. When clouds form over a region in which the air is nearly or quite saturated with moisture, the globules of water which form the cloud unite, and descend through the stratum of moist air underneath, and fall as rain when the temperature is above the freezing point. The rain- drops vary in size from those of the fine mist which we encounter in fog, to the large isolated raindrops which fall from clouds at considerable altitudes. The production of rain has been considered during the past century to be mainly the result of the mechanical mixture of air masses of differ- ent temperatures and humidities. Objections to this theory were made more than half a century ago, but they did not receive proper con- 142 MOISTyRE: PRECIPITATION. 143 sideration. It was not until about 1870, however, that the opposite extreme view was taken, that the mechanical mixture of the air would never alone cause the condensation of vapor and formation of rain, and that this last was entirely due to adiabatic changes. Each of these theories had some truth in it, but neither one of them was satisfactory in itself. A few years later, 1874, the proper relative importance of the two theories was pointed out ; and it was shown, that while the mechan- ical mixture of air might produce some slight rainfall, yet the amount would be small compared with that due to adiabatic causes. Causes of Condensation and Precipitation. As we now view the matter, there are three causes of condensation of water vapor in the atmosphere ; and precipitation may be regarded as but an ultra condition of condensation. 1. The Cooling of the Air by Contact with Cold Bodies or by Radiation is the most powerful and easiest carried on of the rain-forming processes, and a very slight cooling by this means will produce rainfall Very excessive rain- falls in the mountainous regions are due to this process, especially when assisted by the upward movement of air in vertical currents. If we have a mass of warm saturated air at a temperature of 68 F. and 27.56 inches pressure, it requires this air to be cooled by radiation or contact with a cold body, by 1.5 F., that is to 66.5 F., in order to produce 5 grains of precipitation per pound of air. 2. The Cooling of the Moist Air by the Adiabatic Pro- cess. Dry air cools by adiabatic expansion at the rate of about i F. for 183 feet, or 0.55 F, per 100 feet of ascent. When the air contains moisture which has not yet reached the point of saturation, this rate of decrease is nearly the same as for dry air. Where the air is saturated, the cool- ing takes place at a less rapid rate, If the air is saturated, and at 27.56 inches air pressure and a temperature of 68 F., it is necessary to move it upward to a height 144 ELEMENTARY METEOROLOGY. of 1,017 ^et, by which it will be cooled from 68 D F, to 65.1 F., in order to obtain from it the 5 grains of precipitation from a pound of the air. Probably most of the irregular rainfalls which occur are due to this process. 3. The Mechanical Mixture of the Air having different temperatures and humidities is brought about by motions of the air in various directions. The air pressure varies with the altitude ; but since the components of this mixture have a common altitude, then they must have practi- cally the same air pressure. These air masses which mingle together may have, however, a variety of relations depending on the humidity and temperature. Perhaps the most interesting question connected with this is that concerning the amount of humidity necessary in one or both of the component air masses at given temperatures in order to produce condensation or rainfall. Various such combinations may occur. The relative humidity of the mixture must exceed that of at least one of the component air masses before condensation can occur. When one component is warm saturated air, and is mixed with cooler air, the latter may have quite a high degree of dryness, and still condensation may occur. In fact, it is found that when saturated warm air is mingled with unsaturated cold air, condensation occurs much more easily than when the latter is saturated. The determination of the conditions under which two air masses of given temperature and humidity must be mixed in order to produce the greatest amount of precipitation has been investigated, but the method of computation is too complex to explain here. The extreme case for the conditions cited under the first and second causes of condensation will show the amount of precipitation by this method. If we suppose two air masses to have an air pressure of 27.5 inches, and both to be saturated and have temperatures of 32 F. and 68 F. respectively, then the maxi- mum amount of precipitation which could fall from the mixture of the two air masses at 52 F. would be 5 grains per pound of the mixture. In many cases where the one component air mass is very dry and cold, it is impossible to obtain any precipitation by its mixture with moist warm air. This mixing of the dif- MOISTURE: PRECIPITATION. 145 ferent air masses, while not producing much precipitation, plays a very important part in the formation of clouds, as we have seen on p. 136. The Amount of Rainfall is measured by the depth of fallen water, expressed in inches or millimeters. The rain water is caught in a catch basin called a rain gauge ; and when the depth of the water is measured directly, this must take place in a symmetrical vessel having the same water surface area as the receiving mouth of the gauge* Rain fall data inchide snowfall. "Front Vieiv Vertical Section Rain gauges are usually ver- tically placed sheet-metal hollow cylinders of from 5 to 8 inches in diameter. Fig. 41 shows a rain gauge with a funnel-shaped mouth, A, and an inside receiver, C, of less diameter than the outer cylinder. Horizontal Sec. E.F, 516171319202122232} tncftaS Scale FIG 41. RAIN GAUGE. The Diurnal Change of the Rainfall is very marked, as to both the amount and the frequency of occurrence of rainfall. For both of them there is a chief maximum in the warmest part of the afternoon at two or three o'clock or later; and a principal minimum in the morning, two or three hours before noon. A secondary maximum occurs frequently at about two or three o'clock in the morn- ing, and a secondary minimum at about midnight. In the higher latitudes the maximum amount occurs at a later hour in summer than in winter ; but this is not so apt to be the case for the frequency of rainfalL The amount of rain at the hours of maximum is usually from two to three times that at the hour of minimum ; but 146 ELEMENTARY METEOROLOGY. in some extreme cases in the tropics, in certain months, nearly all the rain falls in the afternoon, when the amount may be nearly twenty times that at the morning minimum. The amplitude of oscillation is greater in summer than in winter. The Maximum Amount of Rainfall within a Day is in some instances enormous. On one occasion in Japan 29.5 inches of rain fell in 24 hours, and in India 39.5 inches fell in one day. This is as much as would fall in a favorably situated region in a cold temperate climate in a year. Monthly and Annual Rainfall. The annual march of rainfall is so variable for different places, that many dif- ferent regions would have to be mentioned separately in order to give a true representation of it in all its variety of phases. The direction of the wind, whether it blows from a region of moist air, such as that over the water, or from a region of dry air, makes an enormous difference in the amounts of rainfall for the various months. And when to this is added the influence of elevated land or mountain regions, it becomes impossible to state general laws gov- erning the annual distribution of rainfall, depending on differences of latitude and longitude merely. Such is the effect of mountain ranges especially, that two neighbor- ing places only a few miles apart may have totally dif- ferent rainfall conditions. When the wind blows steadily from the ocean (which mainly supplies the air with mois- ture) towards a not distant inland high range of mountains, and the mountains lie across the path of the wind at right angles to it, we have the greatest differences in rainfall within narrow limits. On the windward side the rainfall is copious, and on the leeward side there may exist a desert where the rain seldom falls. Variation of Rainfall towards the Interior of a Continent. If any law may be expressed concerning the conti- MOISTURE : PRECIPITATION. 147 nental influence on rainfall, it is this: that the rainfall decreases, both in quantity and frequency, with the dis- tance inland from the sea or water surface in general. But, as before stated, a range of mountains may inter- vene, and utterly change the natural tendency to follow this law. Variation of Rainfall with Latitude. The total amount of precipitation in the year is, on the average, greatest in equatorial regions, and decreases towards the poles. It is greatest at a distance of a few degrees from the equator, and there is a slight decrease towards the equator. The variation of the average annual amounts of precipitation with latitude is roughly shown by the following table, in which this average is given in inches for regions between each ten degrees of latitude : AVERAGE ANNUAL RAINFALL (in inches). 90 80 (north) . . 4 8070 " 9 70 60 " . 23 6O 5O o . 32 5040 " .o 36 40 30 (north) . . 28 3020 " . . 48 20 10 " o . 92 10 o " . 90 o 10 (south) . . 128 10 20 (south) . . 62 2030 " . . 42 30 40 " . 30 4050 " o c 44 5060 " . . 40 The Variation of Rainfall with Altitude is such that from sea level up to an altitude of about 4,000 feet there is an in- crease in the amount of rainfall ; while at greater altitudes it again decreases. The increase in the rainfall from low lands up to the altitude of maximum amount of rain is from two to four fold the amount in the low regions. The influence of mountains makes itself felt by the increase of the rainfall to a considerable distance from them on the plains below ; but it is probable that this increase occurs only on the windward side of the moun- tains, and that the condensation is due to the adiabatic cooling of the air, as it rises at some distance from the mountain in order to cross it 148 ELEMENTARY METEOROLOGY. The effect of increase of rainfall with altitude is very curious in some dry regions where at low altitudes only a small amount of precipitation takes place ; and at higher latitudes on the mountain sides there will be such an increase of rainfall as to form belts of land where enough rain is received to cause a considerable or even luxuriant growth of vegetation in an otherwise arid region. Above this watered region there may be such a decrease in the rainfall as to make it impossible for plants to grow. The Geographical Distribution of Annual Rainfall. It has been estimated that about 6% of the land surface has an average annual rainfall of over 75 inches; 16% has from 50 to 75 inches; about 25% has from 25 to 50 inches; over 30% has from 10 to 25 inches; and over 20% has less than 10 inches. The results of rainfall observations on the ocean are unsatisfactory, owing to the lack of perma- nent places of observation. The distribution of rainfall (precipitation) over the land surface is well shown on the accompanying chart (Fig. 42). The darkest shading shows where the rainfall is over 75 inches. The greatest rainfall far exceeds this amount, however, for a number of locali- ties. Among the Chassia Hills in India the average rainfall is over 470 inches, and in some other localities in the tropics the amount reaches 190 inches. On the tropical islands, and especially those containing elevated land, the rainfall is usually excessive ; also on the portions on the west coast of Europe where there are high ranges of hills, on the west coast of North and South America beyond latitude 40, and on the eastern coasts of Newfoundland and Japan, the rainfall is tropical in amount. Copious rainfall, that is, between 50 and 75 inches, is shown by the next lighter shading. This is the rainfall for the greater part of Central Africa and South America, regions of smaller extent in southeastern Asia and in southeastern and northwestern North America, northwest- ern Europe, portions of central Europe, and the eastern coasts of Aus- tralia and central Asia, and the higher eastern coast of North America. The region of moderate rainfall, that between 25 and 50 inches, shown by still lighter shading, covers most of the eastern half of North (149) 150 ELEMENTARY METEOROLOGY. America, central Europe, southeastern Asia, and south-central Africa, South America, and Australia. Northern Africa forms a transition zone to regions of Jess moisture. The region of light rainfall is that having between 10 and 25 inches of rain. This covers an enormous territory in the northern hemisphere, the most of the western half and much of northern North America, northern Asia, and Russia ; and it forms broad transition zones to the arid regions to be next mentioned. The arid regions, those in which the rainfall is below 10 inches, are found in north-central and southwestern North America ; south-central and west-central South America ; southwestern and almost the whole of northern Africa ; southwestern, central, and northeastern Asia ; and central and western Australia. The general seasonal rainfall distribution over the globe is shown on the accompanying chart (Fig. 43). Within each of these broadly characterized regions there are minor subdivisions having special fea- tures, but presenting great complexity of detail. These have therefore been omitted on the chart. The major subdivisions are as follows : 1. (Light Pink) Region of Tropical Rainfall: Principal dry season in winter and spring; maximum rainfall in summer. 2. (Dark Blue) Region of Subtropical Rainfall : Maximum in win- ter; summer rainless. 1.2. (Dark Purple) Transition Region between i and 2 : Rain win- ter and summer. 3. (Medium Blue) Region of Maximum Rainfall in winter ; some- what less rain in summer. 4. (Dark Pink) Region of Minimum Rainfall in late summer, with considerable rainfall at other seasons. 5. (Light Purple) Region of Moderate Rain during all the months. (Snow in winter.) 6. (Light Blue) Wet Region : Copious Rains during the whole year, but most rain in winter. 7. (White) Dry Region : All months of the year deficient in rainfall. Characteristics of Rainfall in the Distinctive Rain Regions of the Earth. The rainfall areas have been divided into the subequatorial or tropical, subtropical, and temperate re- gions. The tropical region lies on each side of the equator, between the tropics. The subtropical region extends mostly 152 ELEMENTARY METEOROLOGY. from about the tropics poleward, and reaches as far as the 4Oth parallel in many cases. The temperate region lies to the poleward of the tropical region. The Tropical Rainfall occurs mostly in summer, and is partly the accompaniment of local storms, and due to the vertical air currents which arise especially at the time of greatest midday heat in summer, and is partly due to the steady winds from the east bearing the moisture-laden sea air on to the mountains of continents and islands. Also, as in India, the monsoon winds bring a rainy season when they blow from the ocean landwards ; but it requires a range of mountains to produce summer rains in the path of these steady air currents. There is a shifting of the tropical or subequatorial rain belt with the shifting of the sun from one hemisphere* to the other, because with this there is also a shifting of the equatorial wind system. In the region of calms (the dol- drums) occur mainly the rains due to the local upward currents ; and the shifting of this region follows the annual course of the sun in the heavens. In the region over which the doldrums pass twice in the year there are two rainy seasons, with two dry seasons between. The excessive rainfalls of the eastern coasts and moun- tain slopes near the equator are due to the steady winds carrying the moisture from the oceans to the land ; but where the excessive rainfalls occur on the western coasts, they are due to the action of vertical air currents in local storms, because, when the winds have passed over the continents from the east, they have been deprived of much of their moisture. The amount of water usually existing in the air as vapor is very great in the equa- torial region, owing to the high temperatures prevailing there. MOISTURE: PRECIPITATION. 153 Long-continued series of rainfall observations have been made at but few places in the tropical rain belt, outside of India, but the follow-' ing table shows the average monthly and annual rainfall at some selected stations. These show the rainfall conditions on both the eastern and western coasts of the continents. TROPICAL OR SUBEQUATORIAL RAINFALL (in inches). LATI- T. F. M. A. M. T. T. A. R. 0. N. D. Y n AFRICA. TUDE Loando (west coast) . 9 S. 2.4 i.i i-3 3-3 o-3 o.o O.O 0.0 O.I 0.2 2 -5 1.2 12.4 Gaboon " 7-5 9.0 15.0 8.7 8.9 0.9 0.5 1.2 8.1 19.4 19.1 7-5 105.8 St. Louis " . . i6N. 0.2 0.5 0.0 0.0 0.2 0.4 2.6 8.0 3-7 o-5 0.0 0.0 16.1 Port Louis (east coast) 2S. 5-7 n.8 5-2 3-i 2.1 i-5 0.9 i-5 0.4 0.7 i-7 3-7 38.3 TROPICAL AMERICA. Rio Janeiro (east coast) 23 S. 5-4 4-7 5-9 3-4 4 .8 1-5 i-3 2.8 3-3 3-9 5-7 5-2 47-9 Georgetown " 7 N. 6.9 5-8 7-3 7-3 I4.I i3-9 II.O 7-4 2.6 2-5 5* 10.8 95.1 Havana 23 N. 3-3 1.6 i.S 3-2 4.1 5-7 4-9 4 .8| 6.0 6.7 2.2 2.2 46.3 Quito (west coast) . . o 3-2 6.4 4.4 6.6 5-i 2-5 1.4 1.8 1.8 4.4 6.2 2.8 46.6 ASIA, i N R 1 fi T 6 8 7.0 6 3 7.2 6.2 9.1 7 pl 8 a 10 8 22 N i 6 q.c 17 2 14 2 ofi 84.7 Calcutta 22 N. 0.4 1.0 i*3 2-3 5-6 n.8 13.0 i3-9 10 S-4 0.6 0.3 65.6 Cherrapungi (Assam) 0.6 2.6 9.0 29.6 50.0 IIO.O 120.5 78.957.1 136 1.8 o-3 474.0 It is seen that in most cases the excessive rainfall occurs during but a few months of the year. The Subtropical Rainfall is characterized by a winter rainy season and summer dryness. The total rainfall is relatively small, and is from three to ten times as much dur- ing the colder half year as during the warmer six months, In some of the regions (as, for instance, in Egypt and Pal- estine) there is scarcely any rainfall during the summer months. This region stretches up to nearly the 4Oth de- gree of latitude over the ocean, but is very variable on the continents, owing principally to the modifying circumstances of mountain ranges and winds. In the neighborhood of the Mediterranean Sea there is a decrease in rainfall from north to south, and from the west towards the east. 154 ELEMENTARY METEOROLOGY. The following table shows characteristics for this region, where the lands bordering on the Mediterranean Sea give the best examples of the type : RAINFALL IN SUBTROPICAL REGION (in inches). T F M A M T T A s O N D. Y L'aguna, Teneriffe . . Madeira. 9.6 6 \ 5-7 2 Q 6.1 ^6 2.2 I e 1-3* I 2 0.4 06 0.0 O O o.o O 3 0.4 I 2 2.2 "6 4.8 c c 10.9 43-6 Sahara I.O I.I I 7 I . 06 O 2 O ^ I 2 08 I O 12 2 France (south coast) Italy (south coast) . Beirut . 2-5 3-i c.4 2.0 2-5 7.2 J -7 2.8 4.0 1-7 2.8 o.q 2.2 2.2 O.7 I.O i-3 0.4 o-5 0.6 o.o I.O ,6 o.o 3-o 2-5 0.7 4.0 3-8 T 8 34 4.1 4-7 1-7 4.1 7.6 24.8 31-5 q62 Alexandria 2 I I 7 I O O I O O O O O O O O O I O 2 I 7 I Q 8 7 The subtropical rains occur in that region in which the annual shifting of the atmospheric circulation with the course of the sun brings into play in the summer the dry trade winds blowing from the east ; and in the winter the general winds from the west, which are rain-bearing, be- cause in that great western current occur rain-producing storms (the cyclones of lower middle latitudes, to be de- scribed later). When the trade winds from the east are present, the season is dry ; when the winds from the west are present, the season is wet. The stormy winds from the west make the west coast in this region very wet during the rainy season, because they bring the moist air from the ocean to the land. On the east coast the winds from the east may still be the most rainy, because they bring the moist air to the land, and the west winds have been deprived of much of their moisture in crossing the continent. Rainfall of the Temperate Region. In the- temperate region, which is to the poleward of the subtropical, MOISTURE: PRECIPITATION. 155 there is usually rain in all the months of the year, but the amount is not usually distributed evenly over all seasons. The rainfall causes are very complex in this region. It is the region of the greatest number of storms which occur in the strong permanent air current from the west. These storms are most frequent in the cold season, which in- creases the amount of rainfall at that time of year. In winter time, too, winds blowing the moisture-laden air from the oceans landward, carry it from a warmer to a colder region ; and, the temperature of the air being lowered, moisture is condensed, and falls as rain. In the summer time the interiors of the continents become greatly heated, and violent local storms arise, which cause an increase of summer rains. In western Europe the least rainfall occurs in midsum- mer, and the greatest in the fall or winter; in central Europe the minimum occurs in the early spring or win- ter, and the maximum in summer ; in Russia and Sibe- ria the maximum occurs in summer, and the minimum in winter. In North America, on the Pacific coast, there is a winter maximum of rainfall, and an almost rainless summer ; for the Mississippi River valley there is a summer maximum, and a minimum in the winter (but sometimes in the spring or fall) ; for the eastern coast of the United States there is a late summer maximum and early summer minimum for the middle regions, a winter maximum and early summer minimum at the north, and at the south a late winter maxi- mum and late summer minimum. In South America, on the west coast, there is a winter maximum and summer minimum ; in the interior this is reversed ; and on the east coast there is a winter minimum. 156 ELEMENTARY METEOROLOGY. On the South African coast there is a winter maximum and a summer minimum, but in the interior there is an autumn maximum and winter minimum. In Australia, on the eastern coast, there is a summer or early autumn maximum and winter minimum, and in the interior a very irregular rainfall. In south and west Aus- tralia there is a winter maximum and early autumn or sum- mer minimum. In the northern Arctic regions there is a late spring or early summer minimum and a fall maximum, except north of Asia, where there is a winter minimum and summer maximum, as for Siberia. The Intensity of Rainfall is obtained by dividing the total rainfall by the number of days on which rain falls. The intensity is calculated for each month and for the year. The distribution of the intensity of rainfall for the months follows more closely the amount of rainfall than it does the number of rainy days. In normal cases the intensity is greatest in the warm seasons, but this does not always so occur. Duration of Rainfall. The average number of hours per .day during which rain falls is quite variable for dif- ferent regions, not only for the year, but for the different months. The greatest average duration of the rainfall occurs in the cold season of the year in normal cases in the middle latitudes. On the coast of Norway the average number of hours is 1 1 : in southeastern England, 5 ; in the northeastern quarter of the United States, 5 ; in the southeastern quarter, 4; in the dry region in southwestern United States, 2.5 ; and in the dry Rocky Mountain region, about 3. The Variability of the Rainfall is obtained by dividing the average deviation of the individual cases from the average rainfall, by the average amount of rainfall. The variability MOISTURE: PRECIPITATION. 157 of rainfall (expressed in percentage of the average amount) in general increases with decrease of the absolute amount of rainfall. In regions of moderate rainfall (30 inches per year) 50 years of observations may leave a possible error of from 5% to 10% in the average annual rainfall. The variability in the number of days of rainfall in the year, and for different months, is found to be the least for regions having the fewest number of days with rainfall, and greatest for those having the greatest number of days with rainfall. The occurrence, during the same months, of the max- imum or the minimum phases of rain amount, rain frequency, and rain intensity, is rarely found in the continental regions of the temperate zone. The Probability of Rainfall is obtained by dividing the number of rainy days by the number of all the days during a chosen period, as a month or a year. This is quite an important element to be considered concerning the rain- fall. The variation is so great from region to region, or even within narrow limits, that of course the phases of maximum and minimum frequency cannot be pointed out unless special mention is made of a great number of regions. It does not necessarily always happen that the greatest frequency of rainfall occurs with the greatest amount of rainfall. Long-Period Fluctuations of Rainfall. There is prob- ably a periodic oscillation in the amount of rainfall extend- ing over a period of 35 years. During about 17 years the rainfall increases slightly in amount, and then decreases again through the same length of time. We have record of rainfall observations going back over 200 years at some places. The records for this century are more accurate and more numerous than those for the last century, and show for the land 15$ ELEMENTARY METEOROLOGY. the following periods of excess and deficiency of rainfall above and below the average amount : 1815, excess of rainfall ; 1831-35, deficiency ; 1846-50, excess; 1861-65, deficiency ; 1876-80, excess. What the cor- responding conditions were over the oceans, we do not know. The Amount of Oscillation of the Rainfall during the probable thirty-five year period is* of as great interest as the times of oscillation. The average oscillation is best expressed in terms of the average amount of rainfall. In Europe the oscillation is about 15%; in Asia, 30%; in North America, 25 % ; in Central and South America, 25 % ; the average for all the regions being about 25 %, or one fourth of the average total rainfall. It has been found that intensity of the oscillations increases towards the interior of the continents. Thus in central Siberia 2.3 times as much rain falls in the period of excess of rain as in the period of de- ficiency, while in England it is but 1.2 times as much. Observations along the coast of the Atlantic Ocean have shown that there is a ten- dency towards a minimum of rainfall at the period of maximum rainfall in the interior of the continent ; and this suggests that on this ocean, at least,, there may exist a compensatory rainfall oscillation the reverse of that for the land. There is, during these long-period oscillations of rainfall, a shifting back and forth of the rainfall areas, or, as the matter may best be expressed, a shifting of the lines of equal rainfall. A rough computa- tion of this shifting of the regions of equal rainfall oceanwards and land- wards shows that the region of 24 inches rainfall lies in Euro-Asia 1,000 miles, and in North America 700 miles, farther inland in the period of excess of rainfall than it does in the period of deficiency. The oscillations in the dry regions of Siberia are enormous. The rain- fall which occurs in \\^est Siberia in the relatively wet period is only to be met with at a distance of 2,000 or 3,000 miles from that region during the relatively dry period. The oscillations are least on the coast, and increase towards the interior of continents. The Permanent Increase or Decrease of Rainfall in special localities, due to the increase or decrease in the forests, has MOISTURE PRECIPITATION 159 oeen the subject of much investigation, and it has not as yet been satisfactorily settled, Experiments for the arti- ficial production of rain by means of explosions high in the air are not likely to be successful, as no appreciable amount of rain could be produced in this manner, unless ascending air currents could be caused and maintained for some time. HaiL When raindrops become frozen in their passage through the air, they fall as hail. Sometimes the rain- drops may be frozen in their downward passage ; but it is believed that they are usually frozen by being first carried upward by vertical air currents into regions where the tem- perature is below freezing, and that they then descend to the earth before they have time to melt. Sometimes hail- stones have a soft snowy center ; and frequently they have several successive coats or shells of ice, probably due to their being subjected a number of times to alternate tem- peratures above and below freezing. SnoWc - When condensation takes place at a tempera- ture below freezing, minute ice crystals instead of water globules are formed, and the union of these crystals gives us snowflakes. Snow crystals assume a great variety of beautiful hexagonal forms, some of which are shown in the accompanying figure (Fig. 44). While the condition of freezing is necessary for the formation of snowflakes in the air above, yet they frequently, by falling rapidly without melting, reach the ground when the temperature of the lower air is much above the freezing point of water. Latitudinal Limit of Snowfall near Sea Level . The temperature of the lower air increases from a temperature at the poles much lower than that required for snowfall, to a temperature at the equator much higher than that at which snow can exist. So there must be some region be- tween the two where snow ceases to fall ; and to the pole- (i6o) FIG. 44. SNOW CRYSTALS. MOISTURE: PRECIPITATION. l6l ward of this region the length of the season during which snow falls increases towards the poles, while on the equa- torial side snow never falls. The equatorial limit of snowfall is in general on the continents at about the Tropics of Cancer and Capricorn, or 23 J of latitude ; but in western South America it reaches much nearer to the equator. Over the oceans the latitude of 35 is about the limit. It is found that in very cold weather snow does not fall ; and as there must be a limit to the height of Lhe temperature at which snowrlakes can exist, so there must be some intermediate temperature of maximum snowfall. The subject has not been very extensively investigated, but some observations in the mountains of Germany have shown that, there at least, the snows occurred between the limits of 48 and 8 F. for the lower air temperature, with an average temperature of 30 F. Measurements of Snowfall are first made by means of a graduated rod, with which the depth of the snow is measured at a number of places which are protected from the wind, and yet freely accessible to the snow. The snow in and over the rain gauge must then be melted, and the amount of the resulting water determined by the usual method of measuring rainfall. The average density of recently fallen snow may be taken as about o.io (or io inches of snow make i inch of water) ; but snow is lighter (less dense) when it first falls than after it has stood for a while, lighter in cold than in warm weather, lighter in woods than on an open plain, lighter in gentle than in strong wind, and the upper layers of fallen snow are lighter than the lower. The density of snow may vary from 0.02 to 0.9 (the density of ice), the density of water being taken as i .00 ; the conditions causing this great variation being the hu- midity of the air, the temperature, the wind velocity, the immediate locality whence the snow is taken (whether from drifts or protected places), the time that the snow has lain on the ground, the form of the snowflakes, the layer of snow chosen, etc. In general, in a cold 1 62 ELEMENTARY METEOROLOGY. climate it may be assumed that the fallen snow is lightest in the early winter (in Siberia, 0.15), moderately dense in midwinter (in Siberia, 0.2), and most dense in spring (in Siberia, 0.3). Dew. Just as soon as the temperature of the surface of the ground (or of any other surface) falls below the dew- point of the adjacent air, the latter gives up part of its vapor in the form of small water drops, or dewdrops as they are called, which are deposited on the cooled surface. When the temperature of the surface lies below the freez- ing point of water (o C. or 32 F.), then the dew is de- posited in small ice crystals, and it is called hoarfrost. On account of the rapid cooling by radiation, especially on clear nights, the temperature of the ground and other solid substances becomes colder than that of the air above, and the dew-point and even frost point are reached by the ground and the adjacent air layer, while the air at the height of a few feet above the ground has a temperature several degrees above those points. It has been supposed that in very dry rainless regions plants depend in a great measure on the dew deposit for their supply of water, but the amount of moisture obtained in this way is quite small. In central Europe it was found that scarcely one inch depth of dew was deposited during a year, which is about 3 % of the rainfall. Night Frosts occur when plants are exposed to tempera- tures under the freezing point, and the conditions are much the same as for hoarfrost formation. Since by the condensation of vapor as dew, latent heat is given out and thus retards the cooling, the temperature during the dew- fall does not go much below the dew-point; so that when the dew-point lies above the freezing point, frost is hardly to be expected; but when the dew-point lies below the freezing point, frost may be expected on clear nights. The drier the air, the more likely will be the dew-point to be MOISTURE: PRECIPITATION. 163 below the freezing point, and therefore the more likelihood of a frost. This is seen on a large scale in the fact that frosts occur more frequently at inland than at coast sta- tions. Evaporation of Water. A portion of the water which is found on the earth's surface is given off to the overlying air by the process of evaporation. The rapidity with which this evaporation takes place varies not only with the surface itself, whether it be a free water surface or wet ground or vegetable growth, but also with the tempera- ture, the relative amount of water already in the air, the motion of the air, and the atmospheric pressure. In any case the amount of water evaporated varies directly with the amount of surface from which the evaporation takes place. There is great variability as to the rapidity of evapora- tion from the ground ; but it may be said that the smaller the size of the particles of earth, the more rapid will be the evaporation, since the capacity of the ground for holding water, and the capillary conduction of the water to the surface, increase with the fineness of the earth particles. From a hard-packed clay ground surface, almost no evapo- ration takes place. The amount of evaporation from plants is enormously great. The passage of water through a plant from the ground to the free air is called transpiration. Five times as much water has been transpired and evaporated from a plant as from a water surface, and more than twelve times as much as from an ordinary land surface during the same time. A short series of experiments has shown, that, accord- ing to the observed temperatures of the dew-point and the surface of the snow, the evaporation from the snow ex- 164 ELEMENTARY METEOROLOGY. ceeds in amount the condensation from the air by means of its cold surface. As a standard for measuring evaporation, the amount evaporated from a freely exposed but shaded surface of pure water is chosen; and the amount of water lost by evaporation is calculated by first measuring or weighing the whole amount when first exposed, and then that which remains in the vessel. Pure water evaporates probably 25 % more rapidly than salt water. A very convenient instrument for measuring the amount of evaporation during the warm sea- son is the Piche evaporimeter, and it can be mounted beside a thermometer and read with as much ease as one. This instrument (shown in Fig. 45) consists of a graduated glass tube closed at one end. The tube is filled with water; and over the open end is placed a cir- cular disk of porous paper, with a very small opening at the center. The instrument is mounted with the porous paper downward, and this last always presents to the air a wet surface from which the water is evaporated. As evapora- tion goes on, the upper part of the tube becomes emptied of the water, and the air makes its way up through the water column to fill the vacated space. The variation of the height of the water in the tube, as read off from the graduated scale, shows the amount of water evaporated from the tube. Since, however, disks of various degrees of porosity offer different resistances to the pas- sage of the water, the instrument must be compared with a standard evaporimeter, or a disk of known porosity must be used. FIG. 45. PICHE EVAPO- RIMETER. Water when evaporated becomes a vapor which tends to distribute itself through the atmosphere somewhat after the manner of the air in an empty space. Vapor-laden MOISTURE: PRECIPITATION. 1 65 air is also transferred by the atmospheric currents to regions where there is a less amount of vapor. Periodic Variations in Evaporation. The amount of evaporation has a daily and annual march following that of the temperature. The evaporation is least a-t night, and greatest shortly after noon. The amplitude of daily variation is greatest in the interior of continents, and least on the coasts. The evaporation during the day is several times greater than during the night. During the year the least evaporation is in midwinter, and the greatest in mid- summer. At Nukuss (continental exposure in central Euro-Asia), the evapo- ration in January was about I inch, and in July nearly 12 inches. At St. Petersburg (seacoast), however, the evaporation was less than 0.2 of an inch in January, and less than 2.5 inches in July. The amplitude of annual variation increases with the latitude and towards the interior of continents. The Average Annual Amount of Evaporation varies greatly, but in general it decreases with the latitude. In the tropics, from a water surface, it amounts to perhaps 90 inches ; in latitude 40, to perhaps 30 inches ; and in polar latitudes, to 10 inches. In hot desert regions, as in southwestern United States, it reaches even 1 50 inches. The amount of evaporation possible over the land is greater than the amount of rainfall in most regions of the earth ; but this amount of water does not evaporate from the ground, because there is not always sufficient moisture at the surface to keep up the rate of evaporation which could take place. WALDO METEOR . IO CHAPTER VII. ATMOSPHERIC OPTICS AND ELECTRICITY. Luminous Atmospheric Phenomena. There are a num- ber of luminous atmospheric phenomena which come to our notice, and which it will be convenient to mention at this stage. They are principally of two classes. The one pertains to the optical effects produced by and in the at- mosphere ; the other concerns the electrical condition of the atmosphere. Transparency of the Air. Any substance that one can see through is said to be transparent. But some sub- stances are more transparent than others; that is, we can see through greater thicknesses of some substances than of others, and with greater distinctness. We call air a transparent substance, but the degree of transparency is not constant. It varies with the density of the air and the amount and form of moisture which it contains. Thus at sea level, and for warm moist air, the transparency is least, while on high mountains with cool dry air the transparency is greatest. Atmospheric Optics. Light, in passing through the air, is subject to reflection, refraction, diffraction, and absorp- tion ; and, as a result of these, there arise a number of phenomena which must be briefly mentioned. Refraction of Light is the bending of light rays. Light, in passing from a rarer to a denser medium, is bent from its straight course so that the angle which the ray of light 1 66 ATMOSPHERIC OPTICS AND ELECTRICITY. l6/ makes with the perpendicular to the boundary surface is less in the denser medium than in the rarer medium ; and where it passes from a denser to a rarer medium, the angle which the ray makes with the perpendicular is increased. This angle in the first medium is called the angle of incidence, and in the second the angle of refraction. In Fig. 46, BA is the ray of light passing through a rare medium (as, for instance, air) ; and upon its entrance into a denser medium (as, for instance, water) the ray will be deflected from the direction of its path BA, and will take the course AE. If the line CD is perpendicular to the dividing surface between the two media, then BAC is the angle of incidence, and DAE is the angle of refraction. The amount of refraction, or the difference between the angles of incidence and refraction, varies with changes in the density of the transparent media ; and in the case of light entering the atmosphere from without, the rays are gradu- ally bent more and more as the denser air is encountered. The accompanying figure (Fig. 47) illustrates the bending of the solar rays entering the atmosphere. When the sun is below the hori- zon, at C, it would be invisible at A, on account of the curvature of the earth, if there were no atmosphere ; but the solar rays entering the atmos- phere near the point B are refracted so that they reach A, and the sun appears to be at D, though really at C below the horizon, either in the morning or in the evening. So that, in the polar regions, the sun is visible while it is in reality below the horizon, and is thus seen earlier and later during the time of polar sunlight. The effect of atmospheric refraction is to slightly in- crease the amount of solar rays reaching a place. 1 68 ELEMENTARY METEOROLOGY. Where the light enters the atmosphere at a small angular distance from the zenith, the amount of refraction can be quite accurately com- puted, but near the horizon it is very difficult to determine. For a zenith distance of 87.5 (i.e., 2.5 above the horizon), the refraction is computed to be 18.5' ; but at the horizon it is about 30'. The rays of light coming from the sun possess different wave lengths of vibration ; and, as these rays pass from FIG. 47. REFRACTION OF SOLAR RAYS BY THE ATMOSPHERE. one medium to another of different density, there is more or less of a dispersion of the rays, and those possessing the longest wave lengths are refracted the least, and those with the shortest wave lengths are refracted the most ; so that the rays which enter the second medium as white light are spread out by the refraction, and separated into the prismatic band of colors, red, orange, yellow, green, blue, indigo, and violet, ranged according to the wave lengths and refrangibility of the rays. Where any one color is observed, it shows that the rays with vibrations corre- sponding to this color are in excess of any others which may be present. Reflection of Light is the throwing-back of light rays from the surface on which they fall. When the rays of light ATMOSPHERIC OPTICS AND ELECTRICITY. 169 strike a surface, they are reflected from this surface again in such a direction that they make the same angle with the perpendicular to the surface before and after reflection. Thus, if a ray of light passes from A towards B (Fig. 48), it will be reflected from B towards C\ and the angle ABF, called the angle of incidence, equals the angle CBF, called the angle of reflection. The reflection of light in the atmosphere is very important in its effects. Reflection of Light from Dust Particles. The reflection of light from the minute particles of dust in the atmosphere diffuses and scatters the light received from the sun in all directions, and illumines the sky. This illumination of the sky is greatest at low altitudes, where the air contains more and larger dust particles than at higher altitudes. Mirage. When one layer of air lies in contact with another layer of different temperature and density, the boundary surface between the two, if it is sharply defined, will reflect light quite perfectly. When the observer is above such a surface, he sees objects beyond and above (but not below) this surface reflected from it as from a horizontal mirror. In case the observer is below this boundary surface, he sees the objects beyond and below (but not above) reflected from it. The images are in- verted. This phenomenon is called mirage. Diffraction of Light is the dispersion or breaking-up of light rays. When a beam of light passes through an aggregation of exceedingly minute particles, or when it strikes a surface roughened or divided into very small surfaces, the rays are broken up and scattered or dif- fracted. The interference of some of these rays causes I/O ELEMENTARY METEOROLOGY. some of the color rays to partially, and others to wholly, disappear. Colors of the Sky. The light rays having the more minute wave length (those nearer the violet end of the spectrum) are diffracted by the more minute as well as by the coarser particles, while the rays of greater wave length (those nearer the red end of the spectrum) are diffracted more by the coarser particles ; and it is on this " selective " diffrac- tion and reflection that much of the coloring of the sky and other more definite optical phenomena depend. The blue color of the sky (away from the sun) is due to the more powerful diffraction of the blue rays, so that rays of that color are reflected towards the observer, while those of coarser wave length pass on and do not reach him. As the observer turns his eyes gradually towards the sun, the sky loses its blue color, and takes on a more neutral or white color, because the direct rays which now reach him do not contain to any great degree the blue component which had been previously reflected toward him from other directions. When the sunlight passes through the atmosphere, we lose more and more of the colors having smaller wave length, the thicker and denser the air mass through which the rays come, so that the sun's rays reaching us contain the most blue when in the zenith; but as the sun ap- proaches the horizon, the rays at the blue end of the spectrum become scattered, and lost in the air, and we receive only those at the red end, and consequently the sun appears red at the horizon. If the sun could be viewed without the intervention of an atmosphere, it would present the distinctly blue color. The glow of the sky which accompanies the rising and setting of the sun is due to the diffraction and reflection of light by the minute particles in the air. As the sun ATMOSPHERIC OPTICS AND ELECTRICITY 171 sets below the horizon, the intervening portion of the earth cuts off the rays from the lower atmosphere (at the point of the observer), and the diffracted rays forming the glow appear in the upper atmosphere only. When the particles of matter in the air are unusually numerous at high altitudes, the sunset glow of the sky is extraordinarily brilliant. Such a condition prevailed during 1883 and a few subsequent years. Corona and Halo. The sun or moon is occasionally surrounded by one or more well-marked rings or circles of light, but not always of the same diameter or color. These rings are divided into two classes, the corona, of small diameter ; and the halo, of greater extent. In the corona, the color of the inner part of the ring is blue, and of the outer, red. For the larger halo, the order is reversed; and red is on the inside and blue on the out- side of the ring. The radius of the halo is about 22, 45, or 90. The corona may vary from i to 10 or more in radius. The corona is a diffraction phenomenon, while the halo is due to refraction and reflection. The corona may consist of several rings concentric with the sun or moon, and occurs when mist or thin clouds par- tially obscure those luminaries. Coronae are formed by the diffraction and interference of light caused by the small water particles in the cloud. The larger the water particles, the smaller will be the ring ; and it is when the particles are of different sizes that coronal rings of different diameters exist at the same time. The halo occurs in connection with the higher cirrus clouds only. Halos are more frequently observed than coronae. The faint halos are much more easily detected by observing in a black mirror the reflection of the light center and its neighborhood. There is great variety in 172 ELEMENTARY METEOROLOGY. the distribution of halos, sometimes the rings are wholly separate, and sometimes they intersect. The points of in- tersection are unusually more or less bright patches, and are called mock suns, or sun dogs (Fig. 49). Mock Snn FIG. 49. HALO PHENOMENA. In general, the halos may be divided into the three following classes : 1. Rings which have the sun at the center, and have a radius of about 22. The rings are about i wide, and are colored with the red on the inner side, where they are sharply defined, but on the outer violet side they gradually vanish. The space directly around the sun is bright, but towards the ring this brightness gradually disappears. 2. Rings which pass through the sun and have a radius of about 45- ATMOSPHERIC OPTICS AND ELECTRICITY. 173 These rings are colored, and have the red side towards the sun. The width of the ring is about 3, and although not so intense as the first kind, yet the colors are more distinctly separated. Sometimes a ring of this nature is visible around the whole sky parallel to the horizon at the altitude of the sun. \ \ FIG. 50. FIG. 51. FIG. 52. 3. Rings which turn a convex side towards the sun, and have a radius of about 90. These are of weak intensity and of indefinite color. Only portions of the ring usually are visible ; and they touch some of the smaller rings which encircle the sun. Explanation of Halos. These ring phenomena are mainly due to the refraction of the light by the ice crystals or ice needles of which the cirrus clouds consist. These needles are as- sumed to be six-sided (Fig. 50), and to hang suspended in the air. When they hang verti- cally, and the rays of light pass through as if the needles were triangu- lar (Fig. 51), then the light is diffracted as shown in Fig. 52, and the ring has a radius of 22. When some of the ice needles are inclined to the ver- tical, then the light may enter at another surface, and be diffracted through 90, and the ring of 45 will result. In order to cause the ring of 90 radius, the light must pass through the crystal, as shown in Fig. 53, and suffer total reflection. FIG. 174 ELEMENTARY METEOROLOGY. The horizontal ring parallel to the horizon is due to the reflection of the light from the outer surface of the ice crystals, which are at the altitude of the sun, and therefore possess the proper inclina- tion of the surfaces to reflect the sun's rays to the observer's eye ; and since no refraction occurs by means of the ice needles, there is no color- ing, and the ring appears white. The broken rings or arcs are likewise reflection and not refraction phenomena. The Glory, Brocken Specter, or Fog Image is analo- gous to the coronal phenomenon. It is brought about by the sun casting a shadow of the observer on a fog or cloud bank. This shadow, sometimes of huge dimensions, is surrounded by a glory of light, which is caused by the diffraction of the rays of light by the water particles (near the observer), and the resulting separation of the prismatic colors which are reflected to the eye of the observer by the fog particles. Sometimes, where the observer is on the top of a mountain, the whole peak may have its shadow thrown against the fog bank. At times a seemingly distant and white fog bow encircles the shadow of the observer and the glory which surrounds the head. A glory of light is also sometimes to be seen around the shadow of the head of the observer when it falls on bedewed or wet grass. Rainbows are produced by the refraction of the sun's rays by means of the raindrops in the air, after which the separated band of colors is then reflected to the observer's eye. The center of this bow is opposite to the sun. The radius of the bow is 40 to 42, with sometimes a fainter secondary bow outside, with a radius of 50 to 54. The inner bow has the blue color on the inside, and the red on the outside ; while this order is reversed for the outer bow. Rainbows are most frequent in the local showers in which the sun suddenly breaks through the clouds at the edge of ATMOSPHERIC OPTICS AND ELECTRICITY. 175 the storm. Where there is a wide distribution of cloudiness, as in our long-continued rains, the proper conditions for the formation of the bow will seldom be present. Rain- bows are visible when the sun is at low altitudes ; and the nearer the latter is to the horizon, the greater will be the length of the bow visible in the sky. The bow is 180 in length, or a semicircle, when the sun is close to the horizon. Atmospheric Electricity We find the atmospheric air usually in a state of positive electrification ; that is, it is generally charged with positive electricity, which is the condition of glass when rubbed with silk. Very great fluctuations, however, occur, especially during thunder- storms, snowstorms, etc. ; and the electrification becomes sometimes negative, which is the condition of glass when rubbed with resin or sealing wax. The origin of atmos- pheric electricity is still unaccounted for, but it may be due to the frictional action of the air. The degree of electrifi- cation of the air is measured by means of an instrument called an electrometer. There are two classes of atmospheric electric phenom- ena which we shall notice : they are auroral displays, which may partly occur in the lower atmosphere, and partly in the highest layers ; and lightning displays, con- fined to the lower air layers. The Aurora called in the northern hemisphere the aurora borealis, or northern lights, and in the southern hemisphere aurora australis, or southern lights is an illumination of the sky which occurs in middle and polar latitudes, and which has a zone of maximum frequency from the 6oth to the /oth parallel of latitude north and south of the equator. It rarely occurs within the tropics. It is pos- sible that some of the widespread, intensely active, auroral phenomena have an electromagnetic origin ; while those of 1/6 ELEMENTARY METEOROLOGY. local occurrence may be simply the electrostatic charge of the air layer, rendered luminous by atmospheric changes. As usually observed by us, the aurora consists of an arch or band of light. It may, however, have various forms, such as that of an arch, ribbon, collection of beams, corona, and haze or diffused light. The auroral arch usually spans the sky in an east-westerly direction. In the middle latitudes it is seen to the north, and in the extreme northern latitudes to the south. The dimensions are variable. (A band observed in 1893 was supposed to be 15 miles wide, perhaps 250 miles high, and over 1,000 miles long." In gen- eral, the arch may be said to vary between an unknown lower limit and 300 miles in altitude. The arch sometimes varies in intensity in different parts. Frequently band-like stream- ers of light are seen perpendicular to the arch, and they extend to an unknown height. Sometimes these stream- ers are constant during their existence, and sometimes they are intermittent and pulsating, flashing out at intervals. Aurorae in the northern hemisphere are least frequent in January and June, and most frequent in March and October. There seems also to be an ii-year period of frequency such as was found by astronomers for sun spots. Lightning. The air between two clouds charged with electricity, or between a charged cloud and the earth, is subject to an electrical strain or pull. When this strain upon the air column becomes too great for further resist- ance, a disruptive discharge takes place. This discharge may vary in character from the invisible silent lightning to the violent impulsive rush discharge which has an enormous amount of energy. The lightning flash is the luminously heated air in the path of the discharge. The flash may have a duration varying from 3-^ of a second to a second. The thunder is but the crackle of the electric ATMOSPHERIC OPTICS AND ELECTRICITY. 177 discharge, but its sound is frequently reenforced and pro- longed by numerous refractions and reflections from adja- cent clouds and other objects. The needle of an electrometer which indicates the amount of the elec- trical strain will show by its fluctuations or " breaks " not only all of the flashes or discharges of lightning which are visible, but also the silent discharges which are of frequent occurrence but are invisible. The typical forms of lightning flashes (which have been studied by the aid of photography) are, 1 . Stream lightning, which consists of a plain, broad, smooth streak or flash of light. 2. Sinuous lightning, which consists of a flash following some one general direction ; but the line is sinuous, bending from side to side. This is the most common type. 3. Ramified lightning, in which part of the flash appears to branch off from the main stem like the branches of a tree from the trunk ; but whether these branches issue from the trunk or unite with it, is un- known. 4. Meandering lightning, in which the flash appears to wander about without any definite course and form irregular loops. 5. Beaded lightning, in which a series of bright beads of light appear along the white streak of lightning. 6. Dark flashes of lightning have also been photographed. The causes of the different forms are not well understood. Since the electrical discharge between the clouds and the ground will take place more readily where the distance is least (other conditions being equal), it is quite probable that the points of the clouds which appear suspended from the main thundercloud mass will the most frequently form the places from which the discharge takes place. In case of a tornado-funnel cloud, in which the cloud reaches nearly or quite to the earth, the condition for a discharge along this would be most favorable. Also at the time of the greatest rainfall the electric discharge would be greatly facilitated by the excellence of the conduction which would ELEMENTARY METEOROLOGY. be established by the falling water between the cloud and the earth. Periodic Changes in Atmos- pheric Electricity. There is a periodic daily change in the de- gree of electrification of the air. There are times of maxima at about eight or nine o'clock in the 1 morning and evening, and of 5 minima at two to four o'clock at s night and in the afternoon. The cause of this periodicity is not known with certainty. There is an irregular annual periodic change, of which we can only say that in general the maximum electrification occurs in the cold season, and the minimum in the warm season. The accompanying diagram (Fig. 54) shows the fluctuations in the *? potential of atmospheric electricity during a couple of days, as observed by means of an electrometer. In the curve not only do we find rapid fluctua- tions in the positive electrification, but there are also sudden changes from a powerful positive to an almost equally powerful negative electrification when the thunderstorm is in progress. St. Elmo's Fire is an elec- trical phenomenon sometimes ATMOSPHERIC OPTICS AND ELECTRICITY 179 seen as a continuous luminous discharge from elevated points. It usually has the appearance of jets of flame issuing from the points of trie objects, and is accompanied by a hissing or buzzing noise. It is most frequent in winter, and is almost invariably accompanied by a heavy fall of soft hail or snow. CHAPTER VIII GENERAL CIRCULATION OF THE ATMOSPHERE Some Preliminary Ideas concerning Air Motions, We come now to the study of the systems of movements of the air, which have already been mentioned in the chapter on winds. It is first necessary, however, to state in a gen- eral way the nature of air motions, and the changes which air undergoes, and its conditions with respect to the sur- rounding air, when subjected to these motions, especially to vertical motions. These are briefly stated in a neces- sarily disconnected manner, and their bearing or impor- tance cannot be fully seen until they are applied to proper cases as they come up in their natural order. . General Air Motions. Atmospheric motions may in general be divided, according to direction, into two classes, horizontal and vertical. Air currents are seldom in- clined at a large angle (such as 20 or 45) to the hori- zontal or vertical directions. Horizontal air motions are those nearly parallel to the earth's surface, and are produced by gradient forces effec- tive in those directions. In horizontal currents the changes in density and temperature of the moved air mass are usu- ally very gradual and relatively slight. Vertical air motions arise from two causes. They are either a natural consequence of the horizontal motions, in order to preserve a continuity of the air taking part in these motions when they have dissimilar directions ; or they are 180 GENERAL CIRCULATION OF THE ATMOSPHERE. i8l due to the endeavor of a local mass of air having an abnormal condition to move to a place where its condi- tion will be normal, i.e., the same as that of the mass of surrounding air. In vertical currents the changes in density and temperature of the moved air are rapid and of relatively great magnitude. Centrifugal Force. If a body (a mass of air, for in- stance) has a motion in any direction, it will continue to move in a straight line with constant velocity, unless it is acted upon by some outside force or is subjected to some sort of resistance. When a body is constrained to move in a circular path, it still has a tendency to move in a straight line, which would be in the direction of the tan- gent to its circular path, at whatever point the body hap- pens to be. The force with which the body presses out- ward, in its endeavors to depart from the circular motion, is called the centrifugal force. This force would be the pull exerted on a cord by a weight attached to one end of the cord, if it were whirled around in a circle of which the other end of the cord (or any part of it) was the fixed center of revolution. In case a mass of free air is forced to move in a curvi- linear path, it will exert a certain amount of centrifugal force outward from the center of rotation ; and the amount of this force is exerted as a pressure against outer adjacent masses of air. Since air is an easily yielding and com- pressible substance, the effect of this centrifugal force is to diminish the amount of air within the circle, and to increase the amount without the circle. Where the whole of a mass of air has a rotation around a point at its center, the centrifugal force causes the air to recede most at this center, and then less and less at increasing distances from the center. WALDO METEOR. II 1 82 ELEMENTARY METEOROLOGY. Conservation of Areas. When a body (air, for example) is acted on by a central force, or the opposite, and has a gyratory motion around this center, then the varying line connecting this body with the center (called radius vector) sweeps over equal areas of space in equal times. There- fore when the air approaches the center of revolution, its velocity of revolution must increase as the radius vector shortens ; and conversely, when it recedes from the center, the velocities decrease. Energy of the Air. While, from the fact that we live in the lower part of the atmosphere, we must deal mostly with the winds near the surface of the ground, yet it must not be assumed that the upper half of the air (that is, the portion above the altitude of about three miles) can be neglected in studying its conditions, for it has been com- puted that in this upper half of the atmosphere there is six times as much gradient force or energy productive of motion as in the lower half. The altitude of about 40,000 feet above sea level, which is about the altitude of the high- est clouds (the cirrus clouds), divides the energy of the at- mosphere into two parts. The amount of gradient force or energy above this altitude is equal to that below it. The Heat in the Air. The heat which is in the atmos- phere is, as we have seen, derived by a number of proc- esses. There is first absorbed about half of the direct solar heat ; then about half of the heat reflected from the ground and water surface of the earth ; then all the heat supplied from the earth's surface by convective currents, and the heat supplied by radiation from the earth's surface (just as it is radiated from the sun). All this heat must be lost from the air through a process of radiation from the air itself, which must not be confused with the radiation through the air from other objects. GENERAL CIRCULATION OF THE ATMOSPHERE 183 Radiation of Heat from Air. The process of radiation of heat from, and of absorption of heat by, layers of air within the atmosphere, depends on the difference of tem- perature, and perhaps on the difference of density, of these layers ; and without doubt a considerable portion of the heat which escapes from lower air layers passes directly through the upper air into space, and the portion of heat which is absorbed by the air passes in its turn through the layers still above, and finally makes its escape. This radiation of heat from the air itself is shown by the fact that the cooling of the air which takes place on a clear night is quite general; while, if it depended on the conduction of the heat from the air to the cooler ground to replace that lost by radiation from the latter, there would be a cooling of the air to a height of not more than 10 feet during, say, twelve hours of darkness. The rate of actual cooling of the air by radiation varies approximately as the difference in temperature of the air and its surroundings, and it is therefore probably least at the middle altitudes of the atmosphere. Comparison of Solar Heat directly absorbed by, and radiated from, the Air. The amounts of heat absorbed by the air and radiated from it again are nearly equal, under the maximum conditions of absorption, during the same period of time. But since the radiation from the air into space goes on at all times, and since the direct absorption of the solar heat rays can take place only when the sun is shining, the outward radiation is much more effective in cooling the air than is the direct action of the sun's rays in warming it. Adiabatic Heating and Cooling of Air Dynamically. - The adiabatic heating and cooling of masses of air by the variation of the air pressure, due to the ascending or 184 ELEMENTARY METEOROLOGY. descending movement of the air, is one of the most impor= tant factors in atmospheric events. Cooling of Dry Air in Ascending, and Heating in Descend- ing Currents. Theoretically, the rate of cooling in ascend- ing and the rate of warming in descending dry air is i F. for each 183 feet change in altitude, whatever the temper- ature of the air. If the air is in any way subjected to a change of pressure equivalent to that which would occur in passing through 183 feet in altitude, it will experience the same changes of temperature as it would if actually moved through this distance. Action of Ascending Currents of Moist Air We have just been considering the air when it is dry ; but when moisture is present, as it always is in the atmosphere, we have no longer the simple conditions just mentioned. The air, as it cools adiabatically in its' ascent (as just discussed), has a diminished capacity for moisture ; and when the limiting temperature is reached, at which con- densation of the moisture present occurs, then, with the condensation which follows, there is a freeing of latent heat, which retards the further cooling of the air. Thus there is heat added to the air mass ; and if the air keeps on in its ascent, the decrease of i F. per 183 feet of ascent is diminished by the amount of latent heat supplied by this condensation. But until condensation does actually take place, the rate of this decrease of temperature is nearly i F. per 183 feet, even when moisture is in the air. Descending Currents of Moist Air. The rate at which air becomes heated in descending is i F. for each 183 feet, whether the air is dry or saturated (except where rain is held in the air, or where supersaturation exists). Thus the rate of change of temperature may be very dif- ferent in ascending and descending currents of moist air. GENERAL CIRCULATION OF THE ATMOSPHERE. 185 The Actual Rate of Decrease of Temperature with the alti- tude is less than that assigned by theory, and varies some- what on account of various irregularities in the existing atmospheric conditions. It is greater in summer than in winter, and differs in cloudy weather from that in clear weather. Still these irregularities are but departures from the laws just mentioned, and their causes would have to be examined separately. Indifferent Equilibrium. If the temperature of the unsaturated quiescent or horizontally moving air adjacent to or surrounding a vertical current of air decreases i F. for each 183 feet increase in altitude, then the air is said to be in indifferent equilibrium. For unsatu-rated air which is moved upward or downward, and which changes its tem- perature at this rate, will remain in the new position in which it is placed, because its change of temperature has been just such as to accommodate itself to the temperature of the surrounding air in its new position. Stable Equilibrium. Unsaturated quiescent air which decreases in temperature at a rate less than i F. for each 183 feet of increase in elevation is said to be in stable equilibrium : for when a portion of such air receives an upward motion, its density 'gradually becomes greater than that of the surrounding air at the same level, and it would sink back again to its starting place if the force which moved it ceased to act. The differences in air density depend on the relation of decrease of temperature with altitude, in the air which is in motion, to that in the air which surrounds it. Since the rate of decrease of temperature of the unsaturated ascending air is about i F. for each 183 feet of ascent, then, if the temperature of the surrounding quiet air decreases at a rate less than this, the ascending air will cool the more rapidly, and will become the denser as it ascends ; and when the force which made it 1 86 ELEMENTARY METEOROLOGY. ascend is spent, the air which has ascended will fall back again to its orig- inal position. Similarly, if a downward current of air becomes heated faster than the surrounding air, then it becomes lighter than the air at any lower altitude at which it may arrive ; and when the force which causes it to descend is spent, the air will rise again to its former position. Unstable Equilibrium. Unsaturated quiescent air which decreases in temperature at a rate greater than i F. for each 183 feet of increase in elevation is said to be in imstable equilibrium: for when a portion of such air is moved up- wards it becomes lighter than the surrounding air, and when moved downwards it becomes still heavier than the surrounding air, at any level which it may reach ; and thus there is a tendency for the moving air to continue in the direction in which it is started. This condition occurs when the temperature of the moving air varies at a rate less than that of the surrounding quiescent air. Conditions of Equilibrium illustrated numerically. In the accom- panying table (from Ferrers " Winds "), the first column gives the alti- tudes ; the second column, an assumed temperature at these altitudes which would produce stable equilibrium ; the third column, the tem- peratures which would produce indifferent equilibrium ; and the fourth column, an assumed temperature which would produce unstable equi- librium in case vertical currents arise. ALTITUDE. TEMPERATURES FOR STABLE EQUILIBRIUM. TEMPERATURES FOR INDIF. EQUILIBRIUM. TEMPERATURES FOR UNSTABLE EQUILIBRIUM. Feet. F. F. F. 9,840 42.8 32 21.2 8,200 50.0 41 32-0 6,560 57-2 50 42.8 4,920 64.4 59 53-6 3,280 71.6 68 64.4 1,640 78.8 77 75-2 ooo 86.0 86 86.0 GENERAL CIRCULATION OF THE ATMOSPHERE. 187 Here we see, that since a vertical air current has a change of tem- perature with the altitude of about i F. per 183 feet, then, when the rate of change of temperature in the surrounding quiescent air is less than this, the condition is that of stable equilibrium ; when it is equal to it, the condition is that of indifferent equilibrium ; and when it is greater, the condition is that of unstable equilibrium. General Circulation of the Atmosphere. The winds be- longing to the general circulation of the atmosphere are the hemispherical systems of winds extending from the equa- tor to the poles. It has been found by observations of wind vanes, that in the northern hemisphere, in the lower air, there is a decided prevailing wind blowing from the east or northeast over the region extending from near the equator almost to the tropics ; while from a little beyond the tropics to the polar regions there is a prevailing wind from the west or southwest. Directly at the equator and in the region of the tropics, there is no such prevailing direction of wind, and in fact these may be designated regions of relative calms. These surface wind directions are shown in a general way by the arrows (drawn flying with the wind) on the charts, Figs. 26 and 27, and are also shown by the heavy arrows on the inner, shaded portion of Fig. 55. At high altitudes, mostly in the regions above the highest clouds, there are prevailing winds blowing from the west or southwest over the entire hemisphere from the equatorial to the polar regions. Between the winds of the upper and lower regions, in which the wind direction is mostly poleward, and extending to the ground in the polar regions, there exist return currents flowing equatorward^ which prevent the accumulation of air in the polar regions. It must not be supposed that these general wind direc- tions are continuous and uninterrupted. As a matter of fact, there are continually occurring within them local air i88 ELEMENTARY METEOROLOGY, movements of sometimes thousands of miles lateral extent, which entirely change, during their continuance, the direc- tion of the main air currents. So it is only by observing the winds for a great length of time that the average nor- mal direction can be determined. In the soutJiern hemisphere these wind directions are somewhat reversed : the prevailing lower currents between the Tropic of Capricorn and the equator are from the FIG. 55. DIRECTIONS OF PRIMARY AIR CURRENTS (AFTER FERREL). southeast, and those to the poleward of the tropic are from the northwest. The upper air has a general movement from the west and northwest, while for middle altitudes the movement is from the pole equatorward. In the polar regions this counter or return current reaches to the ground. Schematic Diagram of the General Atmospheric Circula- tion. The accompanying diagram (Fig. 55) shows the main features of the general atmospheric circulation. On GENERAL CIRCULATION OF THE ATMOSPHERE. 189 the shaded sphere, the complete arrows show the direction of the lower air currents, and the broken arrows that of the upper currents ; the arrows flying with the wind. The outer, unshaded portion of the figure shows the circulation projected into the plane of the meridian; and the position of an isobaric surface at low altitudes is rep- resented by the inner, and that of an isobaric surface at high altitudes by the outer, encircling line. On the body of the sphere we see the eastward upper current at all latitudes, and the lower current eastward in the higher latitudes, but westward in low latitudes. The causes of these wind movements, and the connec- tion between them, have been studied out with great care ; and some of their details are now to be entered upon, after comprehending which, the general air motions shown in Fig. 55 can be more readily understood. The Cause of the General Circulation of the Atmos- phere is primarily the large but not constant difference in temperature existing between the equatorial and polar regions. These atmospheric motions embrace, generally speaking, a whole hemisphere each side of the equator. The difference in the air temperature between the poles and the equator amounts (at a few feet above the ground) to about 80 F. for the average of the entire year. Since the colder the air the denser it is, any isobaric surface is nearer the ground at the cold pole than it is at the warm equator. There exists, then, a declination of isobaric sur- faces and barometric gradients in the direction of the poles, and there is a movement of the air down these isobaric surfaces after the manner already described. This movement down the decline of the isobaric surfaces from the equator towards the pole continues until the alti- tude of these isobaric surfaces becomes so lessened at the ELEMENTARY METEOROLOGY. A equator and increased at the pole as to bring them to the same level throughout. This motion originally occurs at all altitudes where the isobaric surfaces decline towards the pole ; but it increases with the altitude, because the difference between any two upper air layers must be added to the differences which al- ready existed for the layers below. In Fig. 56, D is the equator, CC are the poles, and BAB an isobaric surface. The arrows show the direction of the flow of air. As a result of these air currents towards the poles, there arise (as is explained later) other nearly horizontal currents towards the equator, and upward and downward currents connecting them (see Fig- 57)- It has been com- puted that if the average ^_ ^ temperature of an air layer 33,000 feet in thickness, at latitude 30, were 40 F., and if the temperature at the pole were 90 F. less than this, then the upper isobaric sur- face of this layer would be depressed so much that it would be nearly 6,000 feet lower at the pole than at latitude 30 ; and a mass of air gliding down this inclined surface would attain at the pole a velocity of over 600 feet per second, if it started at latitude 30 from a position of rest, and its motion were frictionless and undisturbed, and under the influence of gravity alone. At the earth's surface the observed greatest pressure is at about lati- tude 30, and the calculated depression of the isobaric surface is 260 feet at the equator, and in the southern hemisphere at the pole it is flbout 650 feet ; while in the northern hemisphere the maximum de- FIG. 57. IDEAL MERIDIONAL AIR CIRCULATION. GENERAL CIRCULATION OF THE ATMOSPHERE, 191 pression is only about 325 feet, and is at latitude 60 ; from thence to the north pole the depression lessens. On the basis of observed air pressures at sea level, it has been com- puted that at an altitude of 33,000 feet the depression of the isobaric surface amounts to over 2,800 feet at the north pole, and 3,150 feet at the south pole. Diurnal Rotation of the Earth. In addition to the baro- metric gradient due to the difference in the temperature at the equator and the pole, there is another influence power- fully affecting the general air motions, and this is the diurnal rotation of the earth on its axis. In considering the matter of air motions, however, it is best first to examine the conditions resulting primarily from the differ- ences of temperature alone, and then to add the modifying influences of the earth's rotation. Ideal General Circulation of the Air without Rotation of the Earth. If the motion of rotation of the earth on its axis is not considered, then the movement of a particle of air is either north and south, or upward and downward, in the plane of its meridian. Starting with the high temperature at the equator, and the low temperature at the pole, and the resulting inclina- tion of the isobaric surfaces from the equator towards the pole, we find the equatorial air flowing poleward down these surfaces in the direction of the lowest level with a velocity which increases with the altitude. The upper air, then, flows poleward most rapidly ; but at the earth's surface there is no motion, because at first there is no gradient at this lowest surface, the amount of air above being every- where the same. Just as soon, however, as any of the upper air leaves its position in lower latitudes, there is a lessening of the actual weight of the air (pressure at the surface of the earth) there, and an increase 192 ELEMENTARY METEOROLOGY. towards the pole, by reason of the air which has flowed thence; and a countercurrent of air 'sets in along or near the surface of the earth from the region of the pole towards the equator, to take the place of the air which moves away above from the latter region. We thus have a current of air above flowing poleward, and a current below flowing equatorward ; and in order to satisfy the conditions of continuity, which allows no gaps to occur in flowing fluids, there must exist vertical currents at the ends of these two horizontal currents, in order to complete the circuit by connecting them. The direction of the upward and downward connecting currents is conditioned by the direction of flow of the main horizontal currents, for they must be such as to keep up a connected flow of air. Thus an upward current at the equator and a downward current at the pole are needed to meet this requirement. The circuit is shown in the diagram (Fig. 57). Since the temperature differences at the poles and the equator are perpetual, and the force (gravity) which causes the poleward motion is a continuous one, this circuit of the air current is uninterrupted in action after it once begins. And, moreover, since the cause of the motion acts continuously, there would be a continually increasing velocity in this circuit, if it were not for the loss of motion through friction, and the interference or mixing of air masses having different directions and velocities of motion. Regions of Separation between the Horizontal and Verti- cal Currents, respectively. It is evident that there must be some region between the upper poleward air current and the lower current towards the equator, in which there is no horizontal motion one way or the other ; and likewise between the upward and downward currents at the equator GENERAL CIRCULATION OF THE ATMOSPHERE. 193 and the pole there must be a region where there is no vertical motion, and where the motion changes from one direction to the other. These regions of no motion are called neutral planes (see Fig. 58). While at the neutral plane between the upper and lower air currents there is no horizontal motion in either direc- tion, yet, with increase in altitude above this plane there is a gradual increase in the poleward velocities, and below this plane there is an increase in the equatorward veloci- ties down to within a few hundred feet of the ground, where the friction of the earth retards the air motion, and makes it decrease again with nearer approach to the Equator TIG. 58. REGIONS OF No AIR MOTIONS. ground. Likewise at the neutral plane between the verti- cal air currents at the pole and the equator, there is an increase of the downward motion towards the pole, and an increase of the upward motion towards the equator. At the intersection of these neutral planes there is no motion in any direction, and consequently a calm. Effect of the Convergence of Meridians. We have been picturing the motion along the meridian. Now, it is neces- sary to take into account the fact that the meridians con- verge more and more as the equator is departed from, and meet in a point at the pole. The effect of this conver- gence is to gradually decrease (with the decreasing size of the parallels of latitude) the amount of space (between the different meridians) which can be occupied by the air above, 194 ELEMENTARY METEOROLOGY. which flows poleward ; and to gradually increase (with the increasing size of the parallels of latitude) the amount of available space which can be occupied by the air below, which is flowing from the polar regions equatorward. Since, with assumed constant velocity of motion, about half of the air must be to the poleward, and half to the equatorward, of the vertical neutral plane, the effect of the convergence of the meridians is to force the vertical neutral plane equatorward ; and it must lie somewhere about latitude 30, be- cause this divides the surface of a hemisphere f ) (and consequently the cubic space between the earth's surface and FIG. 59. REGIONS OF AIR MOTIONS AND CALMS the OUter limit of the ALONG A MERIDIAN. air)into two equal parts. This affects both the velocities and paths of moving air masses. Meridional Movement of an Air Mass. If we follow the motion of a mass of air starting out from a position near the equator but at a considerable altitude above the earth's surface, and above the level of the horizontal neutral plane (Fig. 59), we shall find that the horizontal motion of the air mass as it moves towards the pole is accelerated for a time until it reaches some intermediate latitude, and then it is retarded as it approaches its polar limit ; and with this re- tardation there is a gradual downward fall of the air until it reaches the point where it is nearest the pole, which occurs at the altitude of the horizontal neutral plane. GENERAL CIRCULATION OF THE ATMOSPHERE. 195 The air ma&s now commences its return journey in the lower current towards the equator : there is at first a con- tinued downward motion, and then it becomes more nearly horizontal and continues so until after the neutral vertical plane is passed; there is an acceleration of velocity until an intermediate latitude is reached, after which the horizontal motion is retarded, and there is a gradual rise of the air mass until it reaches its place of starting. The circuit made by the air masses is somewhat elliptical (oval) in form ; and it is seen that some of the air masses move in large orbits which extend from the equator almost to the pole, while others move in smaller orbits, some of which extend only a short distance each side of the inter- section of the two neutral planes. All these motions are along the meridians. The horizontal current reaches its greatest velocity in middle lati- tudes, and vanishes at the equator and the poles. The vertical current disappears at the earth's surface and at the outer limits of the atmos- phere. The descending velocity near the pole is less than the ascending velocity near the equator, because the downward current meets with more resistance than the upward current. The downward-moving air meets air of increasing density, and the surface of the earth offers resistance to its motion. The upward current carries the air into masses of less den- sity, and there is no resistance from a rigid body such as the earth's surface. The velocities of the vertical currents are relatively small, but their great meridional extent magnifies the importance of their action. The relation of the horizontal flow of air to the vertical is, roughly speaking, as the radius of the earth to the height of the atmosphere, because, while the greatest length of the horizontal current is from the equator to the pole, that of the vertical current is from the earth's surface to the upper limit of the atmosphere. The Effect of the Earth's Rotation, on the air currents passing between the equator and the poles along meridians, and on the vertical air currents which connect these hori- 196 ELEMENTARY METEOROLOGY. zontal currents, is a matter which has not yet been inves- tigated in all of its details, and it is so difficult that only a meager outline of it can be presented here. There is one general theorem on which are based all modern explana- tions of this action, and this is as follows : If a free-moving particle (such as air) moves along near the earth's surface, there is a force arising from the diurnal rotation of the earth which deflects it to the right of its course in tJie northern hemisphere, and to the left of its coiirse in the southern hemisphere. The amount of this force in- creases with the mass and the velocity of the particle, and also with the increase of latitude. It is zero at the equator, and greatest at the poles. 1 The following illustration will show the general nature of this deflecting force : In Fig. 60 we have a view of the northern hemisphere as it is seen from above the north pole. The direction of axial rotation is that shown FIG. 60. DEFLECTION OF THE WINDS BY EARTH'S by the curved arrows along the ROTATION. outer circle representing the equator. This direction is opposite to that of the hands of a watch. Suppose that at A there is a wind blowing along the meridian towards the north pole, as shown by the arrow a. While the meridian at A is advancing to the position B (due to the diurnal rotation of the earth), the air which was at A has been moving in a straight line, that is, in the direction indicated by the arrow a, so that this same direction 1 The formula expressing the amount of this force is 2 MVW 'sin D, where J/is the mass, Fthe velocity, W the angular rotation of the earth on its axis, and D the latitude. The meaning of this formula is too complex for explana- tion in this elementary treatise, but is fully given in Ferrel's " Winds." GENERAL CIRCULATION OF THE ATMOSPHERE. 197 when the meridian reaches the position B will be indicated by the arrow b parallel to the arrow a. But it is seen that this direction of the arrow b is to the right of the meridian A in this new position B. Similarly, when the meridian A has assumed the positions C and D, the air will have a motion in the directions shown by the arrows c and d. Another set of arrows on the right hand shows the effect of the deflection (due to the earth's rotation) on air motions towards the west ; another set of arrows at the top shows the effect on motions from the pole towards the equator ; and still another set of arrows on the left shows the effect on motions towards the east. A revolving terrestrial globe with a fixed brass meridional scale gives a still clearer idea of this matter if one takes a pencil in hand as if for writing, rests the hand, by means of the end of the little finger, on the globe, and then slowly revolves the globe from west to east, letting the hand move with it, but at the same time giving a slight move- ment to the pencil point (as in writing), and keeps the latter moving in a direction either perpendicular or parallel to the fixed brass meridian scale. The path traced out by the pencil point will show the direction of deviation which the moving air would take for an east-westerly or north-southerly initial direction. Any other directions than the four cardinal ones mentioned can be followed out by making the path of the pencil always keep a fixed angle with the brass meridian scale. The amount of the influence of the rotation of the earth is best shown by an example. At latitude 50, a rifle ball moving 1,700 feet per second, discharged at a target 3,300 feet distant, would deviate about 4 inches to the right of the target in the northern hemisphere, and the same amount to the left of the target in the southern hemisphere. This effect may seem slight ; but when the force operates on masses of air moving day after day, and over distances of thousands of miles, its results are of great importance. Effect of the Deflecting Force on the Polar and Equatorial Air Currents. The qualitative effect of this force on the horizontal meridional motions which have been described, is to cause the free-moving air currents to depart to the right of their course in the northern hemisphere, and to the left in the southern hemisphere. The upper poleward currents are, then, deflected towards WALDO METEOR. 12 198 ELEMENTARY METEOROLOGY the east, and tne lower currents flowing towards the equa- tor are deflected towards the west, in both hemispheres. We thus see that there will be opposite directions of mo- tions high in the air and near the earth's surface, not only in a north-and-south direction, but also in an east-and- west direction (the upper horizontal current tending towards the northeast, and the lower one towards the southwest, in the northern hemisphere) ; and the friction, and the intermingling of the air which arises at the boundary of these oppositely directed motions, are very important in their effects. There results then a gyratory motion around the pole, toward it aloft, and away from it below ; and, from the law of conservation of areas, the ve- locities aloft tend to increase, and those below to decrease. Easterly Motion of the Air in High (North) Latitudes. - Since the effects of this deflecting force accumulate as long as the current continues, therefore in northern latitudes the motion to the right, or easterly motion, of the upper poleward current, becomes very great in comparison with the westerly motion at the beginning of the equatorial current beneath it. So that this upper easterly motion, by friction, and especially by intermingling of the air, gradually entirely overpowers the lower westerly motion, and a general easterly motion results in both the upper and lower layers of the atmosphere in these latitudes. This general eastward air movement, it is found, actually takes place around the pole. Westerly Motion of Air in Low (North) Latitudes. In passing towards the equator, however, the lower equatorial current gains strength in its endeavors to flow towards the west, in obedience to the force carrying it towards the right; and after reaching a certain latitude it is able, through friction and the intermingling of air in the cur- rents, to just neutralize the (to that point overpowering) GENERAL CIRCULATION OF THE ATMOSPHERE. 199 effects of the upper eastward current, so that there will be no east or west motion in the lower air layers. Passing still farther south, however, the effects of the force towards the west increase, and a westerly air current arises at low altitudes ; and as lower latitudes are reached, the power of this western current becomes relatively so great as to over- come the upper easterly current, which is weak there ; and in very low latitudes a general westerly motion of the air takes place at all altitudes. So that very near the equator there is a wind directed a little to the north of west in the upper, and a little to the south of west in the lower, air layers. The Limits of Velocities of Air Currents. -^ Since the deflecting effect of the earth's rotation is zero at the equator and increases with the latitude, then the poleward upper current is more and more deflected towards the east as it moves northward ; and since the cause of the original motion acts along the meridians from the equator towards the north, then the deflection towards the right, due to the earth's rotation, must be directed partly against this movement, and must have a retarding influence on it. This deflecting force increases with the increase in velocity ; and it is evident that if the velocities in- creased enough, there would be a deflecting force towards the right (in this case towards the equator) sufficient to counterbalance the force producing the original poleward current, and this motion would cease. We thus see that there is a practical limit to the velocities which the polar current may attain ; for, if these increase too much, the deflection to the right will also increase to such an extent as to actually over- come the poleward current in its course. This acts like a controlling governor on the air circulation. A second control over excessive air motions is the fact that when a current is impelled by a constantly acting force, the current is contin- ually breaking up into relatively small atmospheric whirls (analogous to the whirls in flowing water) which impede the regular progress of the current by mixing up the various air masses. A third impediment to excessive velocities is the frequent occurrence of vertical currents in the air. 2OO ELEMENTARY METEOROLOGY. Vertical Air Movements in the General Circulation. The vertical currents in the general atmospheric circulation those which connect the horizontal currents are the descending current in polar latitudes, and the ascending current in the neighborhood of the equator. Two addi- tional vertical currents shown in Fig. 55 at about latitude 30 are explained on p. 203 in connection with the effects of the air-pressure distribution on the general circulation of the air. Downward Current at High Latitudes. The upper air, as it approaches the pole, begins to descend, or settle down towards the earth, and the comparatively great easterly motion which the air has in this region is mostly used up in overcoming the friction between the successive air layers as the air makes its way downward ; but aloft there is a sufficiently powerful motion towards the east to give an easterly tendency to the air even at the surface of the earth in this region. Moreover, when the air descends in the higher latitudes, it approaches a little nearer to the axis of rotation of the earth ; and its velocity becomes slightly increased, according to the conservation of areas. Upward Current at Low Latitudes. The upward cur- rent, which extends to a considerable distance from the equator poleward, must in its lower course have a motion towards the west imparted to it by the lower horizontal air current flowing towards the equator. This west component of motion decreases with the altitude up to a certain height above the earth's surface, where it changes into an east- ward motion ; for at high altitudes there is an easterly cur- rent over, the whole region from the equator to the pole, except, perhaps, directly over the equator, where the westerly motion extends up to very high altitudes. The altitude where the change of direction takes place de- GENERAL CIRCULATION OF THE ATMOSPHERE. 2OI creases up to about latitude 30, beyond which the wind becomes generally easterly (towards the east). Effect of Air Motions on Barometric Pressure. The effect of these general motions of the atmosphere on the distribution of the air pressure and the isobaric surfaces is of great importance. If the air temperature were everywhere the same, then the isobaric surfaces would be level. Now, we have seen that by heating the air at the equator the successive isobaric surfaces are ele- vated there ; but directly at the surface of the earth there is no elevation of the isobaric surface, and there- fore there is no pressure gradient at the earth's surface such as exists at various altitudes above it. But when there are easterly or westerly air currents at (or close to) the earth's surface, this uniformity of the isobaric sur- face at the earth's surface no longer exists, but there arise surface gradients ; and where there are no east or west motions, there are no gradients at the earth's surface. The Barometric Gradients at the Earth's Surface are accounted for by the action of the centrifugal force which arises from the eastward (or the westward) movement of the air, this movement being equivalent to a whirling mo- tion around the pole as a center (or, in more local whirls, around some other point as a center). Whenever this whirling motion occurs, the particles so moved tend to leave their curvilinear track, owing to the centrifugal force. In the case of a whirling mass of water such as is to be seen in any whirlpool, the water particles are free to move, and there is a heaping-up of the water around but away from the center of the whirl, and a de- crease of the water towards the center of the whirl. The water recedes from the center to a distance such that the centrifugal force causing it to move away is just equal to 202 ELEMENTARY METEOROLOGY. the downward push, or gradient force, of the heaped-up water towards the center of the whirl. In the case of the air motion around the north pole, which is the center of the whirl, the current towards the east, in the middle and northern latitudes, gives rise to the action of centrifugal force, which causes a heaping-up of the air at the latitudes away from the pole, and consequently to the right and to the equatorward of the easterly current, and a depletion or lessening of the air on the inner, the left or polar, side of the current. Thus we have a lessening pressure towards the pole, just as there is towards the center of a water whirlpool. The air on the equatorial side is thus thrown off by the centrifugal force, and distributes itself over the equatorial regions of the northern hemisphere, where it meets with a corresponding heaping-up of the air coming from the polar whirl of the southern hemisphere. Barometric Pressure Gradients at Various Altitudes. Up at considerable altitudes above the earth's surface, this heaping-up of the air and increase of pressure take place clear to the equator, as just mentioned ; so that at high altitudes there is a maximum pressure at the equator, and a continuous decrease from thence to the pole. But the maximum pressure at the earth's surface is found at about latitude 30, where the easterly motion ceases, and there is no east or west motion. When, however, the westward motion sets in, as we pass from this point towards the equator, then there is a slight decrease in the pressure, due to the opposing centrifugal effects of the westerly wind. It is not so much a real decrease in the pressure as it is a lessening of the pressure which would exist if the easterly wind alone existed all the way to the equator. This is brought out more plainly in the variation of the pressure GENERAL CIRCULATION OF THE ATMOSPHERE. 203 with the latitude in the southern hemisphere (see t table, p. 96), where the disturbing influence of the land is not so markedly felt. The maximum pressure, then, is to be found at about latitude 30 at the earth's surface, but with increasing altitude it lies nearer and nearer to the equator. Up above this dividing line of maximum pressure there is a motion from the west, probably all the way to the equator, and over the lower westerly motion which exists between the equator and about latitude 30. This allows, then, a motion towards the east up at high altitudes in all latitudes. Cause of the Equatorial and Poleward Surface Winds in Lower Middle Latitudes. In the meridional projection of the general air circulation (Fig. 55) the poleward motion is represented above, with the equatorward motion along the surface below it ; but there is still a peculiarity which must be mentioned, and which can be best illus- trated by the diagram. It is seen that (on the meridi- onal section) at about latitude 30 there is a current along and near the surface of the earth moving towards the pole, which is contrary to what might be expected ; and another towards the equator, although . this is in the direction to be expected. These currents have as their cause the heaping-up of the air shown by the ring of high pressure at latitude 30. The atmosphere here exerts a greater downward pressure than that either side of it, and so forces out the air from the lower layers beneath, in the attempt at equalizing the pressure. Part of the air from this lower layer is then forced poleward, and part equatorward. These currents are below the great equa- torial return current, and are fed by a downward vertical current in the lower air layers : they are indicated in Fig. 55 at the latitudes marked " Tropical Calm and Dry Belt.'* 2O4 ELEMENTARY METEOROLOGY. The Air Circulation, viewed as a Whole, can be considered as consisting of two huge atmospheric whirls with the poles as centers (Fig. 61). The direction of rotation of these whirls is determined by the rotation of the earth on its axis ; in the northern hemisphere it is opposite to the movement of the hands of a watch (face upwards), and in the southern hemisphere the direction is the same as that of the hands of a watch. At the outer edge of each of these whirls (that is, between the tropical region and the equator) there is an atmospheric ring or belt, Equator Equator Lower Currents Upper Currents FIG. 61. GENERAL HEMISPHERICAL DIRECTION OF HORIZONTAL AIR CURRENTS. decreasing in latitudinal width with increase of altitude above the earth's surface, in which the motion of rotation is in the opposite direction to that of the respective whirls. There is an outward flow of air, at middle altitudes, from the polar centers of these whirls. In the transition zone between this inner (polar) region and outer (equatorial) ring with opposite directions of rotation, there is a heap- ing-up of the air, and a consequent increase of the air pressure, due to the centrifugal force of these motions. The moving air obeys the law of the conservation of areas (see p. 182). GENERAL CIRCULATION OF THE ATMOSPHERE. 2O5 An Interchange of Air between the Northern and Southern Hemispheres also takes place, owing to the transference of the region of greatest temperature, first to one side, and then to the other side, of the equator, with the annual movement of the sun. When, for instance, we have our winter in the northern hemisphere, and summer exists in the southern hemisphere, then the isobaric surfaces are more elevated by the intenser heat and subsequent expan- sion of the air over the equatorial regions in the southern hemisphere than in the northern hemisphere, and air flows down these surfaces from the southern into the northern hemisphere. Similarly, when we have our summer in the northern hemisphere, there is a flow of air across the equa- tor from the northern into the southern hemisphere. Inequality in Air Motions Due to Variations in Tempera- ture. The difference in temperature between the polar and equatorial regions is much greater in the winter time than in the summer time, and consequently the slopes of the isobaric surfaces are greater ; and since the velocities of the air currents depend on the gradient or slope of these surfaces, these velocities are much greater in win- ter than in summer. From the observed data of wind velocities it is seen that the winter velocities are in many cases double those of the summer. Theoretical Computation of Easterly and Westerly Velocities of the Wind. Attempts have been made to compute theoretically the east- erly and westerly velocities of the wind at various altitudes on the different parallels of latitude, by means of the gradient forces result- ing from the inclination of the isobaric surfaces due to differences in temperature and air pressure between the equator and the poles along an average meridian. These velocities in miles per hour, for the yearly average, are given on p. 206. It is seen from this table that the westerly motions disappear at a height of about 10,000 feet above sea level to the poleward of about 206 ELEMENTARY METEOROLOGY. COMPUTED EASTERLY OR WESTERLY WIND VELOCITIES ALONG A MERIDIAN. LATITUDE. EASTERLY (E.) OR WESTERLY (W.) VELOCITY OF WIND IN MILES PER HOUR, AT VARIOUS ALTITUDES. INCREASE IN EASTERLY VELOCI- TIES WITH EACH (ABOUT) 3,300 FEET IN ALTITUDE. Sea level. About 3,300 feet. About 13,200 feet. Miles per hour. N. Lat. 75 W. 2.7 E. 0.2 E. 9.2 E. + 3- 70 W. 2.0 E. 2.0 E. 14.3 E. 4.1 65 E. o.i E. 4.9 E. 19.3 E. 4.8 60 E. 2.4 E. 7.6 E. 23.1 E. 5.2 55 E. 3.4 E. 8.7 E. 24.5 E. 5-3 50 E. 3-3 E. 8.7 E. 24.9 E. 54 45 E. 3.0 E. 8.5 E. 25.0 E. 5-5 40 E. 1.6 E. 7.2 E. 24.0 E. 5.6 35 W. 0.7 E. 5.0 E. 22.4 E. 5.8 30 w. 5-3 E. 0.6 E. 18.2 E. 5-9 25 W. 8.9 W. 3.1 E. 14.4 E. 5.8 20 W. 9-4 W. 3.8 E. 13.0 E. 5.6 N. Lat. 15 W. 7.8 W. 4.3 E. 6.1 E. 3-5 Equator S. Lat. 15 W. 15.6 W. 10.5 E. 4.8 E. 5.1 20 W. 13.0 W. 8.2 E. 6.4 E. 4.8 25 W. 6.4 W. 1.7 E. 12.5 E. 4-7 30 E. 2.4 E. 7.0 E. 21.0 E. 4.7 35 E. 7.7 E. 12.3 E. 26.1 E. 4.6 40 E. 1 1. 6 E. 1 6. 2 E. 30.0 E. 4.6 45 E. 14.9 E. 19.5 E. 33-3 E. 4-6 50 E. 17.1 E. 21.7 E. 35-7 E. 4.6 55 E. 17.0 E. 21.6 E. 35-6 E. 4-7 S. Lat. 60 E. 13.6 E. 18.2 E. 32.2 E.+4-7 latitude 15. And we see here, derived by theory, the observed fact of the easterly wind velocity near the earth's surface in the northern hemisphere reaching a maximum at about latitude 50 or 60 ; but it is to be noticed that this region of greatest velocity decreases in latitude with increase of altitude. GENERAL CIRCULATION OF THE ATMOSPHERE. 207 Observed Circulation of the Atmosphere. Usually the arrows showing the direction of the winds on charts are drawn partly from direct observation, and partly from the known laws of the direction of wind as depending on the distribution of the air pressure. Such are the indications of wind directions given on the charts of air pressure for the globe (see Figs. 26, 27). The lower air currents are difficult to observe on the continents, owing to the irregularities of the ground, and so we must turn to the observations made on the ocean for the best obtainable presentation of their characteristics. Published charts showing the general wind direc- tion for the Atlantic Ocean have been already given (see Figs. 36, 37). It is seen on these, that while the surface wind directions follow in general those deduced by theoretical reasoning and shown in the pre- vious sections, yet there are localities in which the local pressure con- ditions exert such a predominating influence that the main currents are interrupted. Thus, on the northern North Atlantic and middle South Atlantic oceans, there are local atmospheric whirls formed in the win- ter of the northern hemisphere ; and in the summer there is still another interrupting whirl formed on the middle North Atlantic. In the northern hemisphere, the observed general direc- tion of the upper air currents is from N. 45 W. to S. 45 E., in high latitudes like Lapland. In middle lati- tudes (at about latitude 40) the direction is nearly due easterly. In lower latitudes (at about latitude 25) it is perhaps from S. 75 W. to N. 75 E. In the equatorial regions, however, the direction is in general from the east towards the west, and in some cases it is from S. 45 E. to N. 45 W. ; but there is a much greater irregularity of direction in this region than near the poles, on account of the shifting of the region of greatest heat from one hemisphere to another, There exist compara- tively few wind observations, on the land, for the inter- tropical regions. 208 ELEMENTARY METEOROLOGY. Trade Winds. These are the lower winds belonging to that part of the general atmospheric circulation embraced between the equatorial region, of calms and the tropical belt of high atmospheric pressure. They blow from the northeast in the northern hemisphere, and from the south- east in the southern hemisphere. The general location and direction of these winds are shown on the diagram of the general circulation of the atmosphere (see Fig. 55), where they are indicated by heavy arrows between the Tropical and Equatorial Calm Belts. They have not very great velocities, but they are relatively very constant in their continuance and in their velocity of motion. The whole system of trade winds shifts from north to south, and the reverse, following the sun in its course ; and it also varies somewhat in width, not only with the longi- tude, but also with the season of the year, the latter being shown by the following table : LATITUDINAL LIMITS OF THE TRADE WINDS. AVERAGE OVER THE PACIFIC AND ATLANTIC OCEANS. IN MARCH. IN SEPTEMBER. In northern hemisphere "t (northeast trades) / In southern hemisphere 1 (southeast trades) / Lat. 4 N. to 25 X. Lat. 1 N. to 26 S. Lat. 10 N. to 33 N. Lat. 5 N. to 22 S. Doldrums. Between the surface winds of the two hemispheres there is an equatorial region of calms, called the doldrums. This region lies a little to the north of the terrestrial equator ; but it varies somewhat in locality and width, not only regularly with the season of the year, but also irregularly with the longitude. In March the dol- GENERAL CIRCULATION OF THE ATMOSPHERE. 2OQ drums extend from o to 3 north latitude on the Atlantic Ocean, and from 3 to 5 north latitude on the Pacific Ocean; in September they extend from 3 to n north latitude on the Atlantic Ocean, and from 7 to 10 north latitude on the Pacific Ocean. Monsoons, or continental land and sea winds, occur dur- ing the hot and cold seasons of the year. During the sum- mer the interiors of the continents become heated, and this elevates locally the surfaces of equal air pressure, and the air flows down these towards the ocean at some distance above the ground; and a countercurrent near the ground sets in from the ocean towards the land. In the winter time the opposite interchange of air takes place, and along the ground the air flows from the land towards the sea. In case of an uninterrupted low-level region extending from the ocean inland, the monsoon effects are not so marked as when the inte- rior is high or mountainous. One reason for this is, that where there is a long surface slope the conditions are best for an extended general movement of the air. In winter the lower air will flow downward much easier than on a level ; and in summer the warm air flows upward along the slopes more easily than along a level surface, just as it will rise better in a somewhat vertical flue than move in a horizontal flue ; and there is less tendency for local currents to form. Another reason is the fact that the temperatures on elevated land, especially on plateaus, are greater in summer and less in winter than at the same altitudes above the ocean or the low land ; and there is .thus a contrast in the temperatures maintained to a height depending on the altitudes attained by the land. This principle is illustrated by the greater air circulation (or draught) in the case of a tall chimney as compared with a short one. This effect is felt mostly in the summer monsoon, as in winter the air temperature of the land is not so much below that of the surrounding air as it is above in summer. While the monsoon winds occur in many regions, yet nowhere else are the conditions so favorable as in India and the north Indian Ocean. There the low lands are backed by the Himalaya mountain range, to the north of which Ires the Kuenlun range, with the plateau of Tibet (210) (211) 212 ELEMENTARY METEOROLOGY. between ; and still farther to the north lie vast stretches of quite high plateau desert land. During the summer monsoon the region of great surface heat of the Indian Ocean and the adjacent low lands, which consequently affords the air an excessive capacity for moisture, lies close to the elevated region just described ; and when the air moves u.p the steep slopes, as a result of the adiabatic cooling excessive pre- cipitation takes place, and enormous quantities of freed latent heat increase the temperature of the air over the continent above what it would be in the case of no precipitation, and the wind velocity is thereby increased. Strongly developed monsoon winds so act that they greatly influence, and may even reverse, the direction of the regular primary wind circulation. The accompanying figures (Figs. 62, 63) show the very powerful monsoon winds of the north Indian Ocean and southern Asia in midwinter and midsummer. Heavy arrows indicate strong winds ; light arrows, light winds ; and circles, calms. Regions of variable winds are denoted by short arrows, while the long arrows designate steady winds. In January and February the outflow of air from the interior of the Asiatic continent is southward across the north Indian Ocean 4:o 10 of latitude beyond the equator. In July and August there is an inflow of air towards the interior of the Asiatic continent, not only extending to the equatorial region of the Indian Ocean, but also form- ing a continuation of the winds of the lower latitudes of the southern hemisphere (to 30 south latitude) which flow in towards the equator. This continuous northward sweep of the winds over the ocean from 30 south latitude to over 20 north latitude, caused by the great indraught over the heated continent, readily accounts for the excessive force of the southern monsoon winds of midsummer on the Indian seas and mainland. CHAPTER IX. SECONDARY CIRCULATION OF THE ATMOSPHERE. WE must next examine those more local conditions of the atmospheric circulation which interrupt the general circulation, and occasion the variability in the winds during brief periods of time. We will take up those conditions which are most closely connected with, and to a great extent depend on, the cur- rents of the general air circulation. Such are the local atmospheric whirls which occur irregularly in the greater hemispherical currents, and which appear to gradually form in the greater current, follow it in its general course for a time, develop in intensity to maturity, and then gradually disappear. Movement of Air in Whirls. It is a marked charac- teristic of atmospheric motion that it occurs mainly in the form of whirls or vortex motion. Even when the wind blows in an apparently straight line, if it is followed out far enough, it will usually be found to belong to some system of whirling motions. Thus the great air currents forming the general circulation of the atmosphere have been found to be but a part of the great hemispherical vortical movement described in the preceding chapter. Secondary Atmospheric Whirls. Within these mighty air currents of the hemispherical air circulation there exist limited systems of atmospheric whirls in which, in some cases, the air moves spirally inward towards a center, and WALDO METEOR. 13 21^ 214 ELEMENTARY METEOROLOGY. in others spirally outward from a center. In such local whirls there exist gradients of air pressure such as occur in the great hemispherical whirl : and these may be arranged in two ways, first, there may be an area of lower atmospheric pressure at the center of the whirl, in which case the air movement is spirally inward, and such a system is called a cyclone ; second, there may be a region of higher atmospheric pressure at the center, in which case the spiral air movement is outward from the central area, and such a system is called an anticyclone (Fig. 64). FIG. 64. DIRECTION OF WHIRL FOR A CYCLONE AND AN ANTICYCLONE (NORTHERN HEMISPHERE). Near the center of a cyclone there is an upward movement of the air, while in an anticyclone there is a central down- ward movement. Cyclonic conditions may be characterized thus : a cen- tral region of low air pressure, with a gradual increase from thence toward the outer limits ; a spiral motion of the lower air inward toward the center ; and in our middle lati- tudes, on the eastern and southern sides of the center of cyclones, a warm, cloudy, more or less rainy condition, with strong winds from the east and south, and on the SECONDARY CIRCULATION OF THE ATMOSPHERE. 21$ western and northern sides a cool, clear, or clearing condi- tion, with strong winds from the west and north. Anticyclonic conditions may be characterized thus : a central region of high air pressure, with a gradual de- crease from thence to the outer limits ; a spiral motion of the lower winds outward from the center; and in our middle latitudes, on the eastern and southern sides of the region of highest air pressure, a cool, clear, or clearing condition, with winds from the north and west, and on the western side a warm condition, with increasing cloud, and winds from the south and east. These whirls have in our latitudes a motion of transla- tion of about 30 miles per hour, usually, in an easterly direction ; and as they pass over any place lying in their path, first the conditions on the eastern or front side are experienced, and then those on the western or rear side. A , cyclone or an anticyclone may sometimes assume such large proportions as to cover half the United States, and may exist for several days, and cross the entire country. The direction of the inward spiral movement of the air circulating around an extended cyclone is, in the northern hemisphere, always opposite to that of the hands of a watch ; and in the southern hemisphere it is with the hands of a watch. The direction of the outward spiral movement of the air circulating around an anticyclone is always with that of the hands of a watch in the northern hemisphere, and opposite in the southern hemisphere. Whether the air is flowing down the isobaric surfaces inward towards the center of low barometric pressure of the cyclone, or outward away from the center of high barometric pressure of an anticyclone, the rotation of the earth on its axis causes the current of air to depart to 2l6 ELEMENTARY METEOROLOGY. the right in the northern hemisphere, and to the left in the southern hemisphere. This gives the distinctive direc- tions to the whirl of the air masses in cyclones and anti- cyclones. These cyclones and anticyclones are sometimes of so great extent and they are so constantly undergoing changes, that, in order to study such in their true relations, it is necessary to obtain simultaneous observations of the meteorological elements over extensive regions, such as continents, oceans, or even a whole hemisphere. CYCLONES. Classes of Cyclones. The term cyclone is applied by meteorologists to all kinds of atmospheric disturbances in which the air pressure decreases, and there is a wind movement, inward towards the center. They are not, however, all alike in their other characteristics. There are the cyclones of the torrid zone, which in their greatest development are called hurricanes and typhoons. These atmospheric whirls vary from a few miles to several hundred miles in diameter, and are characterized by great violence of wind and copious rainfall. Cyclones of the temperate and colder zones are of greater extent, usually covering a region at least several hundred miles, but sometimes a couple of thousand miles, in diameter. These are of less intensity than the cyclones of the torrid zone, and are the cyclonic areas, or areas of low barometer, usually spoken of in our government weather reports. The latter cyclones are frequently accompanied by secondary cyclones of limited extent but intense energy, manifested by strong winds. These are called tornadoes. SECONDARY CIRCULATION OF THE ATMOSPHERE. 217 The cyclones of the torrid zone sometimes pass north into the temperate zone, where they gradually lose their former characteristics of relatively limited extent and great energy, and spread out and assume the main character- istics of the cyclones of the temperate zone. Characteristics of Cyclones. We do not as yet under- stand to a certainty the process of formation of cyclones, and so shall at first describe their characteristics, and afterward mention their probable causes. The main features of the cyclonic areas which must be studied are the distribution of air pressure, air tempera- ture, moisture, cloud, and rainfall ; the direction and veloci- ties of air currents or winds ; and finally the direction and velocity of propagation of the cyclonic areas, and the paths usually pursued by them. Distribution of Air Pressure in Cyclones. In cyclones, at the earth's surface and in the lower layers of the atmosphere there is a central area of low air pressure from which there is an increase of the pressure in all directions (see within the dotted circle, Fig. 65). This is best represented by means of isobaric lines drawn for equal intervals of barometric pressure (after the observed air pressures have been reduced to seq. level). Such iso- baric lines show where the isobaric surfaces intersect this level. The isobars in the simplest ideal case are concentric circles, the inner one of which has the low- est pressure, the outer ones having successively increas- ing pressures ; but the isobaric lines seldom present these regular features, and in nearly every case they are elongated. The ratio of the longer axis to the shorter may be mqre than 4 to i ; on the continent of America it averages about 2 to I ; while on the Atlantic Ocean the average is only 1.7 to i. 218 ELEMENTARY METEOROLOGY. The longest axis in our middle latitudes usually lies in a direction from southwest to northeast, and its average direction for a great many cyclones in the United States is N. 36 E. The barometric gradient, or rate of chang" FIG. 65. AIR PRESSURE AND AIR MOTION WITHIN A CYCLONIC AREA (AFTER FERREL). of air pressure, is steeper or more rapid, and the isobars lie closer together, over the ocean than over the land. Where the isobars are considerably elongated, that is, are very elliptical, there are sometimes two or more dis- tinct centers of low pressure within the cyclone ; but they OF SECONDARY CIRCULATION OF THE ATMOSPHERE. 2IQ differ only slightly from each other in pressure, and fre- quently have the same pressure. The increase of air pressure from the center of cyclones to the extreme outer regions is not a wholly uninterrupted ascent ; for at some distance from the center there exists a ring of higher air pressure, with anticyclonic characteristics (see Fig. 65), which is especially noticeable in cyclones of great energy, such as the hurricanes of the torrid zone; and in cyclones of great extent it lies in the region of regu- lar anticyclones. This ring of high pressure is similar to that at about latitude 30 in the hemispherical whirl. The barometric pressure at the center of the cyclone sometimes descends as low as 27.5 incheSo Fig. 68, a, on p. 222, shows in an ideal cyclone the isobars at sea level and at an altitude of several miles. Distribution of Temperature in Cyclones. The rear and left-hand sides, facing the direction of translation (western and northern sides of our cyclones), are the coldest; and the front and right-hand (southern and eastern) sides are the warmest. But where the winds blow from over a water surface, these conditions may be modified consid- erably, since the air would then be warmer in winter and cooler in summer than that from over a land surface. In the central and eastern United States there is some- times a difference of over 70 F. in the temperatures on the warmer and colder sides of cyclones; and the isotherms which in their normal condition run east and west will become so changed as to run north and south. The following diagram (Fig. 66) shows the isotherms in a cyclone of, say, 500 miles in diameter, as indicated by the circle, along which also the direction of the wind is showp by means of arrows* 220 ELEMENTARY METEOROLOGY. N d X f FIG. 66. ISOTHERMS IN A CYCLONE. C. (AFTER FERREL.) With the increase in the altitudes there is (at least over the interior of continents) a rapid decrease in the tempera- tures, because in the lower air layers the temperature is relatively high, while above it is relatively low. We need still more observations on this point, however, be- fore definite laws can be formulated. Distribution of Moisture, Cloudiness, and Rainfall in Cyclones. These elements are so intimately connected that they may be considered together. The quantity of moisture in a cyclone depends so much on its location with respect to the main source of moisture, the sea, that no general laws can be enunciated concerning it, or the cloudiness and rainfall re- sulting from it. In our latitudes the eastern, south- eastern, and southern sides of the cyclone have the greatest capacity for mois- ture, because they are the warmest ; and generally the western and northern sides are the freest from cloud. The various kinds of cloud have such differ- ent origins, and lie at such different altitudes above the earth's surface, that their individual distribution in the cyclone is difficult to give with accuracy. The cirrus > - > ' < Miles. 100 200 300 400 500 O Signifies the center of the cyclone. The other lines show the degree of cloudiness in percentage of an overcast sky. FIG. 67. CLOUD DISTRIBUTION IN CYCLONES (AFTER CLAYTON). SECONDARY CIRCULATION OF THE ATMOSPHERE. 221 clouds, however, are usually to be found on the front side of a cyclone, and extend considerably in advance of it. The relative humidity increases towards the center of the cyclone, being greatest on the eastern and southeastern sides. The distribution of cloud around the center of a cyclone in the eastern United States is shown in the preceding diagram (Fig. 67) , where the top is the northern side. The Rainfall occurs most frequently on the side of the cyclone which is nearest the supply of moisture, whence this last is carried into the region of the cyclone by the winds blowing in towards the center of the cyclone. The great downpours of rain, however, occur at the front (eastern) or warmer side of our cyclones two or three times as frequently as in the cooler (western) side. The cyclones which are accompanied by excessive rains have strongly marked characteristics, and have very low air pressures at the center ; while those which have light rainfall, or even none at all, have feebler cyclonic char- acteristics, and but slightly depressed air pressures at the center. The areas within which rain falls at any one time are usually elliptical in shape, and in general are twice as long as they are broad ; and sometimes they are as much as 2,000 miles in length. It is usually found that there is a central area of maximum rainfall, and this is encircled by successive zones of diminishing amount. In the east- ern and northern parts of the United States the area of maximum rainfall lies southeast of the center of the cyclone, and usually at a distance of about 300 miles from it; but the distance varies greatly in individual instances. Cyclones are of more or less gradual development. 222 ELEMENTARY METEOROLOGY. The greatest depth of barometric pressure does not occur at first, but is reached later ; and the maximum amount of rainfall occurs at the time of minimum pressure. Direction and Velocities of Winds in Cyclones. As the air moves spirally inward towards the center of low baro- metric pressure, it intersects the successive isobaric lines at certain angles, which vary not only on the different sides of the cyclone and at different distances from the center, but also over a land and over a water surface. Fig. 65 shows this circulation at the earth's surface, and \ > Direction of lower 'winds. Isobars at the average altitude Direction of lower clouds. of cirrus clouds. -> Direction of cirrus clouds. Surface Isotherms. QQ Central region of calms at the different altitudes. FIG. 68. ISOBARS, ISOTHERMS, AND WINDS AT VARIOUS ALTITUDES IN A CYCLONE. (Koppen.) aloft, from the outer to the inner limits of the cyclone. The wind cuts the isobars at a smaller and smaller angle as the center is approached; that is, it turns away more and more from the center as it moves inward, and finally close to the center flows nearly along the isobars. The direction of the wind circulation around a cyclone at various altitudes is shown in the accompanying diagram (Fig. 68, ), in which the arrows fly with the wind. The heavy unbroken arrows show the SECONDARY CIRCULATION OF THE ATMOSPHERE. 223 direction of air motion near the ground ; the light unbroken arrows, that at the height of the ordinary lower clouds, at perhaps an altitude of from 6,000 to 9,000 feet ; and the broken arrows, that at the height of the cirrus clouds, at an altitude of perhaps from 20,000 to 30,000 feet. The arrows in Fig. 68, b, show that in the lower air layers the direction of the wind is towards the interior of the cyclone, at a moderate altitude it is tangent to the isobars, and at a high altitude it is away from the cyclone. This relation expressed more concisely is as follows : the direction of the air movement around a cyclone in the northern hemisphere departs more and more to the right with increase of altitude. The wind velocity increases from the outer edge towards the center of a cyclone, but the maximum velocity is reached at some distance from the center (several hundred miles in some cases); and then the velocities decrease again, sometimes with great suddenness. Directly at the center there is usually a calm, or but very little horizontal motion of the air. The velocities at the place of maximum wind depend on the intensity of the cyclone and the depth and steepness of the gradients, and even near the ground frequently reach 40, 50, or 60 miles an hour. Magnitude of Cyclones. The usual diameter of cyclones over the United States is from 1,000 to 1,500 miles, while over the North Atlantic Ocean it is a few hundred miles more. The barometric gradients are about 15% steeper over the North Atlantic than over the United States, and this is perhaps due to the greater friction of the lower air layers over the land than over a water surface. The cy- clones thus develop more force over the ocean than over the continents. Direction and Velocity of Movement of Cyclones. Cy- clones, in general, move in the direction of motion of the 60 EQUATOR FIG. 69. PATHS OF NUMEROUS CYCLONES IN THE NORTHERN HEMISPHERE (AFTER LOOMIS). (224) SECONDARY CIRCULATION OF THE ATMOSPHERE. 225 great mass of air carried by the primary atmospheric currents. The preceding chart (Fig. 69) shows the paths pursued by a great many individual cyclones. The arrow-headed curved lines show at their beginning the place of formation, their length shows the path pursued, and the points of the arrows show the direction of motion and place of dissipation of individual cyclones. The direction of translation of cyclones for the middle latitudes is easterly, while for the lower latitudes it is west- erly, corresponding to the direction of the air currents in the general atmospheric circulation at those latitudes. The cyclones of our middle latitudes have an average direction of translation of about N. 80 E., or 10 north of an easterly direction ; but this varies somewhat for dif- ferent regions. In the central United States the direction is a little south of east, and in the eastern part it is a little north of east. The cyclones which are confined to our tropics have a direction of about N. 64 W. ; and the Asiatic or East Indian cyclones have a velocity of motion of about 9 miles per hour, while those of America or the West Indies move about 12 miles per hour. The cyclones which first appear in the tropics with a westerly motion, anr 3 then cross over into middle latitudes and move in an easterly direction, have in the West Indies at first an average direc- tion of N. 64 W. and a velocity of 17 miles per hour, and in middle latitudes a direction of N. 52 E. with a velocity of 20 or 21 miles per hour; while those in the East Indies have at first an average direction N. 52 W. and a velocity of 8 miles per hour, and later in the middle latitudes a direction of N. 55 E. and a velocity of 10 miles per hour. There is a tendency for these cyclones to pursue some- what the same tracks, according to the place of origination (226) SECONDARY CIRCULATION OF THE ATMOSPHERE. 22 J and the special characteristics of the individual cyclones. Sometimes two cyclones coalesce and form a single cyclone, and at other times a single one will divide up into two, and these will pursue quite distinct paths. The average tracks of observed cyclones over portions of the north- ern hemisphere are shown in Fig. 70. The velocity of translation of cyclones is nearly twice as great in winter as in summer, and seems to increase up to a middle latitude, and then decrease again. The cyclones of the United States move with an average velocity of about 30 miles per hour, which is considera- bly faster than that of cyclones of the Atlantic Ocean and Europe. The Seasons of Greatest Frequency of Cyclones. The cyclones of the West Indies and the China Sea occur most frequently in July, August, September, and October ; those of the Java Sea and the South Indian Ocean, in December, January, February, March, and April (in the latter months for the regions more distant from the equa- tor); those of the Arabian Sea and Bay of Bengal, from April to June, and again in October and November. In the middle northern latitudes the cyclones occur most frequently probably in February, and least frequently in July. Region of Maximum Number of Cyclones. The region over which the maximum number of cyclone tracks have been traced in the northern hemisphere lies along the fol- lowing course : commencing in the region south of Japan and Corea, it passes thence through Bering Sea to the southern part of Alaska ; thence along the coast almost to Oregon ; thence nearly due east across the continent to the southern part of Newfoundland; thence east-northeast over the Atlantic Ocean to the Orkney Islands, and ( 22 8) SECONDARY CIRCULATION OF THE ATMOSPHERE, 22Q farther on to northern Norway and Sweden ; thence south- east to the interior of Russia and Siberia, in which dry region they probably die out for lack of the moisture in the air which seems necessary for their long continuation. The freedom of the region within the Arctic Circle from cyclones is due partly to the low temperature and smaller amount of moisture there, and partly also to the descending air currents of that region, and to the absence of the great horizontal currents of middle latitudes. The frequency of occurrence and region of maximum number of cy- clones in various parts of the northern hemisphere are shown in Fig. 71, where the numerals signify the number of cyclone centers which have passed, during a period of 10 years, over the regions along the lines. The heavy arrow lines show the most frequented paths, Hurricanes and Typhoons. The hurricanes of the West Indies and the typhoons of the East Indies are cyclones possessing both great extent and power. Commencing usually between latitudes 10 and 20 N., they pursue a course a little to the north of west until they reach latitude 30 , when, if they last long enough, they swerve around towards the east, and pursue a course north of east. They diminish in intensity with their northward progress, and finally enter upon the path frequented by the cyclones of middle latitudes, and disappear in the same manner as these. The time which it takes for a hurricane to pass over a given place lying within its path varies from a few hours to a day or more, when we count the time during which the barometer is falling as the hurri- cane approaches, and rising as it recedes from the place. Hurricanes are never more than a few hundred miles in diameter ; but they have long paths, and thus their force is distributed over a large territory. As the hurricane is approaching, there is a decrease of pressure, a slight decrease of temperature (during the 230 ELEMENTARY METEOROLOGY. western or northwestern progress), a rapid increase of wind velocity, and a quite constant wind direction. At the center there is almost a calm. At the rear of the center there is an increase of pressure, an increased tem- perature, first an increase and then a decrease of the wind velocity; and the wind blows quite constantly from a direc- tion nearly opposite to that on the front side. The baro- metric pressure sometimes falls two inches at the center of one of these cyclones. One typhoon was traced from near Manilla, where, on Sept. 27, 1882, it had a movement forwards of but 5 miles per hour, to the coast of Japan (about Oct. i), where it moved at the rate of 33 miles per hour, which increased to 51 miles per hour on Oct. 3. just east of Japan ; thence it passed to the Aleutian Islands and to Oregon, which it reached on Oct. 10; thence across the Rocky Mountains at a rate of 37 miles per hour, and through the northern United States and Canada, Hudson Bay, and Labrador, to Davis Strait; and thence past the southern point of Greenland to longitude 27 west, latitude 55 north, in the Atlantic Ocean, where it united with another cyclone. After this union, the cyclone remained stationary for nearly a week (from Oct. 19 to Oct. 24), when it suddenly took a southeast course towards England, passed over the Bay of Biscay, and reached France on Oct. 27, took another northeasterly trend, and vanished in the region of the Baltic Sea on Nov. i, having thus traveled over 14,000 geographical miles in 35 days. Eye of the Storm. Directly at the center of the cy- clone there is a region in which the air may be quite free from cloudiness, and in which there is but little hori- zontal air motion. This is undoubtedly due to the exist- ence of a descending current at higher altitudes above the ascending current at the interior of the cyclone at low altitudes. The descending current prevents the ascend- ing current from reaching a sufficient altitude to cause precipitation, and perhaps even the formation of clouds, SECONDARY CIRCULATION OF THE ATMOSPHERE. 231 and tends to dissipate clouds already formed. This phe- nomenon becomes so marked, in the case of the limited but intense tropical cyclones, that the limits of the clear space at the center can be distinctly seen, and it is called the " eye of the storm." In the cyclones of higher latitudes it is not so pronounced in character, but is of wider extent. Secondary Cyclones sometimes occur within the bounda- ries of our larger cyclones, usually to the southeast of the center of the main cyclone, where the air is warmer and moister than in other quarters. They have an air circulation against the hands of a watch, like the primary cyclones. Sometimes they are so slight as to cause merely a small distortion of the isobars of the main cyclone, with but little disturbance of the wind direction; and sometimes they cause a fairly well developed cyclonic wind circula- tion. The local upward currents of these secondary cy- clones cause an increase in the cloudiness, and more rainfall at the points where they exist. Tornadoes and kindred phenomena are closely allied to the secondary cyclones, and it is scarcely possible to draw a dividing line between them. It has been deemed best, however, to treat of tor- nadoes separately in the next chapter. Origin of Cyclones. When air flows along in a current, there is a tendency for it to break up into whirls or eddies on the outer edges of the current, and so, near the borders of the great currents of the primary air circulation, such vortices are formed (see Fig. 72); and the centrifugal force resulting from this local whirling motion causes the air to recede from the center of each whirl, and to be heaped up on the outer edge, making a more or less reg- ular increase of air pressure (gradient) from the center to the outer edge of the whirl. These whirls might form the cyclones, or areas of low air pressure ; and the inter- WALDO METEOR. 14 232 ELEMENTARY METEOROLOGY. vening spaces, where the air is heaped up, the anti- cyclones, or areas of high air pressure. This theory, however, has not been worked out with any degree of completeness. g) G) (3) 6) 6) <3) ) 6) g) (3) >- Central > Air y Current 1-> ~~Q) Q) Q) (3) Q) Q) Q) Q) > Q) FIG. 72. BREAKING-UP OF AN AIR CURRENT INTO WHIRLS. Another Source of Cyclonic Motion is to be found in the local temperature conditions. If anywhere, as in the inte- rior of the continent, a local region becomes excessively heated, the air expands upward, and the isobaric surfaces become elevated, and cause an outflow of air aloft into the colder surrounding air. The increased air pressure in this outer region causes the air near the surface of the ground to flow inward toward the warm central area, where the air pressure has been diminished by the overflow of air aloft Connecting these two horizontal currents, there is an as- cending air current over the central warm region, and a descending current on the outer cooler boundary. The direction of these horizontal currents, where they extend over hundreds of miles, is affected by the deflecting force of the earth's rotation. The upper outflowing cur- rent and the lower inflowing current are deflected towards the right (in the northern hemisphere); and so, instead of having a flow of air directly from and toward the heated center, there results a curvilinear or spiral motion in both currents. The inflowing surface current is deflected towards the right of the central area, and flows spirally around it in the direction opposite to SECONDARY CIRCULATION OF THE ATMOSPHERE. 233 that of the hands of a watch (face upwards) . As this air current ap- proaches the center, the velocities are much increased, because the same amount of air has to pass in a given time through a much narrower space nearer the center than farther away from it. The horizontal motion be- comes more and more changed to the vertical as the immediate center is approached, and near the center becomes an upward current ; and this upward current, too, may have great velocities. The upper horizontal current flowing away from the center is like- wise deflected towards the right, and its spiral motion is in the same direction as that of the hands of a watch (in the northern hemisphere) ; and since the channel for the air constantly increases with the distance from the center, the velocities of the current decrease, and on the out- skirts of the whirl there is but a gradual settling-down of the air, instead of a swift down-moving current. The lower inflowing current is much more clearly marked than the upper outflowing current, and its gyratory motion is so great that the opposite gyratory motion of the upper current is at first required to overcome this lower gyratory motion before it can assert itself in its proper direction ; hence, in the neighborhood of the warm center, the gyration of the whole air mass is against the hands of a watch. As the upper outward current increases its distance from the center, the effect of the deflecting force becomes greater, and finally on the out- skirts of the whirl overcomes the less developed deflecting forces of the lower current, and imparts to the whole air mass its distinctive direction of rotation. Therefore, for a cyclonal region of this kind, nearly the whole mass of air at the interior has a circulation around the center in a direction opposite to that of the hands of a watch ; while outside of this there is a ring or zone in which most of the air has a distinctive rotation around the same center, but in the opposite direction to that at the interior. Since the friction with the ground is such as to make the gyratory velocity less below, and the friction of one air layer on another is very slight aloft, then the less the altitude of the air disturbances above the ground, the more powerful will be the effect of the upper air current, and the nearer to the center will be the region where the air circulation changes from the cyclonic to the anticyclonic. This is analogous to the general hemispherical circulation of the atmosphere. The directions of these air currents are shown in Fig. 65. The full arrows indicate di- rections of lower currents, and the broken arrows those of upper currents. 234 ELEMENTARY METEOROLOGY. It is probable that some cyclones are due entirely to the breaking-up into vortices of steady air currents, while others are due entirely to the local unequal heating of the air, and still others are due to a combination of the two processes. Moisture most probably plays a very important part in the maintenance of cyclones, since the large quantities of heat liberated by condensation must retard the normal cooling of ascending air very appreciably, and thus con- tinue the condition of instability. Aside from the fact that local unequal heating of the air may arise from the freeing of heat by condensation when upward currents of moist air are caused by outside influences, there is very great uncertainty as regards the influence of moisture in the formation of cyclones. Air Density. The local density of the air is dependent mainly on its temperature, and to a slighter degree on its humidity. It has been shown that the lines connecting regions having air of equal densities (over Europe at least) run parallel to the coast, and that the density decreases towards the interior of the continent in summer, and towards the ocean in winter. The study of air densities in connection with cyclones and anticyclones has not been thoroughly carried out ; but it has been stated that when the closeness of the lines of equal density signify an unstable condition of equilibrium, then cyclonic disturbances tend to form, and they move along parallel to the lines of equal density. It is very probable that the accurate and careful study of the densities of the air will lead to a better knowledge of the formation and behavior of secondary or local cyclonic disturbances. ANTICYCLONES. Classes of Anticyclones. There are several classes of areas of high barometric pressure. One class is due to the heaping-up of the air thrown off by centrifugal force from SECONDARY CIRCULATION OF THE ATMOSPHERE. 235 the cyclonal whirls. Such is the ring of high pressure which encircles the globe at about latitude 30. When cyclonal areas enter this ring of high pressure from with- out, the air is still further heaped up on their outskirts ; and these masses of air are forced out from the main ring especially on its poleward side, and are made to project far into the higher latitudes between the successive cyclones. Another class of high-pressure areas is due to local cooling of the air. These are the isolated areas of high barometric pressure, which are essentially phenomena of the colder and continental latitudes, where the lower air is cold and dense. The isobaric surfaces being thus brought close together at the cold center, the upper air flows down them towards the center from the warmer regions without, and increases the pressure at the center. Cyclones and Anticyclones contrasted. Anticyclones are by no means so well marked in their limits and char- acteristics as cyclones. In anticyclones the air pressure is greatest at the center, and decreases thence in all direc- tions somewhat irregularly ; and on the side where a cyclone is to be found, this decrease continues practi- cally uninterruptedly to the region of lowest air pressure in the cyclone. Anticyclones are usually an accompaniment of cyclones ; and in fact where a cyclone exists, with its deficiency of air, there must lie at not a great distance trom it one or more anticyclonic areas into which the air has overflowed from the upper region of the cyclone. Sometimes the cyclones are surrounded by a ring of anticyclones, but the continuity of the latter is nearly always broken by slight depressions or troughs in which the air pressure is lower than the maximum barometric pressure in the anticyclones, which in most cases is above 30 inches. 236 ELEMENTARY METEOROLOGY, Air Pressure in Anticyclones, In ordinary anticyclones of our middle latitudes the barometric pressure frequently reaches 30.6 inches, and in some cases it exceeds even 31 inches. The isobars are even more irregular than in cy- clones, and are usually elongated (elliptical), with the great- est axis pointing most frequently about northeast and south- west, although it may be directed towards any point of the compass. The longest diameter is on the average about twice the length of the shorter, but sometimes exceeds it fourfold. The relative length and direction of the axes de- pend a good deal on the adjacent areas of low barometer. When an anticyclone is situated between two well-devel- oped cyclones not far apart, then the anticyclone becomes much elongated; and when this latter condition obtains there are sometimes two or three distinct centers, which differ in maximum pressure by only slight amounts. The dimensions of anticyclones exceed those of cyclones. It has been found that the diameter of those in the winter time, with extraor- dinarily high pressures, was from 3,000 to 4,000 miles ; and the pre- ceding and following cyclone centers were distant fully 2,000 miles from the center of the anticyclone, and the preceding cyclone was much the deeper of the two, and the steepest gradients were on this side. Such relations are shown in Fig. 74. Moisture in Anticyclones. Anticyclones are character- ized by a dry cool air and little cloud, and but slight precip- itation. The descending air at the center tends to dissipate the cloudiness. This is due to the fact that as the air descends it is subject to a greater compression, becomes denser, and the temperature is raised ; by which means the relative humidity is decreased, and cloud is dissipated. Ground fog frequently occurs in anticyclones, because on unclouded nights just at and very close to the ground the SECONDARY CIRCULATION OF THE ATMOSPHERE. 237 temperature abruptly decreases to below the point of con- densation, and fog is formed. Temperature Distribution in Anticyclones. Near the earth's surface the temperatures are generally below the normal air temperatures ; and usually the more pronounced the anticyclone (i.e., the greater the air pressure), the lower the temperature. At higher altitudes, however, the air is but little colder than near the ground, and in many cases there is even an inversion of the usual temperature decrease with altitude, and it is actually warmer above than below. This is due to the excessive radiation of heat from the ground into the clear sky. In the anticyclones of our middle latitudes the air is colder on the eastern side. Direction and Velocity of the Wind in Anticyclones. - Near the surface of the ground in anticyclones the air motion is directed spirally outward. The winds at high altitudes are, on the contrary, directed spirally inward ; and at the center of high pressure there is a vertical downward current connecting the inflow above with the outflow below. The wind velocities in anticyclones are much smaller than those in cyclones. The inner downward current has probably very much less velocity than the upward current in cyclones, and usually amounts to but little more than a settling-down of the air towards the earth's surface. Calms are very frequent within anticyclones. The Direction and Velocity of Propagation of Anticyclones. Anticyclones, like cyclones, drift along in the general direction of the great air currents. Since they occur mostly in the middle and higher latitudes, their direction is generally easterly In the United States the anticy- clones move, on the average, in a southeasterly direction, that is, at right angles to the direction of greatest elon- gation; but in winter the direction is a little more east- 238 ELEMENTARY METEOROLOGY. erly. The velocity of translation is not so great as that for cyclones, nor are the paths pursued so long. The reason of this is not well understood, but it is supposed that some process in the formation or continuance of cyclones accel- erates their movement in certain directions ; and this fea- ture is lacking in the anticyclones, which lag behind. Excessive High Pressure in Anticyclones - - In the case of an ordinary anticyclone at the rear of a cyclone, we have very little cloud, and consequently the outward radiation is very rapid and great, and the ground and lower air layers are still further cooled. This causes the lower air to become still denser, and the isobaric surfaces are lowered so that the air at greater altitudes in the sur- rounding region flows down these surfaces and in towards the center. This increases the air pressure, which was already high, and causes it to become very high if the air becomes sufficiently cooled. This accounts for the extremely high air pressures observed in winter. In considering both cyclones and anticyclones, it must be remembered that their lateral extent is perhaps a thou- sand times their thickness, and the mass of gyrating air may be regarded as a relatively thin disk. There is thus a chance for a great waste of force through friction, and innumerable opportunities for the interference with the horizontal circulation by means of vertical currents. The original forces which give rise to these phenomena can thus soon become used up. Relation of Cyclones to Anticyclones. Anticyclones are a natural consequence of cyclones, because, when there is a diminution of pressure in one place (in the cyclone), there must be a corresponding increase in another place (in the anticyclone). The air movement between a cyclone and an anticyclone is shown in the accompanying diagram Lower current Upper current SECONDARY CIRCULATION OF THE ATMOSPHERE. 239 (Fig. 73), where, for the lower layers of the air, the baro- metric minimum or cyclone is shown on the left, and the barometric maximum or anticyclone on the right; and the wind directions there are shown by the full-drawn arrows. At higher alti- tudes, however, in the cloud region, the wind di- rection for which is shown by the broken-lined arrows, above the cyclone the di- rection of motion of the . . , FIG. 73. INTERCHANGE OF AIR BETWEEN CY- air IS Seen tO be antlCyClO- CLONE AND ANTICYCLONE (NORTHERN HEMI- nal and away from the cen- SPHERE). ter, while above the anticyclone it is cyclonal and towards the center. Thus the upper current from the cyclone re- plenishes the air lost from the anticyclone at the surface, and the lower current from the anticyclone replaces the air lost from above the cyclone. The cyclones coming from the tropics have to break through the ring of high pressure at about latitude 30 ; but there the absolute mass of air is so great, and the cyclone so limited in extent, that the air thrown out by them does not cause much of an increase in the pressure ; so that, while there are well-marked cyclones in the lower latitudes, yet the anticyclones with excessively high pres- sure are not to be found there. The simultaneous distribution of cyclones and anticyclones over the northern hemisphere is shown in Fig. 74, which gives the isobars on Dec. 15, 1882. This shows an area of high pressure (an anticyclone) over North America, and one over Asia, with areas of low pressure (cyclones) over the eastern Pacific Ocean and the North Atlantic Ocean. A less pronounced anticyclone is also to be seen over the southern part of the North Atlantic Ocean. 240 ELEMENTARY METEOROLOGY. 110 100 90 80 FIG. 74. CHART SHOWING ISOBARS, DEC. 15, 1882 (AFTER LOOMIS). Permanent Cyclones and Anticyclones. In the charts showing the average air pressures over the globe, it is seen that there are local regions of high and of low pres- sure, the so-called permanent anticyclones and cyclones. CHAPTER X. LOCAL AND MISCELLANEOUS WINDS. Irregular Local Atmospheric Disturbances. In those regions where there arise disturbances of the condition of equilibrium of the atmosphere over a limited area, there is found a class of phenomena of which the thunder squalls, spout phenomena (such as waterspouts, sand spouts, and tornadoes), and the straight blows (dereckos) form the principal members. These are all somewhat closely related as to their cause ; and the final development of one form or another is due to variations in the conditions under which they arise, and somewhat also to the geograph- ical locations. While these phenomena are individually of but small extent, yet the conditions favorable to their formation may extend over a considerable area, so thai, instead of individuals, groups may exist at any one time, and these progress in parallel tracks. TORNADOES. A Tornado is a progressive, limited, local, violent whirl- wind, characterized by a funnel-like cloud which hangs suspended from an intensely black mass of storm clouds ; the apex of the funnel cloud sweeps over the earth's sur- face, sometimes touching it, and sometimes receding from it, to come down again to the ground farther on in the course of the cloud as it moves forward. An ideal picture of a tornado in the United States, as viewed when looking toward the east, is shown in Fig. 75. The peculiar funnel-like cloud may be seen at the center, receding in a northeasterly direction from the scene of destruction. 241 242 ELEMENTARY METEOROLOGY, The first visible precursor of one of these tornadoes is a heavy bank of clouds appearing in the southwest, and later in the west and northwest ; in these a violent commotion is observable, with adjacent clouds rushing in from the south- east, east, and northeast, towards the center of the disturb- ance. If the clouds are light in color, they resemble smoke clouds from a large fire ; if they are dark, they are of a peculiar greenish hue, which increases in intensity as the FIG. 75. PASSAGE OF A TORNADO (THE OBSERVER FACING THE EAST). clouds approach. Sometimes these dark clouds look like the dense volumes of smoke emitted from the smokestack of an engine. It is among such clouds that the first ap- pearance of the tornado funnel cloud is to be noticed. It then seems to descend toward the earth by gradual length- ening or growth of the funnel. The near approach of the cloud is announced by a heavy rumbling sound like that of an approaching railroad train, or of distant thunder. LOCAL AND MISCELLANEOUS WINDS. 243 The noise increases as the cloud draws nearer, and its passage creates such an uproar that all other sounds are deadened. The observer* in front of this rapidly moving mass of clouds finds himself in the region of a gentle south breeze of warm air, or of calm air and oppressive heat. This quiet condition is suddenly disturbed by the tempestuous whirlwind caused by the tornado itself, which cuts a swath in its narrow track, felling everything before it in an irresist- ible manner. The passage of the funnel-shaped cloud is succeeded by a sudden lowering of the temperature, and by a calm or gentle breeze. Air Circulation in Tornadoes. There is a horizontal current of warm moist air flowing in from all sides (but not from a great distance) towards a center, and an upward cur- rent at this center which is fed by the horizontal current just mentioned. Above these there is an outflow away from the center. The motion of the upward current is not di- rectly upwards, but takes a spiral form ; and its position is shown by the outlines of the peculiar tornado cloud which hangs in funnel shape, suspended from the intensely black mass of storm clouds (see Fig. 75). The direction of this spiral rotation has been observed to be opposite to that of the hands of a watch, in the cases where this phase has been noticed, but it has not been possible to obtain frequent careful observations of it. Formation of Tornadoes. Tornadoes are caused by local differences of temperature. The air having become abnormally heated over a central area, there results a differ- ence in pressure between the air of the inner region and that surrounding it ; from this there arises a flow of air spirally inward towards the center, and as it is approached, the velocity is increased. The principal condition for the formation of a tornado is the local unstable condition of 244 ELEMENTARY METEOROLOGY. the air, due to the abnormal heating of a mass of air either at the earth's surface or at some locality above it. This mass of air, being warmer than the surrounding air at the same level, is in unstable equilibrium ; and when some slight disturbance frees it from its abnormal posi- tion, it is forced upwards, by the pressure of the air below and around it, to that altitude where its tempera- ture and density are normal (that is, the same as the temperature and density of the surrounding air); there the power causing the vertical current ceases. 1300? r i 8-6 6 In this ascent the air cools by expansion ; but in ascending moist air, when condensation occurs the cooling is slower, and consequently the difference in temperature between the out- side and inside column of air becomes greater (Fig. 76), and the upward velocity is increased. But this ascending air at first accumulates above, and the pressure thereby increases, so that at some altitude the air pressure over the ascending column becomes greater than in the surrounding air. The difference between the interior and exterior temperature increases up to this level ; but above this point it decreases (owing to the greater density of the air above it), and at some higher altitude the temperature is the same without and within the current, and still above this (and also below the altitude of condensation) the decrease of temperature with the ascent is more rapid within than outside. At the altitude where the tern perature becomes the same inside and outside of the current, the air pressure also becomes the same. Below this level the temperature is greater and the pres- sure less in the ascending current with condensation than outside of it, and above this level the condition is reversed. Adjacent Space temperatures Altitude Central Space Altitudes in meters inC. FIG. 76. VERTICAL DECREASE OF TEMPERATURE (SPRUNG). LOCAL AND MISCELLANEOUS WINDS. 245 This increased pressure above forces the air out on all sides as it moves upward, and this air flowing over into the adjacent air mass makes it heavier, and the air below is forced in from it to replace the air which has moved upwards. There is thus caused a vertical air circulation upwards within the central area, and downwards on all sides around it ; while an outward horizontal current above, and an inward horizontal current below, connect the two vertical currents. This cir- culation may take place when the unstable condition exists for any air mass, whether this condition be near the earth or at high altitudes, or whether it extend through the whole vertical air column or not. If this circulation took place without any rotary motion, the velocities would be small, and the flow of air gentle ; but because of rotation the veloc- ities increase, and near the center become very great, according to the law of conservation of areas. This rapid gyratory motion near the center gives the tornado winds their tremendous velocities. With the increasing wind velocities towards the center there is an increasing centrifugal force, and the air is forced away from the center as in the case of cyclones, and thus gradients arise in the air pressure ; and these gradients are steeper the nearer the center is ap- proached, so that there is a decrease in the air pressure towards the center of the whirl, and near the center this decrease becomes rela- tively very great in the case of the great rotational velocities met with in tornadoes. Sustaining of Tornadic Action. The difference in tem- perature within and without the central upward current of the tornado, which is necessary for its continuance, could not persist long enough for its full development if it were not for the freeing of latent heat through conden- sation. This latent heat becomes sensible, and retards the cooling of the inner air column. It is therefore when the air is saturated with moisture that the conditions are best for the maturing of a tornado ; and when conden- sation no longer takes place, the tornado action must soon cease. When and Where Tornadoes Occur. The conditions necessary for the formation of tornadoes are to be met 346 ELEMENTARY METEOROLOGY. with at low altitudes in the moist atmosphere of the lower and middle latitudes. They occur in western Africa, 1 southern Asia, occasionally in the lower middle latitudes of Europe, but most frequently in the central and eastern portions of the United States. Most of our knowledge of these phenomena is obtained from the study of those occur- ring in the last-named country, and the following remarks apply to them. The true tornado is probably not observed above the 5Oth parallel of latitude. Tornadoes in the United States form most frequently several hundred miles (300 to 500) to the southeast of the center of a cyclone which has a long troughlike central depression extend- ing north and south or northeast and southwest. This region is the one of greatest heat and moisture, and of surface winds of considerable force from the south. The regions of greatest frequency of tornadoes are the lower Missouri, the central Mississippi, and Ohio river valleys ; and the season of their greatest frequency is the late spring and summer for these regions, but farther south (in Georgia, for instance) many tornadoes occur in the early spring. Few tornadoes occur in the United States west of the looth meridian. The hours of greatest fre- quency of tornadoes are from 3.30 to 5 P.M. Paths of Tornadoes. The path of destruction which marks the central region of a tornado varies from a few feet up to 2 miles, the average being about a quarter of a mile. The length of such tracks, which mark the places where the tornadoes reached to the earth's surface, vary from 300 yards to, 200 miles, with an average of about 25 miles. The funnel-shaped cloud at the earth's surface 1 Where they occur in the southern hemisphere, the directions of motion are the reverse of those in the northern hemisphere, because in the former the cyclonal rotation is with the hands of a watch. LOCAL AND MISCELLANEOUS WINDS. 247 varies in diameter from a few yards to an unmeasured limit, but the average is probably several hundred feet. Velocity of Motion of Translation of Tornadoes. The velocity of progression of tornadoes varies from 7 to 100 miles per hour, with an average of about 44 miles per hour. The same tornado cloud may sometimes remain almost stationary for a while, and then dart forwards with great velocity. The average time occupied in passing a point is a little over a minute. The direction of motion of translation of the tornado is nearly always from southwest to northeast. Air Pressures in Tornadoes. The normal air pressure is about 14.7 pounds per square inch; and if, as is possi- ble, the air pressure is reduced one fourth of this amount at the center of a tornado, then it is lessened there about 3.7 pounds per square inch, or 533 pounds per square foot. Supposing, now, that the tornado passes over a house in which the air is at the normal pressure (that is, on the inner and outer walls there is a pressure of 2,117 pounds per square foot), then the pressure on the outer walls during the passage of the center of the tornado will be suddenly diminished by 533 pounds per square foot, and there will be an excess of pressure by this amount from within outward. The diminution of the outside pressure comes very suddenly, and this inside pressure acts like an explosion. It is not to be wondered at, then, that doors and windows are blown out, and walls even torn down, when this great force so acts suddenly on them in an out- ward direction. Destructive Winds in Tornadoes. The enormous wind velocities in tornadoes are measured only indirectly by the mass or strength of objects which they have moved or broken. It is probable that they reach or even ex- WALDO METEOR. 15 248 ELEMENTARY METEOROLOGY. ceed 400 or 500 miles per hour in some cases. The winds are greatest near the center, and decrease from thence out- wards ; they are less in front than at the rear. Tornadoes are usually accompanied by hail fall, and very frequently by manifestations of atmospheric electricity. On the right side of the tornado the winds are stronger than on the left-hand side of its direction of progression ; and the path of destruction extends much farther from the center on the right than on the left side. The winds on the left (or northwest) side are the weakest, because, for one thing, the direction of progression is there directly opposed to ^_ that of the gyratory motion. ^_ >^ In Fig. 77 the long arrow ^. shows the direction of pro- " ^ ^ gression, and the small arrows > ^- > the gyratory motion ; and it FlG< 77 ' is seen that on the right-hand side the two motions are in the same general direction, while on the left the directions are opposed to each other. The wind force in tornadoes depends directly on the velocity of the air current and the density of the air. It has been computed, that, in the case of a tornado in which the gyratory velocity was 10 feet per sec- ond at a distance of 3,300 feet from the center, the velocity at 70 feet from the center would be 460 feet per second, or 310 miles an hour, which would give a pressure force of about 300 pounds per square foot on any plane surface exposed squarely to the wind. In case the ascend- ing current in a tornado had a velocity of 100 miles an hour at the alti- tude where the air pressure is 15 inches, this would prevent a hailstone 2.5 inches in diameter from falling to the ground. Large raindrops o.l of an inch in diameter require an upward current of only 16.5 miles an hour to sustain them when the air pressure is 23 inches ; while small drops only 0.003 mc ^ in diameter require an air current of only 3 miles an hour to prevent their falling to the ground. The size of raindrops may be taken as indicating roughly the veloc- ity of the ascending air current. When the raindrops are fine, there is a weak upward current ; but when they are coarse, there is a more powerful one. LOCAL AND MISCELLANEOUS WINDS. 249 Safety in Tornadoes. Concerning the safety of those who find themselves in the path of a tornado, it may be said that it is best to enter some subterranean vault, such as an ice cellar. The cellar of a house is preferable to the house itself if the house is of wood ; but if it is of stone or brick, one should not enter the cellar, but leave the build- ing. The southwest corner of a cellar is the safest. If one must escape a tornado in an open place, he should first take the bearings of the approaching tornado and its course. If the tornado is discovered to the west or south- west of him, he should move off with dispatch towards the southeast or northwest, respectively. If the tornado is far to the south or to the north, one is usually safe. If one is caught in the whirlwind's blast, it is best to lie prone on the ground in an open field rather than seek shelter in a wood or under trees. Frequency of Tornadoes. During the 10 years from 1877 to 1887 the number of tornadoes reported in the United States averaged 146 a year. Observations do not as yet prove that tornadoes are increasing or decreasing in frequency. Multiple Tornadoes. Frequently on the same afternoon a number of tornadoes will form. They may be close to- gether or several hundred miles apart, and they move in parallel directions. Thus they form a group or band of isolated whirlwinds, all of which have for their origins the same general conditions. THUNDERSTORMS. Thunderstorms and their Attendant Phenomena. Thun- derstorms are local progressive atmospheric disturbances occurring in most latitudes inhabited by mankind, in ELEMENTARY METEOROLOGY. regions where there is considerable atmospheric moisture. They receive their name from the electrical phenomena which distinguish them. They are usually accompanied by rainfall. Their attendant phenomena in our latitudes are as follows : First dark clouds are seen lying low in the western sky, and light southern winds are experienced, the air being warm and sultry. Perhaps an hour later the clouds have mounted to near the zenith, but the same rel- ative quietness of the air continues near the ground ; and the heat is perhaps not so great, owing to the sheltering influence of the clouds. The latter usually present an appearance something like this : On the front side of the thunderstorm there are grayish white or reddish clouds hanging over and in front of the main rain cloud. Above these, dense dark-gray and violet cumulo-stratus are seen, as also dome-like cumulus clouds, which are separated from the cumulo-stratus. Often these are interspersed with one or more thick cumulo-stratus-like cloud layers, and above all is the widely distributed cover of cirro-stratus. The first thunder is heard before the cloud reaches the zenith, and the first rain commences after it. The interval between the first thunder and the beginning of the rain varies from a few minutes to perhaps half an hour or more. Just before (usually only five minutes or so) the rain begins, there comes from the west or northwest a brisk wind, which suddenly increases in violence, and be- comes a squall, which is the name given to a sudden vio- lent and not long-continued wind ; this, however, dies down very much, after the rain begins. The time of heaviest rainfall varies: sometimes it occurs at the begin- ning, and sometimes in the latter part, of the time when the rain falls. The thunder is loudest, and lightning strokes occur some minutes after the rain begins. LOCAL AND MISCELLANEOUS WINDS. 251 Soon after this the western horizon loses its dark aspect, and begins to lighten up a little, and finally the clouds break there, and blue sky is seen. The storm clouds then pass by overhead, and the rain ceases some minutes before their western edge reaches the zenith. Although the tem- perature may have fallen even as much as 25 F. during the storm, yet it frequently regains its former condition after the storm has passed. Rainbows appear in the latter part of the rain period of the storm. The last thunder is usu- ally heard after the rear edge of the clouds has passed the zenith. Individual thunderstorms seldom last (between the first and last thunder) over two hours, but frequently one thunderstorm follows another in such quick succession as to appear to be part of the same storm. In our lati- tudes the usual direction of translation of thunderstorms is easterly. Classes of Thunderstorms. -- Thunderstorms are now generally divided into three classes, cyclonic, heat, and winter thunderstorms. Cyclonic thunderstorms accompany the well-developed areas of low barometric pressure, and they have a circula- tion somewhat analogous to that of cyclones, becoming in extreme cases tornadoes. Heat thunderstorms are the result of the local heating of the lower air, which makes its condition unstable. They need for their development a moist quiet air warmed by the sun's rays, and they occur in their most distinct state outside of the regions of strong ascending and de- scending currents of cyclonic and anticyclonic areas. Probably many, if not the majority, of our thunder- storms are a combination of these two forms. Winter thunderstorms occur most frequently at night, especially in high latitudes, and they are more frequent 252 ELEMENTARY METEOROLOGY. '-^-IS '* near the coast than inland. These winter storms are of more limited area, travel faster, are shorter in duration, and their lightning is more destructive, than the summer thunder- storms. Heat or Stationary Thunderstorms. In a stationary thunderstorm, that is, one having no or but very slight progres- sive motion, we have at -5 first an ascending cur- FIG. 78. UPWARD AIR CURRENT IN THUNDER- rent of moist Warm air over some limited region. STORM (AFTER FERREL). Air then flows in from all sides, and a vertical circulation arises (Fig. 78). When the ascending air reaches an alti- tude where its dew-point occurs, then condensation begins, and a cloud is formed which has a flat base. Above this, precipita- tion occurs ; and the rain falling through the air beneath cools it, so that it contracts, becoming heavier than the sur- rounding air, and, being also pressed downward by the falling rain, falls downward, after over- coming the gentle upward current which already existed there. The movement of the air downward causes an out- flow of the air below it, and, forcing its way into the origi- a c c' c FIG. 79. AIR CURRENTS IN THUNDERSTORM (AFTER FERREL). LOCAL AND MISCELLANEOUS WINDS. 253 nai ascending current which surrounds it, causes it to spread out, thus increasing the size of the atmospheric disturbance and the cloud region overhead. The air thus forced out from below at the center also takes part in the vertical cir- culation of the surrounding ascending air (Fig. 79). Since this occurs on all sides, the outer upward currents are ar- ranged symmetrically around the inner downward current. There is, then, a central region of greatest air pressure caused by this central cooling and downward flow of air, which reaches its maximum at the earth's surface; and outward from this the temperature increases and the pres- sure decreases, and this change is very rapid within a short horizontal distance at the dividing line between the descending and ascending currents. It is in this region of steep gradients that the squall or sudden wind occurs. The air which descends at the interior, forces its way out under the ascending current encircling it, and, in part at least, very soon joins the latter in its upward movement, after its horizontal force is somewhat spent. By this mix- ture of the cool interior air with the warmer outer air, the latter becomes cool on the inner side, and changes there to a descending current, so that the ring of steep gradient and squalls is enlarged. This process lasts until the stable equilibrium is restored, when the storm disappears. Cyclonic or Progressive Thunderstorms. It must be rare, at least in middle latitudes, that the air is quiet enough to permit the formation of a symmetrical thunderstorm such as has just been mentioned. In the southern and southeastern parts of the large cyclones, and at a distance of several hundred miles from the center, where our thun- derstorms usually occur, there is an easterly motion (the wind varying from northwest to southwest) of the air. It is this easterly motion which carries the thunderstorm on- 254 ELEMENTARY METEOROLOGY,, wards ; and since the motion is greater above than below, new masses of air are introduced on the front side near the earth's surface, and thus a fresh supply of moisture is drawn into the vertical currents to keep up the precipita- tion; and on the rear side this air is left drier and cooler. Not only does the storm progress in this easterly direction with the upper air currents, but it increases also on the side of greatest moisture and heat, which in this case is the preceding or front side. The local gradient which occurs between the air of the inner and that of the outer regions, and which is indicated by the sudden rise of barometric pressure at the central region in the thunderstorm, is sufficient to cause a wind as powerful as that observed at the front edge of the storm. Thunderstorms seem to occur under much the same conditions as those favorable for the formation of torna- does. The latter are nearly always accompanied by light- ning ; and where the tornado is too far distant to render its distinctive funnel cloud visible, but not too far for its thunder to be heard or its lightning seen, it is regarded as a thunderstorm. Progressive Movement of Thunderstorms. Although the thunderstorm thus commences small in size, it does not remain so. As the storm progresses (moving generally in a direction somewhat easterly), it broadens out so that it presents an ever-increasing length of front, until it is dis- sipated or else breaks up into smaller storms. This is illustrated in the following diagram (Fig. 80), in which the relative length of the front at successive hours is shown. Sometimes thunderstorms expand in all directions from a central point, but this occurs usually when the storm is very local and stationary, or in a stationary stage. The direction of propagation of thunderstorms is usually that of LOCAL AND MISCELLANEOUS WINDS. 255 the local wind belonging to the cyclonic system in which it appears. In Europe, thunderstorms move with an average velocity of from 22 to 25 miles per hour ; and in general the velocity is probably nearly equal to that of tornadoes. Time of Occurrence of Thunderstorms. Our thunder- storms occur most frequently in June, July, and August, and least frequently in winter; they occur most frequently in the middle of the after- noon, and least frequently in the early morning hours. In some places it has been FIG. 80. PROGRESSIVE GROWTH OF THUNDER- found that more thunder- STORMS. storms occur one or two hours after midnight than in the hours just preceding and following. Variation of Number with Latitude. Thunderstorms are in general most frequent near the equator, and decrease with increase of latitude ; but their number depends also on the temperature, the moisture, and the movement of the air. For instance, in the tropics, during the season of calms, thunderstorms are frequent ; but during the season of trade winds they are rare, because, the air as a whole being then in constant motion, the initial condition of unstable equilibrium necessary for the formation of thun- derstorms seldom occurs. Changes in Meteorological Elements during Thunder- storms. In the passage of thunderstorms the meteorolog- ical elements undergo the following changes : Before the thunderstorm the air pressure and the relative humidity de- crease, and the temperature rises, the wind being generally weak ; so that at the beginning of the storm the pressure and relative humidity are at their lowest, and the tempera- ture at its highest point. At the moment of the bursting of 2 5 6 ELEMENTARY METEOROLOGY. the storm the air pressure and relative humidity increase very rapidly, and the temperature falls ; the wind also suddenly becomes strong, and sometimes it as suddenly subsides almost immediately afterwards, while at other times it increases until near the close of the thunderstorm. Towards or at the end of the thunderstorm, the air pressure 28C. 26 4 24 3 22 2 20 1 18 81 53 26 N.W. W s.w. s 754 S.E, 753 "752 mm, 'Bctr'on FIG. 81. METEOROLOGICAL ELEMENTS D.URING A THUNDERSTORM (SPRUNG). and relative humidity reach their maximum, and the tem- perature its minimum. These relations are shown by the preceding diagram (Fig. 81), which shows the variation of the meteorological elements, at a place in the path of a thunderstorm, during the afternoon and evening of the day on which a thunderstorm occurred at half-past six o'clock. LOCAL AND MISCELLANEOUS WINDS. 257 The rain areas in progressive thunderstorms increase with the spreading-out of the upward air currents and gradual growth of the storm; and at any one time the rain area has an average form as shown in Fig. 82. The thun- derstorm is carried along by the wind which is least at the sur- face of the earth; so that the warm moist air on the front side feeds the air in the thun- derstorm like a lower current FIG. 82. FORM OF RAW AREA IN A in a westerly direction, and the storm grows in extent towards this lower feeding cur- rent, which supplies it with moisture. In the rear the storm dies out from lack of moisture to sustain it ; also the storm spreads toward the south when warm moist southern winds blow into it, and with the diminution of this moisture by precipitation there is a dying-out of the storm. The growth of the storm is therefore toward the east and south. On the north side the storm increases in size but slowly and in a more regular form, and is even wider than on the south side, because on the north the changes of moisture from one stage to another take place less rapidly, and the moisture is held in the air longer. Ideal Sketch of a Thunderstorm. The following dia- gram (Fig. 83) gives a detailed sketch, in vertical section, of the clouds and their circulation in a thunderstorm, or in a rain squall not accompanied by thunder. On the right-hand edge or front there is seen the outflow of air below, with a gentler inflow towards the cloud above. In the rainy region, shown by the shading, there is a downward flow of air, and to the rear of this is the upward current already mentioned. The 2 5 8 ELEMENTARY METEOROLOGY. lower edge of the roll-like clouds at b and () may be from 1,000 or 1,200 to 1,500 feet above the ground, while the dome-like vault of the cloud in the rain section is perhaps 2,500 feet. At (J) there may be almost or quite a cessation of the rain, and then it may commence again, and perhaps another cessation may occur as shown. The cloud FIG. 83. -CLOUDS AND WINDS IN A THUNDER SQUALL WHICH is MOVING TOWARDS THE RIGHT (VERTICAL SECTION). (AFTER KOPPEN.) forms are distinguished as follows : At c are the cumulus cloud heads or tops projecting from the mass of dark storm cloud a. At b and (6) are dense rolls of clouds hanging down lower than the main mass. At ee there is a continuous cirro-stratus layer ; and lower, at e', is another FIG. 84. CLOUDS AND WINDS IN A HAIL SQUALL. layer partly broken, and pierced by the heads of the cumulus clouds below. At d instead of rounded tops there are cloud tops of fantastic forms resulting from the outstreaming of the air. The downward cur- rent between b and (b) is due to the heaping-up of air at a, and replaces LOCAL AND MISCELLANEOUS WINDS. 259 the air carried upward on either side at b and (). At the rear of the storm are frequently seen, as at g, isolated cloudlets due to the local complete saturation of the cold air just after the storm has passed. In case of hail formation and hail fall, the frozen rain will fall in the region at the front side shown by the coarser dotted shading in Fig. 84. Hailstones may form in the upward current of moist air, and then drop out to one side of the upward current down into the region of rain, where they will have water added to their exteriors ; and then they may be carried up, in the ascending current into which they have fallen, into the region of freezing. The hailstones may become largely increased in size by repetition of this process. Sometimes the central kernel of a hailstone is a snowflake, which at first becomes thoroughly wet by con- tact with rain, and then is carried up by an air current to the region of freezing ; and then, following the process just described for a hailstone, it may increase to a very large size. A variety of such processes must occur in order to account for the different kinds of hailstones which fall during storms of this kind. Thunderstorms, tornadoes, and hailstorms have a common relation to cyclones as far as position is concerned, since they all depend on the same condition of great humidity and instability of the atmosphere. In addition to those conditions necessary for the formation of a thunder squall, tornadoes require the local conditions which cause the gyratory motions peculiar to them ; and hailstorms require the powerful ascend- ing current which extends up to altitudes where freezing can take place. SPOUTS. Spouts are cloud-like or fog-like phenomena of a slender, more or less tapering form, like a long trunk or funnel, formed in the air under certain atmospheric conditions The funnel-shaped cloud of a tornado is a spout, but there are also column-like spouts which occux- in fair weather, especially over water surfaces, and which extend from the earth's surface to the clouds. These are waterspouts, and are the cloud brought down to the earth by the rapid gyra- tory motion such as occurs at the center of a tornado. 260 ELEMENTARY METEOROLOGY. Spouts are due to the rarefaction of the air caused by the centrifugal force driving some of the air from the center of a rotating column of air in which the motion is spirally inward and upward. As the moisture-laden air enters this central region, where the air pressure is reduced so much as to bring the air temperature down to the dew-point, the vapor is condensed into cloud. It is this cloud of con- densed moisture which we see as a dark column. Water is sucked up some distance above the water surface at the foot of the column, but probably not very high up into the air. The gyrations being most violent at certain al- titudes above the earth's surface, the central rare- faction is in the begin- ning greatest up there ; and the funnel or spout cloud forms there first, and afterwards extends down to the earth's surface if the gyratory velocities are sufficiently great to diminish the air pressure in the central region by such an amount as to cause conden- sation of the moisture in the lowest air layer. Fre- quently spouts of little energy are not visible clear to the earth's surface. Fig. 85 illustrates waterspout phenomena of various degrees of energy, and shows how a water FIG. 85. WATERSPOUTS. LOCAL AND MISCELLANEOUS WINDS. 261 surface is stirred up by the whirlwind at the base of the spout Dust Whirlwinds are really spouts of feeble energy, in which the air is too dry to permit the cloud formation by the condensation of vapor, The dust, however, is carried inwards and upwards, and is held suspended in the air where the motions are strong enough to- sustain it; and the outline of this region is rendered visible by the dust particles. White Squalls, or. fair-weather whirlwinds, occur when the conditions are ripe for spout formation, but there is not enough moisture present to form the cloud through- out the length of the central vortex. Up at a considerable altitude, however, a cloud may be formed which will indi- cate the location of the top of the vortex; and at the bot- tom the gyratory inflowing wind may create a disturbance which is especially noticeable over a water surface, and in well-developed cases proves of great danger to ships. These squalls are sometimes called bull's-eye squalls, on account of the peculiar appearance of the small isolated cloud which marks the top of the invisible spout at the center of the whirlwind. Cloud-bursts are sudden and excessive downpours of rain, or rain and hail, which have been carried upward or merely sustained and kept from falling by the ascending air currents, until a large amount has been accumulated aloft, when, by some weakening or breaking-up of the ascending currents, the whole or a part of the accumu- lation suddenly falls to the ground. Cloud-bursts are of most frequent occurrence in connection with tornadoes, where the immense velocity of the ascending current is favorable to the collection and support of great masses of water. 262 ELEMENTARY METEOROLOGY. PERIODIC LOCAL WINDS. Periodic Local Winds are periodic winds which do not depend on the general or secondary circulation of the atmosphere. They are the land and sea breezes, moun- tain and valley winds, and monsoons. Land and Sea Breezes occur daily on the coasts of large bodies of water. When the sun rises, the land is warmed more rapidly than the water surface ; and this heat, being communicated to the air above the land, causes an expan- sion upwards, and the surfaces of equal air pressure are thus raised higher over the land than over the sea. The air over the land flows down these inclined surfaces (so to speak) towards the sea, and the pressure of the air over the sea is thus increased, and the air pressure over the land correspondingly decreased. The increased pressure over the sea causes a return surface wind to set in toward the land region of deficient pressure. This latter air current commences at some distance out at sea, away from the shore, and moves inward towards the land. This circula- tion is kept up until the air above the land is no longer warmer than that over the water. At evening, when the sun goes down, the land cools more rapidly than the water, and the air over the land becomes cooled more rapidly than that over the sea. The result is, that the surfaces of equal air pressure are lowered over the land, and the air above flows from above the sea down these surfaces, causing an excess of pressure over the land, and an under wind from the land to the region of defi- cient pressure over the sea. This land breeze continues until the temperature and pressure differences are adjusted. Mountain and Valley Winds occur in mountainous regions in relatively quiet conditions of the air. There is LOCAL AND MISCELLANEOUS WINDS. 263 then, especially in clear weather, an upward motion from the valleys during the day, and a downward motion towards the valleys at night. These winds have some analogy to the land and sea breezes. In the night the cold denser air descends from the mountain tops, which cool very rapidly by radiation, toward the valleys below. These downward currents commence first in the narrow valleys on the mountain side, in which the air is least heated during the day ; but the descent soon becomes general, and the air flows down to wider valleys below. In the morning the air over lower levels of the ground becomes heated, and the surfaces of equal air pressure over those levels at some little distance away from the mountain are elevated, but directly on the mountain they are not, because there is no air below to expand upward; so that the air flows down these nearly horizontal surfaces towards the upper parts of the mountain, which it strikes at such an angle as to de- flect it upward along the mountain side. In addition to this motion, there is an upward motion along the mountain side, due to the action of the sun's rays in the warming of the ground along the slope and the air immediately above it ; so that the upward motion of the air along the surface of the ground is the result of these two motions combined. The accumulation of this air at the top of the mountain is prevented by horizontal currents which exist there. MISCELLANEOUS WINDS. Under this title we shall mention certain winds which have distinctive features, and have therefore received separate names. Some of them belong to the great pri- mary circulation of the atmosphere, and some to the second- ary circulation, and others are of local origin. WALDO METEOR. 1 6 264 ELEMENTARY METEOROLOGY, Cold and Hot Winds. When the cyclonic and anti- cyclonic conditions of the middle latitudes are such as to carry to warmer regions the air which has become exces- sively cooled by local causes, there arise the cold winds which have received their distinctive names in the various regions in which they occur. When the conditions are such as to carry to colder regions air which has been made excessively warm by local causes, warm winds are experi- enced. The warm winds usually have a poleward and the cold winds an equatorward movement, but either wind may have a motion in other directions. Cold Winds, The cold north and northwest winds of the United States, called in the south northers and in the north blizzards, the bora of the Adriatic, the mistral of Mediterranean France, the buran of Russia, and the pam- pero of South America from the southwest, 1 are due to cur- rents of cold, dry, dense air which flow from the cooler to the warmer latitudes under the control of the secondary atmospheric disturbances. These cold winds occur usually when the temperature changes are very rapid within short geographical distances, and consequently the cold air need not be blown far to reach regions of greater warmth. They consequently occur in winter, and most frequently in those regions where the temperature gradients are steepest. The extent to which they penetrate the warmer regions depends on their persistency and the velocity as well as the volume of the moving mass of cold air. When the region from which these northerly (in the northern hemisphere) winds blow is very much elevated, as for instance is very markedly the case along the north- ern shores of the Adriatic Sea, then the air which has be- 1 It must be remembered that in the southern hemisphere the polar winds come from the south. LOCAL AND MISCELLANEOUS WINDS. 265 come excessively cooled by radiation on these highlands is partly blown, and partly descends from its own greater den- sity, on to the low lands or sea below; and, notwithstanding the adiabatic warming, it reaches the low levels still much cooler than the surrounding air, which it would not do un- less there were steep horizontal temperature gradients in the air. Thus are formed the bora of the Adriatic, and the tramontane* negra or black norther of Greece. Hot Winds. The hot winds from a southerly direction, the sirocco of the central and western Mediterranean Sea, the leste of Madeira, the kahnisin of Egypt, the leveche of Spain, or our own hot winds of the central and eastern United States, and the zondas of the South American pampas from a northerly direction, 1 are likewise due to the influence of the secondary atmospheric disturbances of middle latitudes, and are the reverse of the cold winds above described. The air is blown from the excessively heated regions to the cooler. Sometimes it is dry, as in the case of the dry sirocco of Spain and occasionally of Italy and Sicily, and the hot winds of our great plains ; and sometimes it is very moist, as in the case of the usual Italian or Sicilian sirocco. In the latter case the very dry hot air is blown from the Sahara and the northern coast of Africa across the Mediterranean Sea, where it takes up the moisture to almost the point of saturation ; and then it reaches the northern shores of the sea very warm and very moist, which renders the air very oppressive. Foehn Winds. The foehn winds, so called by the natives in the Alps, where they frequently occur, are warm dry winds or hot waves peculiar to some mountainous re- gions, and are the result of the following process : The 1 In the southern hemisphere the equatorial winds come from the north. 266 ELEMENTARY METEOROLOGY. warm moist air is blown against the side of a mountain range, and a not necessarily rapid upward air current arises along the slope ; and condensation usually takes place either by the cooling of the air coming in contact with the colder mountain top or by adiabatic cooling of the ascending air, or by both. When the moisture has thus been abstracted from the air in its passage over the mountains, the air, which begins its descent on the other side of the mountain range, is much drier; and in this condition its adiabatic increase in temperature as it moves downward is much more rapid than the adiabatic cooling in its ascent, for the latter was retarded by the liberated latent heat when the moisture of the air was condensed ; and with the descent the relative humidity decreases. Thus, when it reaches low altitudes, the air becomes very warm and very dry. The decrease in temperature for saturated ascending air is only about half as much per 100 feet as the increase of temperature for the dry descending air which is subject to dynamic heating by compression. The greater the amount of moisture lost out of the air, and the farther the descent of the air, the hotter it becomes, and the more intense are the foehn characteristics. The degree of intensity of heat in the foehn depends on the amount of water lost by condensation high up in the mountains, and on the distance of the descent of the air. If no water were lost out of the air, there would be no foe/in. One peculiarity of these foehn winds is that the descent of air fre- quently, or perhaps generally, occurs in isolated or disconnected streams, and thus the resulting excessively heated condition of the air below is not continuous, and is found only at intervals ; the intervening air not being much, if any, above a natural temperature. Winds of a. foe/in-like char- acter are found to the leeward in most mountain regions where warm moist winds blow against mountain ranges. The hot Chinook winds of LOCAL AND MISCELLANEOUS WINDS. 267 the western part of America at the eastern slope of the Rocky Moun- tains, and on the plains at their base, are due to this foe/m process, as this drying and warming of moist winds is called. They are especially frequent near the northern border of the United States, where the ex- cessively moist air from the Pacific Ocean coast of Oregon and Wash- ington is carried inland, and it is probable that they extend southward even to Texas. A single case, which occurred at Fort Assiniboine, Montana, on Jan. 19, 1892, may be mentioned, in which the temperature rose about 43 F between 2 A.M. and 2.15 A.M., changing from 5.5 F. to 37.5 F. within these 15 minutes. In some cases the temperature rises 80 F. in the course of 6 or 8 hours. It is very probable that the foehn process takes place in the free atmosphere, where air currents ascend, and some water is lost by rainfall ; and then the same air enters a descending current, and is brought down to or near the ground again. This would naturally occur in the circulation between cyclones and anticyclones, where the air ascends in the former, and, after losing some of its moisture, descends in the latter, and reaches its former al- titude warmer than when it started. The same reasoning also applies to the case of the rising and falling air in the local thunderstorm and tornado phenomena. The effect of the hot, dry foehn winds is very marked. When they occur in the cold season in regions where snow falls, the snow disappears almost as if by magic, and these warm winds are thus sometimes called snow eaters. When they occur in the warm season, vegetation is frequently withered, and growing crops entirely destroyed. Avalanche Winds are the movements of air masses which are pushed ahead of masses of land or snow as they de- scend mountain sides in land or snow slides. The mass of air is compressed and becomes locally denser-as it is pushed ahead of the moving substance. This air movement is 268 ELEMENTARY METEOROLOGY. of such violence that it does great damage even at a distance of many feet from the solid mass of snow or earth moved in the avalanche. In the case of a snow avalanche, the limits of these winds are well shown by the cloud of snow particles which they carry forward. The effects of such winds have been especially noticeable in Switzerland, where trees have been broken off by the wind at a distance of 1,500 feet from the mass of snow, and the particles of snow driven to a distance of more than a mile from it CHAPTER XI. WEATHER AND WEATHER PREDICTIONS Main Features. By the term iveather we mean the atmospheric conditions as shown by the meteorological elements at a particular time or during a short specified period. Climate is the aggregate of weather conditions, and weather is but a phase of climate. Thus we might speak of the weather at any instant or for a day, season, or even a year, but not for such a long period of years as would give the average conditions. These last with their oscillations would be climate. The degree of heat, the amount of moisture, precipitation, and cloud, and the di- rection and force of the winds, are the main features of our weather. There are certain combinations of conditions which occur together. Thus, in general, heat and wetness, and cold and dryness, go hand in hand ; although in middle and polar latitudes, where the seasonal changes are most marked, in the summer time cold and wetness, and in the winter time cold and dryness, go together. Absolute and Relative Weather Conditions. In the different localities on the earth's surface we get used to certain normal or average conditions of weather, and the ordinary or average diurnal changes where such exist ; but any accidental deviation from these average conditions and changes attracts our notice. Thus, when we speak of a hot day, it usually means hot in relation to the average temperature at that hour and 269 2/O ELEMENTARY METEOROLOGY. season of the year. A condition which we should term hot at one hour of the day might be cool when measured by that usual at another hour. Thus, in summer a temper- ature of 80 F. at eight o'clock in the morning would seem hot ; but the same temperature at two o'clock in the after- noon would not seem hot. The expressions which we apply to heat and cold are relative in their significance, and we make allowance for the diurnal changes. This is not so much the case when we speak of the wet or dry weather, for then we have the absolute and distinct dividing line between precipitation and no precipitation. The dampness of the air without precipitation is esti- mated on a scale reaching from extreme dryness with a clear sky to the dense cloudiness of fog. The wind ranging from a calm to a storm is a prominent feature of the weather ; and it is partly noticeable in its direct effects, as when it moves branches of trees, and partly in its indirect effects, as when it causes rapid evapo- ration or blows the cold air through our clothing, and makes the cold more readily felt. In going from a region with one climate to that with another, or, as we say, from one climate to another, one must become accustomed to the average conditions in the new climate before he can judge of the weather and its changes by the same standard as that customarily applied in that region. Seasonal Weather. During long periods of time which take account of average conditions only, the weather con- ditions are seasonal ; and although there may occur during any one season marked departures from the averages of many seasons, yet these irregularities are small as com- pared with those which occur during short intervals of time. WEATHER AND WEATHER PREDICTIONS. 2/1 If during some particular season the average temperature is, say, only 5 F. higher, or the rainfall a few inches greater, than usual, then we speak of it as being an exceptionally warm or wet season. The study of these long-period weather conditions belongs more properly to climatology. Current Weather Conditions are studied in a very different manner from the average conditions just mentioned. The true cause of the peculiar weather conditions at a particu- lar time at some one place, and the relation of these con- ditions to those existing at the same time at adjacent or even quite distant places, have been found out by writing on maps of extensive regions the meteorological conditions existing simultaneously at a great number of places in- cluded on the maps. Such maps are called weather maps ; and they not only permit us to see what kind of weather actually existed at the time the observations were made, but, as we shall see, they also enable us to say with a considerable degree of certainty what the weather has been for a day or two before, and what it will be for a day or two after, that time ; and this last we call weather predicting. In the present chapter we shall mainly consider these weather maps and the methods of using them for predict- ing the weather, or for weather forecasting ; and we shall confine ourselves to the consideration of the middle lati- tudes of the northern hemisphere, as these are the regions in which we are most interested, and where such studies are mainly carried on. It has been found that our weather in these latitudes is mainly conditioned by the phenomena attending the passage of the cyclones and anticyclones moving in the general direction from west towards the east, or sometimes from the south towards the northeast, as already men- (272) WEATHER AND WEATHER PREDICTIONS. 2/3 tioned. These phenomena, and their relation to cyclonic and anticyclonic areas, have already been described in preceding chapters, where the distribution of the meteor- ological elements in cyclones and anticyclones is given. Weather Conditions in a Cyclone and an Anticyclone. Fig. 86 shows the combined weather conditions of a cy- clonic and an anticyclonic area. On the front (eastern) side of the cyclone we see at first high cirrus clouds moving toward the east or northeast, and towards the center lower clouds moving from the east or southeast, with increased cloudiness and rain ; the air pressure diminishes, but the temperature does not change much. Near the center of the cyclone the rain ceases, and cloudiness diminishes. On the rear (western) side of the center there is some cloudiness and perhaps showers, and proceeding from the center the temperature falls, the air pressure increases, and the wind blows from the west and northwest. At the south side of the cyclone, proceeding from the front (east) to the rear (west), the winds veer around from southeast through south to the west, and the weather changes from rainy to clearing, or showery before finally clearing with approach to the anticyclonic area. The tem- perature is at first quite warm and constant, but with the shifting of the winds to west and northwest, it rapidly becomes colder. At the north side of the cyclone, proceeding from the front (east) to the rear (west), the wind changes from east through north to northwest. With the east and northeast winds there are clouds and rain, but clear weather with north and northwest winds. The temperature falls rapidly with the beginning of the north wind. For the anticyclonic area there is usually clear, cool weather, with north and northwest winds on the front side, 2/4 ELEMENTARY METEOROLOGY. and east winds on the rear. On the south (passing from east to west), the winds shift from northeast through east to southeast, the temperature at first falling, and then rising ; while on the north they shift from northwest through west to southern, the temperature not varying much. On the west side (passing from north to south), the winds shift from southeast to northeast through the east, the temperature grows warmer, and the clouds of the following cyclone begin to make their appearance. Methods of Making Weather Predictions. -- Weather predictions are made, first, from local observations, and refer to the place at which they are made ; second, from weather charts of an extended region, and refer to any region on the chart ; third, from weather charts and local observations, and refer to the region at and around the place at which the local observations are made. Weather Predictions from Local Observations. It has been found that the probability of the continuation of existing weather is greater than that for a change of weather ; and so we may assume that the present state will continue, unless some phenomenon presents itself which foretells a change in the weather. We have such phe- nomena in the movement and condition of the atmosphere, as shown by the direction of the wind, and kinds of the clouds. If we take as a standard or normal condition dry weather and a clear or but slightly cloudy sky, then the change to generally wet weather and an overcast sky is heralded by the appearance of the cirrus, cirro-cumulus, cirro-stratus, and stratus clouds which almost invariably precede, or we may say outrun, a cyclonic area. The next point is to locate with some accuracy the center of the approaching storm. This can be done by placing the back towards the direction of approach WEATHER AND WEATHER PREDICTIONS. 2/5 of the surface wind, when the region of low air pressure (cyclonic area) will be on the left and a little in advance, and that of higher pressure (anticyclone) on the right and a little in the rear, of the position occupied by the observer. The latter can then make a guess as to which of the quadrants of the storm area will pass over his lo- cality ; and, from a knowledge of the general distribution of the meteorological elements in the ideal cyclone or anti- cyclone, he can foretell with some degree of accuracy what the coming weather will be. For making local predictions for the morrow from the aspect of the sky, the evening at about sunset is a very favorable time. Observed Shifting of the Wind. In the case of a cyclone approaching from the west, if the wind gradually changes from the east around through the north to the northwest, then the cyclone center is passing on the south side of the observer, and he encounters in succession the weather conditions on the northern or left side of the cyclone ; . but if the wind changes from the east or south- east around through the south to the west and northwest, then the cyclone center is passing to the north of the observer, and he encounters in succession the weather conditions on the southern or right-hand side of the cyclone. In the case of a cyclone approaching from the south, when the wind changes from northeast around through the west the observer is on the left or western side of the cyclone ; but when the wind changes around through the east, the observer is on the right or eastern side of the cyclone. When the observer's position lies in the path of the center of the cyclone, the wind at first remains steady in direction, but almost ceases with the passage of the center, and immediately afterwards blows from the proper quarter 2/6 ELEMENTARY METEOROLOGY on the following side of the cyclone, without the gradual shifting of direction ordinarily noticed. A rapid and great fall in the barometric pressure indicates the near passage of the center of the cyclone. Introduction of Weather Charts and Storm Warnings. About the middle of the eighteenth century it was dis- covered that great atmospheric disturbances known as storms occurred first in the south and the west of the American colonies, and moved in a northeasterly direction over them ; but the portrayal of the construction or mech- anism of these atmospheric disturbances was not shown until after the first third of the nineteenth century, when, by means of charts showing the simultaneous distribution of the meteorological elements over a large ' region of country, the existing conditions were brought out, and the movement of translation over a limited portion of the earth's surface could be followed. The introduction of the telegraph made it possible to collect meteorological data from a large section of country in time to make it of use in following the weather changes over the whole region at the time the events are actually taking place, and also to transmit storm warnings in ad- vance of the approach of the storm. Systematic work of this kind was begun in a limited way about the middle of the nineteenth century, and during the last quarter of the century most civilized countries have had regularly estab- lished services for carrying on a work of this nature. Use of Weather Maps for Weather Predictions. Briefly stated, the usual method of using weather maps for pre- dicting the weather is as follows : The charted meteoro- logical observations, and especially observations of the air pressure, show us the location, on the map, of cyclones and anticyclones, and the accompanying weather conditions. WEATHER AND WEATHER PREDICTIONS. 277 These pass over the country from the west towards the east (or sometimes from the south towards the northeast) ; and, when we see the rate at which they are moving, it can be calculated about where they will be at any time during the next 24 or 48 hours. Thus we follow the movement of cyclones and anticy- clones and their accompanying weather conditions across the country in much the same manner that we can follow the movements of a railroad train if we know its time and place of starting, and its route and speed. But the cyclones vary so much, in intensity, in the paths which they take, and in the velocity of movement, that their positions and conditions can usually be foretold only day by day. Once having fixed the position of a cyclone or anticyclone with regard to any place, we know the general weather conditions at this place as shown by the distri- bution of the meteorological elements in cyclones and anticyclones. We have now to see exactly how weather charts are constructed, and then more particularly how they are used for making weather predictions. Simultaneous Meteorological Observations for use on weather maps are made at certain hours of the day (at present in the United States at 8 A.M. and 8 P.M., 75th meridian time) by regularly appointed observers at the government meteorological observing stations distributed throughout the country. The results of these observations are at once telegraphed to some central office, where specially appointed 'officials receive them, and by means of them make charts of the condition of the various meteorological elements, and the weather prevailing over the whole region from which such reports have been received. 278 ELEMENTARY METEOROLOGY. There are over a hundred observing stations in the United States, and perhaps a dozen more in Canada, from which complete reports of this nature are received by telegraph at the United States Weather Bureau office at Washington twice a day just after the observations are made. In England such reports are received at London ; in France, at Paris ; in Germany, at Hamburg and other cities ; in Austria, at Vienna ; in Russia, at St. Petersburg ; etc. Construction of Weather Maps. In the construction of weather maps there is used as a basis an outline geograph- ical chart of the whole region from which meteorological data are received by telegraph. Air Pressure. The barometric pressures reduced to sea level are written down at the stations to which these belong, and then the isobaric lines are drawn. These full, heavily drawn lines (which are placed on United States weather charts at intervals of one tenth of an inch air pressure) have marked on each in large figures the baro- metric pressure which they represent, and they show at a glance where the cyclonic and anticyclonic areas are located. The cyclone centers are then marked in large letters LOW to signify low barometric pressure. The anticyclone centers are marked HIGH to signify high barometric pressure. These isobars are the most important feature of the weather map ; and a person skilled in the study of such maps could make good weather predictions from a chart containing them alone, as he is famil- iar with the usual distribution of the meteorological elements in and around cyclones and anticyclones. Temperature. On this same chart the observed tem- peratures are written down at the stations to which they belong. Isothermal lines are then drawn at intervals of 5 or 10 F. usually. These isotherms are heavy dotted lines, and each has marked on it in large figures the proper temperature. WEATHER AND WEATHER PREDICTIONS. 279 These dotted lines show at a glance the regions of cold and warm air. Where they bend towards the equator, we know that the local air is relatively cold ; and where they bend towards the pole, the air is relatively warm. On the American maps, regions in which marked changes of temperature have occurred during the past 24 hours are in- closed by a heavy dotted line ; predicted cold waves are marked C. W. Wind, The wind's direction is indicated at each station by drawing an arrow through the center of the station ; the arrow flying with the wind. The velocity of the wind is sometimes written down in miles per hour beside the arrow ; or lines may be drawn in the tail of the arrow to indicate feathers, the greater number of lines signifying the greater wind velocity. Wind velocities vary too greatly to easily permit the drawing of lines of equal velocities. Moisture. The aspect of the sky, and the amount of moisture as indicated by the degree of cloudiness, are shown on the map by a circle drawn around each station. Thus a O is made around the station when the weather is fair and the sky is clear, and the space within the cir- cle is filled up as the degree of cloudiness increases ; and when it is raining, the circle is entirely filled up. In the American maps a bar is drawn through the circle ((D) when the sky is half covered with cloud. A small white center is still left (O) when the sky is entirely covered with cloud ; and the circle is entirely filled up () when it is raining. When it is snow- ing, several bars are drawn through the open circle ((). The arrows showing the direction of the wind pass through the centers of these circles. Such symbols are subject to change. Also on the American charts, the regions in which rain has fallen in the past 12 hours are shown by a slight shading on the map. In cases of cyclones with low pressure and correspondingly strong winds and much rain, the previous path of the cyclone over the chart is marked by an arrow-headed line, and on it are placed circles with a cross (-0-) to indicate the location of the center of the cyclone at the previous times of observation, which are shown by the attached dates. WALDO METEOR. 17 (280) (28 1) 282 ELEMENTARY METEOROLOGY. Such a weather map shows, first of all, the absolute and relative positions of the areas of high barometric pres- sure (anticyclones) and low barometric pressure (cyclones), which, in the main, control the characteristics of the weather in our latitudes. It shows the area dominated by the weather conditions pertaining to each. It also shows the magnitude of the excess and deficiency of pressure on which the intensity of the resulting condi- tions depend. Tabular data accompany most maps. In order to present the idea of weather charts more clearly, there are given here (Figs. 87 and 88) two actual maps for the United States, showing the simultaneous weather conditions in midsummer and again in midwinter, the two extreme seasons of the year. Very pro- nounced specimens of the winter and summer types of weather have been selected in order to bring out their characteristics. These maps will be comprehended from the foregoing details concerning their con- struction, and they should be carefully studied. Scope of Weather Predictions, Weather predictions are made either for a single place or for a whole region ; and they are made for a definite time, say for 12, 24, or 36 hours in advance, or else the weather changes are designated which are gradually to occur during the com- ing 12, 24, or 36 hours, It is best at first to select some definite point for which the weather is to be predicted ; and the best one will ordinarily be that at which the person making the prediction lives, because it will then become a matter of practical interest. After some experience has been thus acquired, predictions can be made for some distant point, and still later for a whole region. Irregular Movements of Cyclones and Anticyclones. These atmospheric disturbances move (in our latitudes) gen- erally in an easterly direction, and the probability is that the velocity of translation for a few hours preceding any WEATHER AND WEATHER PREDICTIONS. 283 definite time will remain the same for a few hours after this time ; that is, if the cyclone or anticyclone has moved at the rate of 25 miles per hour during the past 12 or 24 hours, then it will probably continue so to move during the next period of like length. Some weather predictors consider it best to take as the probable future velocity the average velocity of the cyclones, or some compro- mise between it and the velocity during the hours just passed. Cyclones entering our continent from the west, and cross- ing it, frequently move at first in a southeasterly direction, and then in a northeasterly direction ; and those which enter the continent from the south or southeast usually move off in a northeasterly direction. Cyclones have a tendency to move slightly spirally around to the right of areas of low temperature, and they have a tendency to drift in the direction of the main atmospheric motion, as shown by the direction of the higher clouds, and to move in the direction of greatest rainfall. When areas of high barometric pressure (anticyclones) and of low pressure (cyclones) occur simultaneously within a limited region, as on the same weather map, they seem to mutually affect each other's condition, behavior, and move- ments. When an anticyclone is on the north or east side of a cyclonic area, there is a tendency for the latter to be retarded and deflected towards the north ; but if the anti- cyclone is to the south or southeast of the cyclone, the movement seems to be accelerated. Where one cyclone follows another closely, the one in the rear seems to have a retarding and deflecting effect on the one in front After having decided on the probable future path of the cyclones and anticyclones, and their accompanying condi- tions shown on the weather map, the making of a weather 284 ELEMENTARY METEOROLOGY. prediction consists in following out the changes which will take place when the existing conditions are moved along to the places at which they will be if the movement takes place as expected. All of the existing observed conditions may be just moved along over the stationary map in the most frequented path, and at the average rate of move- ment of the cyclones, and we can see on the map where these conditions will fall. The conditions actually do change, but it requires great skill on the part of the predictor to foretell these changes. Variations in Conditions on Weather Charts. On the first consideration of a weather map, and recognizing that it is made up of areas of high and low air pressure, it might be quite natural to think that in short intervals of time, say every few days, the maps would repeat themselves, and that many would be found quite identical. But this is not the case, for a great variety of combinations exists, and two maps are seldom alike. Still it is possible to class weather maps in a general way according to certain types, but the number of these types is very great. Probably over 100 typical forms would be necessary for such classification ; some one of them might occur only at intervals of several years, while others would occur several times in a year. It has also been found that the behavior of the cyclones and the anticyclones is not the same on maps which appear to be almost exactly alike ; so that, because they follow certain movements in one case, it is by no means certain or even probable that the same will occur for other cases of the same type of form and arrangement. It is this variability under apparently like circumstances which pre- cludes a much further advance in accuracy of weather predictions by the methods now adopted. Weather Predictions for Different Regions. It must be WEATHER AND WEATHER PREDICTIONS. 285 assumed, in the case of general weather predictions which are made for a whole continent from a single point (as at Washington), that there will not be much change in the distribution of the weather elements in the cyclonic and anticyclonic areas as they traverse the continent ; but they are as a fact continually undergoing changes, which can seldom be anticipated at the proper time. As the cyclones and anticyclones move over the land, new air is brought into the circulation ; and it makes a great difference in the future conditions within the areas whether the air thus drawn in has a constant or varying supply of moisture. Any portion of the cyclonic area, as for instance the southeast quad- rant, may have certain definite conditions when the southwest wind which supplies it with air comes from a dry region (as, for instance, in Colorado) ; but let the area progress eastward, and the air be drawn from a moister region (say, the Ohio River valley), and the weather would not be of such a nature that a prediction for the dry region (Colorado) would fit the existing conditions in the new locality. Let the storm progress still farther, so that the southeast quadrant would be on a warm moist coast (say, on the southern coast of New England), then the warm moist air entering it (from the Gulf Stream region of the Atlantic Ocean) would cause still other conditions to arise. The meteorological and topographical characteristics of different regions must be studied very closely by the person making predictions for them. Combination of Weather Map and Local Meteorological Conditions. It is now a recognized fact that a person on the field of any region can, with the aid of weather maps, make more accurate weather predictions for that region than a person at a distance, who is unacquainted with the local characteristic weather, and relies entirely on the weather map. In addition to the general conditions of en- vironment, there are many local peculiarities which cannot 286 ELEMENTARY METEOROLOGY. be taken into account in telegraphic reports, and which greatly affect the weather conditions at a place. It seems best, therefore, to have, for each geographical section where the conditions are similar, a local weather predictor who shall receive telegraphic weather reports, and have the use of maps constructed therefrom, and who shall modify the predictions based on this by the local conditions. This method is now adopted by the United States Weather Bureau ; and it seems to be more successful than the older method, where one predictor at Washington had to furnish from the weather map alone the forecasts for all the various regions of the whole country. This combined method has been in use in Europe since the introduc- tion of telegraphic weather services. There the various countries unite in the interchange of the observed data, so that the predictor in each country can have available the whole continental distribution of mete- orological data ; but the official predictor for each country makes the weather forecasts for his own country alone. The form of weather maps varies in the different lands, although the data used in all are about the same. This makes the predictions in each case (except for Russia) apply to a limited region only ; and this is even more necessary in Europe than in America, for in the former the weather predictions are much more difficult to make, on account of both the variability of the paths of the cyclonic areas in high latitudes, and the fact that the atmospheric disturbances come from the ocean to the westward, whence telegraphic communications are not available. Thunderstorm Predictions. Thunderstorms and torna- does belong to that class of local meteorological phe- nomena that may be predicted for a general region, but it is impossible to predict their formation at any particular place. When, however, they have once formed, a knowl- edge of their presence can be transmitted by telegraph or telephone to points which lie to the eastward or northeast of them, and which lie in their probable paths. Undoubt- edly more harm than good would be done by predictions of WEATHER AND WEATHER PREDICTIONS. 287 tornadoes for whole regions in which they are likely to occur. In the case of thunderstorms, however, they trav- erse such a great section of country, that, when it is found that they are active in a certain section of a cyclonic area, their probable occurrence in the same relative position to this area may be predicted in its farther course ; but for any region lying in their path nothing more definite can be said than that local (here and there) thunderstorms may be expected. Prediction of Cold Waves. By cold wave is meant the fall of temperature over a large area from one day to the next; and this change must be more than temporary and local, for the fall of temperature which occurs during a thunderstorm would not be called a cold wave. Cold waves are nearly always connected with cyclonic and anticyclonic disturbances. They almost invariably occur to the west of a cyclonic and to the east of an anticy- clonic area. While cyclonic areas usually move easterly or northeasterly, cold waves generally advance from day to day in a southerly or southeasterly direction. The greatest temperature fall during 24 hours is usually quite near to the center of low pressure in the cyclone, and is most likely to occur a little to the south or west of this center. The direction of most rapid change in the tem- peratures within a cold wave is usually towards the north- west of the center of the cyclone. A change of 60 F. in the temperature from one day to the next (24 hours) was observed in the United States only twice from 1880 to 1890 ; and a change of from 50 to 60 F., on but 16 days in 10 years. Cold-wave Areas, which may in extreme cases cover 1,000,000 square miles, are usually elliptical, with the long axis extending from southwest to northeast, and parallel to 288 ELEMENTARY METEOROLOGY, the long axis of the cyclone in front of it. While the length and breadth are sometimes nearly equal, yet the one axis may be eight times as long as the other. In the United States Weather Bureau, the name cold wave is given when the following temperature changes take place in 24 hours (for instance, from 8 A.M. one day to 8 A.M. the next day): in the northwestern States, from Minnesota to Montana, a fall of 20 F., and temperature going as low as 32 F. ; in the central region, extend- ing from Colorado to Maine, and as far south as Tennessee, a fall of 1 8 F., and temperature going down to 34 F. ; in the southern United States, a fall of 16 F., and temperature going down to 36 F. The approach of a cold wave is announced by the United States Weather Bureau by the display of a square white flag with a square black center. Prediction of Hot Waves. Hot waves may be referred to their three possible causes : 1. The ground and the lower layers of air may be heated to an excessive degree by the continuous action of the solar rays during successive clear days ; and this accumula- tion of heat becomes greatest when the days are clear and the nights are clouded. These hot waves are of local char- acter, but may be very extensive. 2. The air of vertical currents may have its temper- ature raised by the foe/in process. Such conditions may be expected on the leeward side of mountain ranges, where the air on the windward side is moist. These hot waves are usually confined to narrow bands parallel to the moun- tains. Similar hot waves also occur in anticyclones, where the descending current is fed by air which has moved up- wards in a cyclone, and been deprived of a portion of its moisture before it descends again. Such hot waves follow the course of the anticyclones. 3. Warm air may be blown from warmer to colder regions by the continuation of the winds in the proper WEATHER AND WEATHER PREDICTIONS. 289 direction during an interval of several days. Such heated conditions of the air may be expected, in middle latitudes, just preceding the great cyclones and following the anticy- clones; for in both of these cases the air is blown from lower and warmer to higher and colder latitudes some- times for days in succession. These hot waves are of greatest extent. Predictions of Hurricanes. In the hurricanes which come from the West Indies, and skirt the eastern coast of the United States, the storm center sometimes moves along over the ocean, very frequently follows the Gulf Stream, but sometimes penetrates the Gulf of Mexico ; or, pursuing a course somewhat inland, it ravages the whole Atlantic seaboard. Such storms require special predictions. In the incipient stage of the hurricane there is but a small area affected by it, and at this stage no characteristic signs of its approach precede it at any great distance. It is therefore very necessary for the official weather predictor at Washington to receive as early as possible a telegram from the West Indian Islands, announcing the existence of the hurricane ; and additional telegrams should be received stating its progressive movement. The storm center moves so slowly that it may be several days after its announcement before it reaches the coast waters of the United States ; and very frequently the wave movement imparted to the ocean water by its violence will reach the shores of the United States before the storm itself is noticeable on the weather maps. Often the passage of such a hurricane far out at sea is made manifest by the waves which it causes to lash the coast hundreds of miles from the storm center, while no trace of the latter is visible on the land weather maps. The successful predictions of the movement of the hurricanes after they have entered the region covered by the weather maps, is a matter of much difficulty, for their exact direction of motion is too uncertain, and their rate of progress is too variable, to foretell the exact location of the storm's center. The warnings which navigators, especially of small coasting vessels, receive concerning these storms is of incalculable value to thenio 2QO ELEMENTARY METEOROLOGY, When the East Indian typhoons approach the Chinese and Japanese coasts, almost unheralded except at telegraphic points, the loss of life and damage to shipping is very great. Predictions of River Floods. During a river flood there is a period during which the water rises, another when it is at its greatest height, and a third when it falls. A flood may then be viewed as a progressive wave which has the culminating point of the flood for its crest. The movement of this crest is followed on the map by the predictor as it proceeds down the river valley ; and the time of its appear- ance is predicted for points below by means of its speed as determined by the velocity , of the current at the highest point of the flood. Such predictions can be made usually several days in advance for places lying on the lower part of the river course. When a river is fed by several branches, it is necessary to predict the flood conditions of each effec- tive branch, and then combine these predictions in order to determine the flood conditions for the main stream below the point of union of the branches. Frost Warnings (Cold- wave Frosts). When the tem- perature is high and a cold wave makes it fall considerably below the freezing point, then one kind of a frost occurs ; and it will be very general and of a wide distribution, and may begin day or night. When the air is dry, and the temperature does not descend below the dew-point tem- perature, as in many cases of cold waves, little damage is done by the cold, even if the temperature goes below the freezing point of water ; but when the dew-point is high, and dew is deposited and then freezes, the damage to plant growth may be very great, even for a temperature at or just a little below the freezing point of water. Night Frosts. The more local frosts are those which, under certain conditions, may occur during the usual mini- WEATHER AND WEATHER PREDICTIONS. 29 1 mum temperature at night. Briefly stated, it may be said that night frost is to be feared when observations of the temperature and humidity show that the dew-point lies below the freezing point, and it is expected that the tem- perature will descend below the dew-point. Storm Signals. In all civilized countries where weather services are maintained, special signals announcing the approach of storms are displayed at seaports for the bene- fit of mariners. These signals are usually flags by day, and lanterns by night. Distribution of Weather Predictions (especially in the United States). After the official government weather predictor has decided on the probable weather predictions for the next 24 or 36 hours, this forecast must be placed before the public. This is accomplished in a number of ways. Daily weather maps such as have been de- scribed are mailed from various centers to those places which can be reached in time to be of any use, and these are displayed in public buildings, such as post offices. The telegraph and telephone services are more and more used for distributing weather predictions to distant and local points, and especially in transmitting them to all of the great daily newspapers of the country. It is through these daily papers that the widest-spread knowl- edge of the weather forecasts is obtained. Use is also made of the frequent railroad trains in some regions to display on them flags which shall communicate to the inhabitants near the railroad the coming weather conditions ; and the railroad telegraph conveys the forecasts to each station on the route, where they are posted up for public information. The press telegraphic organizations transmit the predictions to the city newspapers. Accuracy of Weather Predictions. -- The accuracy of weather predictions is measured in percentage of their 2Q2 ELEMENTARY METEOROLOGY. complete success. It is estimated that, on the average, from 80% to 85% of weather predictions are successful. The accuracy varies considerably, not only for the differ- ent meteorological elements, but also for different regions of a country. Those regions are the most difficult to pre- dict for which have the greatest and the most sudden weather changes. It is easier to predict the weather during the season of settled weather or least rain than when rainfall is most frequent. It is also easier to predict temperature changes within certain fixed limits in the summer than in the winter. The predictions of wind velocities and cold waves are about the least successful, averaging only about 60%; while the wind direction is the most successful, averaging over 90 %. The weather predictions for the California coast are the most successful of any in the United States. Long-range Weather Predictions. While it is possible in our latitudes to make predictions concerning the weather for the next 24 or even 36 hours, and in special cases even longer, yet, in general, the making of definite forecasts for successive days, or for any specified date several days ahead, cannot be successfully accomplished. All long-range predictions, such as for a coming season, are groundless ; and the prediction of any great storms for certain dates far in the future is utter nonsense, as no one can foretell the appearance of a particular storm until it actually begins to form. It is possible, however, to state the average weather conditions which have been found to exist at different periods during the year. CHAPTER XIL CLIMATE. Climatic Conditions. The term climate signifies the aggregate or average of meteorological conditions. The climate of a place can be determined only by a series of observations of the meteorological elements, carried on through a period long enough to give the average con- ditions, freed from the irregularities due to accidental weather conditions, and to determine the average and possible extreme departures from those average conditions. Weather is but a phase of climate extending over a not too long definite interval of time. We speak of the meteorological conditions for a single year, or a season, or a day, as weather. Thus we say, " last summer the weather was wet," and not, "the climate was wet." We could not say that the climate was wet during the summer, unless "it was so during the majority of summers. We should speak of the annual or seasonal climate, and perhaps also for the months, but not for shorter periods. Climatology is the science of climate, and its object is to present the average joint action (and possible variations therefrom) of the meteorological elements which pertain to climate at the different places on the earth's surface. Climatology is therefore mostly descriptive. The impor- tant climatological elements are the heat, movement, and moisture of the air ; precipitation ; evaporation ; cloudiness ; and solar radiation. 293 294 ELEMENTARY METEOROLOGY. The barometric pressure is of the greatest importance in influencing the distribution of the climatological elements ; but it is only in the limited regions of great altitudes that it can be considered one of them, and then chiefly on account of its effects on animal organisms, due to the decrease in air density. In studying climate, we must, then, study the conditions and relations arising from the combination of the average conditions of the above-mentioned elements, which have been studied separately in the earlier part of this book. The relative importance of the climatological elements, and the data usually required for each, are as follows : Temperature. 1. The monthly and annual mean air temperature. 2. The amount of the daily temperature oscillation in the single months. 3. The average and absolute monthly and annual tem- perature extremes. 4. The average and extreme dates when ice forms in the spring and fall, and the number of days free from ice. 5. The average variability of the daily temperature for each month and the year. It is desirable, also, to have the average monthly and annual temperatures at some morning, some midday, and some evening hour, say 7 A.M., 2 P.M., and 9 P.M. Moisture and Precipitation (for each month and the year). 1. The average relative humidity. 2. The average amount of precipitation. 3. The average number of days with precipitation. 4. The average intensity of rainfall. 5. The average probability of rainfall. Cloud (for each month and the year). 1. The average degree of cloudiness. 2. The average number of hours with sunshine. CLIMATE. 295 Wind (for each month and the year). I The average wind velocity or force. 2. The average number of times the wind blows from each of the eight points of the compass. Evaporation. It is desirable to obtain the average monthly amount of water evaporated both from a free water surface exposed to the sun and all weathers, and from a shaded water surface ; but these are very seldom measured. Solar Climate. The ideal solar climate is that which would exist if the sun's rays reached a homogeneous earth without an atmosphere ; that is, an earth with a smooth surface all land or all water. The solar climate would have a distribution varying with the latitude from the equator to the pole, but would be the same at all points on a given parallel of latitude. There would exist climatic zones extending around the earth, and following the paral- lels of latitude. The Telluric or Physical Climate of the earth is the solar climate modified by the atmospheric conditions, and the existing distribution of land and water surface of the earth. The most important disturbing causes are the unequal distribution of land and water surfaces, and the different elevations of the land surface above the sea level. These give us the three chief forms of climate of telluric origin : i. The land or continental climate. 2. The water or oceanic climate, 3. The mountain climate. The fact that the parallels of latitude run partly over a land and partly over a water surface causes differences of climate in an east-westerly direction, transverse to those of the solar climate. These east-westerly differences of climate are reenforced by the transfer of equatorial heat towards the pole, and of polar cold towards the equator, by the currents of air and water. WALDO METEOR. 1 8 296 ELEMENTARY METEOROLOGY. The Land or Continental Climate is characterized by great extremes of temperature, relatively high temperatures^ during the summer (or during the daytime), and low temperatures during the winter (or during the night). The humidity, cloud, and rainfall are deficient. The Water or Oceanic Climate is characterized by slight temperature extremes : the temperatures remain relatively low during the summer (or during the daytime), and rela- tively high during the winter (or during the night). The humidity and cloud are excessive, and rainfall ample. Continental and Oceanic Climates are best considered to- gether, in order to compare the two extreme conditions. The Temperature. When the same amount of heat falls on land and water surfaces, the temperature of the land is raised nearly twice as many degrees as the water. The land, then, in summer and in the daytime, warms the air above it more quickly than does the water surface, but the latter gives up to the air above it, by evaporation, more moisture than does the land surface ; and when this moisture is con- densed in the higher air layers, it gives out heat to them. At night and in winter, however, the land cools more quickly than the water, and so the air temperatures over the water do not fall so low as those over the land. The greater moisture and cloudiness of the air over the water also prevent the loss of heat as rapidly as over the land. The temperature differences over a land surface in summer and winter are intensified by the decrease of cloudiness towards the interior of a continent, and these differences over a water surface are lessened by the greater cloudiness over the oceans. In higher latitudes a less degree of cloudiness lowers the winter temperatures and raises the summer temperatures. In lower latitudes a decrease of cloudiness increases the temperature* CLIMATE. 297 At about latitude 40, both north and south, the land and water surfaces have nearly equal temperatures. On the equatorial side of this parallel the land surface, and on the polar side the water surface, is the warmer, on the average, for the year. At the equator the temperature for a land surface would be about 113 F., and for a water surface 72 F. ; at latitude 45, both land and water would have a temperature of about 50 F. ; while at the pole the temperatures would be, for land 25 F., and for water 12 F. The average temperature of a parallel can be considered as dependent on the latitude and on the relative amounts of land and water. The Wind. The differences in temperature between the land and water surfaces give rise to the daily land and sea breezes on the coast, and to the seasonal monsoon winds extending from the interior of the continents far out over the oceans. The winds belonging to the cyclonic and anticyclonic circulation are not influenced very much by the land and sea surfaces, except that the winds become stronger over the water surface because the friction is less than over the land. The most important climatic effects of the wind are its transference of moisture from the ocean to the land sur- face ; of heated air to colder regions, and cold air to warmer regions ; and of air from higher to lower, and lower to higher altitudes, thereby causing adiabatic heating and cooling. The flow of air between the continents and oceans is much more general and also more rapid in winter than in summer, because the continents are cooled more in winter than they are warmed in summer. The cold winds blow from the continents in winter, and towards them in summer. 298 ELEMENTARY METEOROLOGY. The windward coasts of the continent are those towards which the wind blows from the ocean, and they have an oceanic climate. The leeward coasts of the continent are those from which the wind blows towards the ocean, and they have a continental climate. The region to the lee- ward of isolated large bodies of water, like the American Great Lakes, is warmer in winter and cooler in summer than the region to the windward ; because the water surface is cooler in summer and warmer in winter than the land, and the temperature of the air which approaches from the windward of a lake becomes tempered by the water before it reaches the land on the leeward side of the lake. The Moisture. The moisture is transferred by the winds from the oceans to the continents. The windward coasts or sides of the continents are therefore moister than the leeward coasts ; because on the windward side moist air is received from the oceans, while on the leeward side the air comes from the interior of the continent, where mois- ture is taken from the air more rapidly than it is added. The flatter and lower the surface of the land, the farther inland does the air penetrate towards the interior of a continent before it loses its excessive moisture. Where mountain ranges or other elevated lands lie in the path of the moisture-laden air, much of the moisture con- denses on the windward side, and causes a wet climate, while on the leeward side the climate is dry. This is most marked when the moun- tains lie close to the ocean, so that precipitation occurs before a large portion of the moisture has gradually been lost while passing over a large extent of land. Climatic Effects of the Oceanic Circulation. The great currents which exist in the oceans, by which the warm water of the equatorial regions is carried towards the pole, and the colder water is transferred from the polar regions CLIMATE. 299 towards the equator, very greatly modify the climatic con- ditions over the oceans and along coast lands. The warm current heats the air above it, and increases its capacity for moisture ; and the cold current cools the air above it, and decreases its capacity for moisture ; and the air thus heated or cooled is blown by the prevailing winds on to the continents, or distributed over the adjacent ocean sur- face. The temperature of the water is communicated to the land, only when the wind blows from the water towards the land. Effect of Vegetation, and especially Forests, on Climate. Local Effects. Temperature of the Soil. -- The general influence of the forest is, on the whole, to cool the soil. The extremes of temperature are reduced as compared with those of the open fields, but the effect is more marked (is greater) on the summer maximum than on the winter minimum temperatures. Temperature of the Air. The air temperature under the crowns of forest trees is, on the whole, cooler lower in summer, and higher in winter than in the open fields. The kind of trees composing a forest has an influence on the tem- perature, although the average for the year is about the same for all kinds of trees. For evergreen trees, the difference in the temperatures under the trees and in the open shows a symmetrical increase and de- crease during the year, being least in winter, and greatest in summer ; but for the deciduous trees the difference is variable, diminishing from midwinter to springtime, but increasing rapidly with the growth of leaves. In young forests the maximum air temperatures are lower, and the minimum highen, than in old forests. The air temperature within the crowns of trees is higher than for the same elevation above ground in the open fields. 300 ELEMENTARY METEOROLOGY^ The difference in temperatures within the crowns and below is more constant in evergreen than in deciduous trees. There is frequently in the woods, especially in summer time, a higher temperature above than below. The air temperatures above the crowns of trees are similar to those over a grassy meadow or cornfield, warmer than the air in the sun- shine by day, and cooler by night. The effect is greatest for evergreen trees and for deciduous trees in full leaf. The temperatures within the tree trunk are a little higher than those of the surrounding air in the early spring and late summer, but for the rest of the year are lower, Local Moisture. The absolute humidity within a forest slightly exceeds that of the open ; and the relative humidity is from 2% to 4 % greater, the excess being much more marked for evergreen than for deciduous trees. It is not known that precipitation is greater over forests than over the open, but the presence of trees retards the rapidity of loss of fallen rain. In middle latitudes the annual evaporation from a sur- face within a forest is about half that in an open field, and the maximum evaporation in the forest occurs in May or June. Influence of Forests on the Surrounding Climate. The water which the roots of trees collect from the ground, and which passes through the trunk and branches, is said to be transpired. The water which is evaporated from the tree crowns through transpiration is carried as vapor by the winds to the adjacent regions, and the moisture on the leeward side of forests is thus slightly increased. Local air currents arise due to the difference between the temperatures of the forest and of the surrounding country. Cooler currents in the lower air strata come from CLIMATE. 301 the forest by day, and warmer ones in the upper strata by night. The forests act as a wind break for the regions close by on the leeward side, and they also reduce the force of the wind in the lower air layers by the friction with the tree tops. This last would consequently reduce the evapora- tion, which depends so much on the wind velocity. When the forest forms a glade around a small open area of land, the climate of the inclosed space is not so extreme as it would otherwise be. In dry regions, when the water is stored underground, and prevented from coming to the surface by overlying hard strata of earth, then the deep penetrating roots of forest trees bring this moisture to the surface, and by tran- spiration give it to the air. Altitude and Climate. The climatic factors of a place due to its location as regards latitude, and continental or oceanic exposure are greatly modified by the altitude of the place above sea level. With increase of altitude, the air pressure, the temperature, and the absolute amount of moisture in the air, and above a certain altitude the rainfall, decrease ; while the intensity of the direct solar rays, the evaporation, the winds, and up to a certain altitude the rain- fall, increase. Since these variations are mentioned under the different meteorological elements in treating them separately, they are not dwelt on further here. Climatic Zones. It has been customary to divide the climates of the earth's surface, according to the solar cli- mate, into three zones. The tropical or warm zone lies between the tropics ; the temperate zone lies between the tropic and the polar circle, and extends from latitude 23 1 to latitude 66J ; and the polar or frigid zone lies within the polar circle, and extends from latitude 66J to the 3I.<)< IV regular progressive motion of these cold waves cause the existing regularity in the line of equal minimum tempera ture in the central and eastern United Stales. The extreme minimum of 63 F. obseived in northern Montana (Fort Assinihoine) is surrounded hy higher temperatures to the east, west, and south. Towards the west from Montana theie is at first a ;radual and then a moie rapid increase too" F. near the coast, and on the North Pacific coast itself the minimum is from -| X 1 to } 15 I 1 '- above zero, Towards the east from Montana there is a gradual increase to 25 F. or 30 F. on the Atlantic coast in the same hi-h latitude. Towards the south !mm Mont, in, i there is an Increase to 4-13 I 1 ', ill southern Texai; towards the southeast there is an increase to +41 I'", at the southern end of Florida; and towards the .soul Invest their is an inciease to +32" F. on the southwestern coast of ( 'alifoi nia. On the summit of Mount Washington in New Hamp- shire the minimum temperature leached is 50" F. at an altitude of about a mile', and on Pike's Peak, Col., about 40 F. at an altitude of about two miles and a half The Absolute Oscillation of Temperature (Max. -M in.) is shown in Fi<;. <)<>. Thisehait brings out verv plainly the- fact that the am- plitude of t he tempera! me depends to a \ it extent on the minimum temperatures rcaehed, and is but slightly influenced hy the maximum temperatures reached. The amplitud' it the interior <>f the continent, but decrease with approach towards the coast and with the latitude , although this latter decrease mav be almost masked on a windward CO n lor s i breadth of latitude as the width of the United Stales. On the leeward (eastern) coast t the I'uited states the conti CLIMATE OF THE UNITED STATES. 329 nental influence is shown by the decrease with latitude coming halfway between that for the interior of the conti- nent and the windward (Pacific) coast. FIG. 99. ABSOLUTE AMPLITUDE OF OSCILLATION OF SHADE TEMPERATURE IN THE UMJED STATES (AFTER GREELY). The Duration of Temperatures below Freezing in the United States is shown by the following chart (Fig. 100), on which is given the number of days when the average daily temperature was below 32 F The most southern line drawn on the map, and marked o, is the dividing line between the regions where the average daily tem- perature goes below freezing, and those where it does not. The next line, marked 30, passes through the localities which have 30 days with the average temperature below freezing. The additional lines are similarly drawn for the successive multiples of 30 days (i.e., months). This distribution follows in a general way that of the midwinter average (Fig. 95) and minimum (Fig. 98) tem- peratures already described. In northern Minnesota the 330 ELEMENTARY METEOROLOGY. number of days is 165, and there is a decrease from there in all directions in the United States. The decrease is very gradual to about 1 20 days on the northern New Eng- land coast on the east, but more rapid to o days on the Delaware coast on the southeast, central Texas on the south, central Arizona on the southwest, and central Oregon on the west. FIG. ioo. NUMBER OF DAYS WITH AVERAGE TEMPERATURE BELOW FREEZING IN THE UNITED STATES (AFTER GREELY). Variability of Temperature in the United States. The change in the average temperature from one day to the next varies greatly in the United States, not only dur- ing short intervals of time, but also on the average for each month from month to month. Such changes are much greater in some regions than in others. The variability of the temperature depends largely on the descent of the minimum temperatures, and follows quite closely the distribution of the minimum tempera- tures (Fig. 98). The variability is greatest in the interior CLIMATE OF THE UNITED STATES. 331 of the continent, but decreases towards the coasts, and towards the windward (western) coast much more rapidly than towards the leeward (eastern) coast. There is also a decrease towards the equator, scarcely perceptible on the windward coast, very marked at the interior, and between the two on the leeward coast. In general, the variability is greatest in midwinter (Jan- FIG. 101. VARIABILITY OF AVERAGE DAILY TEMPERATURE IN JANUARY IN THE UNITED STATES (AFTER GREELY). uary and February), and least in midsummer (July and August). The frequency and magnitude of these changes are mainly dependent on the number, rapidity of move- ment, and intensity of the areas of high and low baro- metric pressure which sweep over the country. The accompanying chart (Fig. 101) shows the average tem- perature variability for the maximum month of January. It is seen that the greatest variability occurs in the cen- tral interior region of the northern part of the United States, where it is about 10.5 F. Towards the west of WALDO METEOR. 2O 332 ELEMENTARY METEOROLOGY. this region of maximum change, the variability decreases at first slowly, and then more rapidly $ to only 2.5 F. on the northern Pacific coast. Towards the southwest the decrease is to 2 F., on the southern Pacific coast. Towards the south the decrease is to 6 F. on the Gulf coast ; towards the southeast, to 3 in southern Florida ; while towards the east there is but a slight, if any, de- crease to the North Atlantic coast. The increase of the average temperature variability with latitude is scarcely perceptible on the Pacific coast and in the extreme western United States in general; it is about i for each 300 miles at the center of the continent; and on the Atlantic coast the increase is quite rapid on the northern and southern coasts, but very gradual on the middle coast. Relative Frequency of Cold Waves. The cold waves that sweep over the country in the rear of cyclonic disturb- ances are an important feature in the climate of the United States, and the number of rapid changes in temperature depends on their frequency. The amount of these sudden falls of temperature is quite a different matter from the aver- age variability of the temperature just mentioned, and bears somewhat the same relation to it as the absolute range of temperature bears to the average range of temperature. It is not the coldest region, however, that has the greatest number of rapid changes in the temperature. It is rather where the cold of the center of the continent struggles to render abnormally cold the air of the region in the neighborhood of the ocean or other large bodies of water, and where the cyclonic areas occur most fre- quently. Thus there arises in that region the alternat- ing control of the continental cold and the oceanic warmth. The accompanying chart (Fig. 102) shows roughly the rela- CLIMATE OF THE UNITED STATES. 333 tive frequency, in the central and eastern United States, of falls of temperature of at least 20 F. in 24 hours. Forty such temperature changes occur in the region of northern Michigan for every five in the region of the southeastern United States. Dates of Earliest and Latest Frosts. The times of earliest frost in the fall and latest frost in the spring are of great practical importance. Average Date of Earliest Killing Frost. This phe- nomenon is dependent both on the average temperature, FIG. 102. RELATIVE FREQUENCY OF FALLS OF TEMPEKATURE OF OVER 20 IN 24 HOURS (AFTER RUSSELL). consequently on the varying altitude of the sun, and on the diurnal range of temperature, and consequently on the cloudiness and distance from oceans. Whenever the aver- age temperature in its autumnal decrease reaches the place where the diurnal range will take the night minimum tem- perature below freezing, then the first frost must occur ; but this regular occurrence is usually anticipated by the ap- 334 ELEMENTARY METEOROLOGY. pearance of one of the coid waves just mentioned, in which the temperature fall is far greater than would occur during the regular diurnal change. Frost occurs earliest (Fig. 103) at the Canadian boundary ; but the time is retarded, at first gradually, and then more rapidly, with southern progress to the Gulf coast. There is also a retardation from Montana towards the northwest coast of the United States. Frost occurs in the northern part of the United States about Sept. i ; but the time is gradually retarded to Oct. I at FIG. 103. AVEKAGE DATE OF EARLIEST HARD FROST IN THE UNITED STATES (AFTER GREELY). about the latitude of the Ohio River valley, and to Nov. i in central Mississippi ; to Dec. i in southern Louisiana, and nearly to Jan. I in central Florida (when it occurs at all in the last-named region). Average Date of Latest Killing Frost, When with the increase of the average temperatures in the spring (with the altitude of the sun) these become so high that the diurnal change is such as to keep the night minimum CLIMATE OF THE UNITED STATES. 335 temperature above the freezing point, then ordinarily the frosts will cease ; but some temperature falls greater than that due to the diurnal change will very likely occur at a later date, and thus cause a considerable retarda- tion in the time of latest frost. The latest frost occurs (Fig. 104) earliest along the coast of the Gulf of Mexico; but the time is retarded, at first rapidly, but afterwards more slowly, to the Canadian boundary. It occurs about Feb. i on the Gulf coast, March I in southern Mississippi, FIG. 104. AVERAGE DATE OF LATEST HARD FROST IN THE UNITED STATES (AFTER GREELY). April i in northern Tennessee, May I in southern Wis- consin, and the latter part of May in North Dakota. Rainfall The rainfall conditions of the United States as a whole are very variable, the total annual amount of precipitation varying from 3 or 4 inches, or less, to over 100 inches. The accompanying chart (Fig. 105) shows the average (336) CLIMATE OF THE UNITED STATES. 337 annual precipitation for the different regions of the whole country. The Regions of Most Rainfall are: the northwest coast, where in the extreme corner over 100 inches fall, owing tr the warm moist air from the Pacific Ocean blowing on to the cooler continent ; the northeastern shore of the Gulf of Mexico, where over 60 inches fall, and where a plentiful supply of moisture is found in the southerly winds from the Gulf; the southeast coast 'of Florida* and the extreme eastern portion of the North Carolina coast, where over 70 inches fall, and where a plentiful amount of mois, ture is furnished by the easterly winds blowing from the Gulf Stream in the Atlantic Ocean, The Region of Least Rainfall is in the southwestern part of Arizona, where the winter winds blow from the dry interior of the continent, and the summer winds blow from a cooler to a hot region, and thus little rain falls. It is seen from the map that there is an increase of rainfall with northward progress on the Pacific coast, and a decrease from south to north on the Atlantic coast. In the eastern inland part of the United States there is a decrease from south towards the north ; in the central part, in the neighborhood of the looth meridian, there is no change in a north-and-south direction, but rapid changes in an east-and-west direction ; while in the extreme west- ern inland region there is an. increase with northward prog- ress. The region separating the western from the eastern type of variation with latitude is almost at the center of the continent. The wind from the west and southwest blows the moist Pacific Ocean air over the land ; and where the land is cooler, as on the northern Pacific coast, an excessive rain- fall occurs, but where the land is warmer, as on the south- 338 ELEMENTARY METEOROLOGY ern coast, but slight precipitation occurs, Moreover, the sea wind at the south is frequently but a day wind, and is not as permanent as in the north. The air flowing into the northwestern United States is deprived of most of its moisture before it has gone far inland, so that the regions to the east have dry air and little precipitation until moist air is supplied by winds from the Hudson Bay, the Great Lakes, the Atlantic Ocean, or the Gulf of Mexico. And the northwest and northern winds carry this dry air far to the south, .almost to the Gulf of Mexico, making the whole region traversed one of deficient rainfall. But from the region where the southwest winds begin to carry the moist air from the Gulf of Mexico up into the Mississippi Valley, the rainfall in- creases again rapidly over the whole eastern United States The excessive precipitation of the southeastern United States is due to the prevalence of warm oceanic winds. The whole United States is divided into three great distinct regions as regards rainfall. The first region, to the east of the 97th meridian (which passes through central Kansas), has an average rainfall of over 35 inches ; and while there is a decrease from nearly 60 to 35 inches with northward progress, yet, on the whole, the distribution of the amount of rainfall is remarkably equable for such an extended region. The second region, that to the west of the 97th meridian (central Kansas), has at first, for the next 5 of longitude, a rapid decrease from 35 to 15 inches of rainfall at the western border of Kansas; but from this (iO2d meridian) westward there is little change in the rainfall north of latitude 42 (the southern boundary of Wyoming) to the region of the northwest coast, where the increase becomes very rapid towards the coast. rom about the looth meridian westward, and south of latitude 40, there is a slow, somewhat variable decrease from 20 inches to a minimum rainfall of from 2 to 5 inches in the desert region to the east of the Sierra Nevada Mountains, from which there is a slight increase to the Pacific coast. The rapid change in rainfall in CLIMATE OF THE UNITED STATES 339 the region just east of the Rocky Mountains is due to the fact that just there the moisture begins to be received from the Gulf of Mexico on the south, and Hudson Bay and the Great Lakes on the north. The third region is on the northern half of the Pacific slope, to the west of the Cascade Mountains. Here the rainfall increases rapidly (from 25 inches) from northern California to the northern boundary of Washington, and from the Cascade Mountains on the east to the ocean on the west. FIG. 106. PERCENTAGE OF TOTAL RAINFALL IN THE RAINY SEASON. MONTHS OF RAINY SEASON. (U.S. WEATHER BUREAU.) Periods of Rainfall. The accompanying chart (Fig. 106) shows the months of greatest rainfall, and the percentage of the total annual rainfall which falls during those months. The dotted lines are drawn merely to roughly separate the various regions, and to show that the months and lines of percentages inclosed by each belong together. The regions between these must form transi- tion zones from one system to another. In most of the United States the period of greatest rain is that usual for continental localities, except on the Pacific coast, where there is a winter rainy season. 340 ELEMENTARY METEOROLOGY. In the winter time, the warm, moisture-laden winds of the Pacific are blown on to the colder west coast of the United States, and precipitation takes place. In the sum- mer time, however, this moisture is rendered still warmer, and thus farther removed from the temperature of con- densation. Much of the summer rainfall of the eastern United States is of local character, and accompanies thunder- storms. Rainfall Types. The classification of the rainfall for the United States by regions is about as follows, when the two types of a single and a double period for the annual amount of rainfall are made the basis for classification : 1. The Pacific type of rainfall is characterized by heavy precipita- tion during midwinter, when the west and southwest winds blow from the ocean ; and an almost total absence of rain in the late summer, when the winds blow from the ocean to the warmer land, or when the north winds blow from the dry continent. This type is found in British Colum- bia, Washington, Idaho, Oregon, California, Nevada, and western Utah. 2. The Mexican type of rainfall is characterized by heavy precipi- tation after the summer solstice, when the southeast winds blow in the moist Gulf air ; and a very dry period after the vernal equinox. This type is found in Texas, New Mexico, and most of Mexico. In Arizona the type is complex, resulting from the meeting of the Pacific and Mexican types. 3. The Missouri type of rainfall is characterized by light winter rains, because the prevailing north and northwest winds bring little moisture ; and heavy rains due to the south and southeast winds blow- ing the moisture from the Gulf in the late spring and early summer, when most of the annual rainfall occurs. This type occurs over the slopes of the Arkansas, Missouri, and Upper Mississippi rivers, and Lakes Superior and Michigan, and includes the States of Montana, the Dakotas, Minnesota, Nebraska, Kansas, Iowa, Missouri, Wisconsin, Illinois, the Territory of Oklahoma, with parts of Arkansas, Texas, Michigan, Indiana, and Indian Territory. This covers the greatest area of any ; and the time of occurrence of the rainfall is most favorable to agriculture, considering the total amount of annual rainfall. CLIMATE OF THE UNITED STATES. 341 The eastern part of Texas has a complex rainfall, resulting from the meeting of the Mexican and Missouri types. 4. The Tennessee type of rainfall has the heaviest rains in the latter part of the winter or the early spring, when moist south winds from the Gulf meet the cold north winds from the interior ; and the least rain in mid-autumn, when the cool, dry wind blows from the north and northeast. This occurs in Tennessee, Arkansas, Mississippi, eastern Kentucky, western Georgia, Alabama, and Louisiana. 5. The Atlantic type of rainfall has a somewhat uniform distribu- tion of rain throughout the year, and it covers the Atlantic coast States except New England. 6. The St. Lawrence type of rainfall has heavy rains in the late summer and autumn months, due to the meeting of warm, moist south, and cool north winds ; and a scarcity of rain in the spring, due to the dry winds from the west and north. It covers the St. Lawrence valley. In N.ew England the Atlantic and St. Lawrence types meet, and cause late summer and late fall maxima, and irregular minima in early summer and spring or early fall. Agriculture Rains. The rainfall during the late spring and early summer months in the United States is very favorable for agriculture east of the Rocky Mountains, and for very limited regions to the west. A well-distributed rainfall of over two inches per month is found east of about the looth meridian, in April; east of the Ii5th meridian (central Idaho) in the northern United States, and east of the iO2d meridian (western Texas) in the southern United States, in May; and east of the H5th meridian at the north (central Montana), the iO2d merid- ian at the center (eastern Colorado), and the looth meridian at the south (central Texas), in June. It is to these widely distributed sufficient rainfalls that the agri- cultural prosperity of the United States is largely due. In most of the centra! and eastern United States the rainfall is between two and four inches per month. Where 342 ELEMENTARY METEOROLOGY. it is over six inches per month in limited regions, mostly near the Gulf coast, the excess is due to a few heavy rains. Excessive Rains are those which are injurious to plant growth, and which cause damaging floods. When the rainfall exceeds two inches in a day, or ten inches in a month, it becomes excessive. Such rains are most fre- quent on the North Pacific coast from January to March, on the South Atlantic coast in summer, on the Gulf of Mexico in the spring, and in the Tennessee and Kentucky region from the late winter to early summer. Occasionally these excessive rainfalls occur north of the Potomac River, chiefly in the late summer or early fall. An excessive monthly rainfall of nea-rly 42 inches has occurred in northern California, one of 37 inches in Lou- isiana, one of over 28 inches in North Carolina, and one of over 22 inches in New Jersey. The greatest daily rainfalls (in 24 hours) do not usually exceed 10 inches, with occasionally double that amount during two successive days. The chart, Fig. 107, shows the greatest daily rainfall in various regions of the United States. On June 15 and 16, 1886, and within a period of 24 hours, nearly 21.5 inches of ram fell at Alexandria, La. On Feb. 11-13, 1886, there occurred a remarkable storm in New England, in which precipi- tation was excessive over a large extent of territory. The amount was 12.4 inches at Canton, Conn. ; and it averaged more than 7 inches over 1,500 square miles, and more than 5 inches over 5,000 square miles. Rain frequently falls at the much more rapid rate of from 5 to 18 inches per hour during short intervals of time. Such rainfalls usually last from 2 or 3 to 15 minutes, and rarely for so long as half an hour. In Galveston, Tex., in June, 1871, there was a rainfall of nearly 4 inches in 14 minutes ; but usually the amount during these intense CLIMATE OF THE UNITED STATES. 343 rainfalls is about 2 inches. We do not know much about the amount of rainfall in the cloud-bursts, except from the overflowing of creeks and the damage which is done. FIG. 107. HEAVIEST DAILY RAINFALLS (U.S. WEATHER BUREAU). Snowfall in the United States. Snow falls with greater or less frequency in almost all parts of the United States. Since snowfall depends on the temperature, snow occurs in very low latitudes if the altitude is sufficiently great ; but it is almost unknown on the Florida peninsula and on the California coast below the 35th parallel Taking the country as a whole, snow seldom remains unmelted on the ground south of about latitude 33 at low altitudes ; and on the eastern coast this limiting latitude is about 31, while on the western coast it is about 37. The region of greatest snowfall would naturally be that in which the greatest amount of precipitation occurs in winter, but it is also necessary that the temperature be low enough to convert the moisture into snow. The total average annual amount of snowfall in inches, for the (344) CLIMATE OF THE UNITED STATES, 345 7-year period 1884-91, is shown on the accompanying chart (Fig. 108). The lines connect the regions of equal snowfall, and the attached figures indicate the amounts of snowfall in inches. The snowfalls rep- resented on this chart refer to the low lands and plains in which the Larger towns are situated. Where there are mountains, the present figures might have to be increased even several fold in order to repre- sent the snowfall at higher altitudes than their bases. In general, for the low lands, the region of greatest snowfall is from northern Michigan (where it is 130 inches) eastward along the Canadian frontier. To the southward 'of this region there is a rapid decrease to 20 inches at about the latitude of St. Louis, and then the decrease is very slow until the practical disappearance of snow near the Gulf coast. On the Atlantic coast the snowfall is 60 inches at the north, and diminishes, at first rapidly and then more gradually, to the limit of practical disappearance on the Carolina coast. To the west of the region of maximum snowfall in northern Michigan, there is a dimi- nution to nearly 30 inches in the Missouri valley, and then an increase to 50 inches on the high plateaus from Utah to Montana, with a decrease farther west towards the Pacific Ocean. In the whole southwest United States the snowfall is slight, except on the mountains. On the Pacific coast the snowfall is less than 10 inches at the north, and decreases to less than i inch at San Fran- cisco ; while south of San Francisco snow seldom falls. Along the Mississippi River there is a decrease from 40 inches at the extreme north, to 20 inches at St. Louis, and 2 inches at Vicksburg. The greatest recorded annual amounts of snowfall are in eastern California and western Nevada, where they reach several hundred inches. At 346 ELEMENTARY METEOROLOGY. Summit, Cal., on the Central Pacific Railroad, the snowfall averaged 378 inches per year during a period of ten years. For the greater portion of the United States it may be said that in the northern half the snowfall exceeds 20 inches, and in the southern half is less than 20 inches. Frequency of Precipitation. The distribution of the number of days with precipitation (rainy and snowy days) follows in part the amounts of precipitation, and in part FIG. 109. AVERAGE NUMBER OF DAYS DURING THE YEAR WITH PRECIPITATION (U.S. WEATHER BUREAU). the distribution through the year. For the whole year, along about the central meridian of the United States, there is precipitation on about 100 days (Fig. 109). To the eastward there is an increase towards the Atlantic Ocean, with a local maximum over the lower Great Lakes. To the westward there is an increase towards the Pacific Ocean in the northern part, and a decrease in the southern part. In the western United States, where the seasons of rainfall are strongly marked, the number of rainy days is CLIMATE OF THE UNITED STATES* 347 greatest in the region of greatest rainfall, and least in the region of least rainfall. In the eastern United States, however, the region of greatest number of days is at the north, where the rainfall is more equably distributed over the whole year, and where the passage of barometric minima is most frequent. The Greatest Number of Consecutive Days with Pre- cipitation is shown on the chart, Fig. no. In general, FIG. no. GREATEST NUMBER OF CONSECUTIVE DAYS WITH PRECIPITATION (U.S. WEATHER BUREAU). the regions of greatest rainfall have the greatest number of consecutive days with precipitation. For the main part of the United States, the number of days is from 10 to 20 ; but on the northwestern coast, where there is such a pronounced rainy season, with steady winds from the ocean, there is an increase to nearly 40 days ; while in the northeastern United States in the St. Lawrence valley, where the cyclonic areas are most frequent, the number of days increases to over 30. 348 ELEMENTARY METEOROLOGY. The Greatest Number of Consecutive Days without Pre- cipitation is of importance as signifying the duration of droughts, and is shown on the chart, Fig. in. The more equable the distribution of precipitation through the year, the less the number of consecutive days without it. In nearly the whole of the eastern half of the United States, the greatest period of uninterrupted drought is from 15 to 30 days, except in the extreme southeastern part, where it varies from 30 to about 50 days. There is an increase in FIG. in. GREATEST NUMBER OF DAYS WITHOUT PRECIPITATION (DROUGHT) (U.S. WEATHER BUREAU). the number of days from the center of the United States towards the northwest, to from 30 to 60 days, and towards the west and southwest to over 150 days. Absolute Humidity. The absolute amount of water contained in the air is shown on the charts, Figs. 112, 113, which give the number of grains contained in each cubic foot of surface air for the various regions of the United States. Since the possible amount of water in- CLIMATE OF THE UNITED STATES. 349 creases with the increase of the temperature, the distribu- tion follows somewhat after that of the temperature, and is least in winter and greatest in summer ; but the dis- tance and accessibility of the supply of moisture, and the direction of the wind, also enter into this distribution. In the Winter (Fig. 112) there is a region of least mois- ture in the region of lowest temperature in the north central United States ; and there is an increase towards the east, south, and west, in the direction of increase of FIG. 112. AVERAGE ABSOLUTE HUMIDITY IN MIDWINTER, JANUARY (AFTER GRKELY). temperature, and roughly in proportion to this increase of temperature. In the Summer (Fig. 113) the humidities are much in- creased over those of winter, and the increase is propor- tionally the greater where there is the greater annual range of temperature, although the increase in grains is nearly the same in all parts. The region of least moisture is now shifted to the high plateau west of the Rocky Mountains, where, with the region to the west of it, the 350 ELEMENTARY METEOROLOGY. average temperatures are lowest. From thence there is an increase towards the higher temperatures to the east, southeast, and south. In the southwestern part of the United States the moisture would undoubtedly become FIG. 113. AVERAGE ABSOLUTE HUMIDITY IN MIDSUMMER, JULY (AFTER GREELY). greater if it were not for the mountain ranges cutting off the winds, and the unfavorable wind direction with regard to the supply of moisture. The Relative Humidity (Fig. 114), depending as it does on the absolute moisture and the height of the tempera- ture above that necessary for condensation, is least in the southwestern United States, where the amount of moisture is low and the temperature is high. There is an increase in all directions from this region to the limits of the United States. This increase is fairly symmetrical, although not with the same rapidity on all sides. The relative humid- ity, while lowest in summer and highest in winter, is so irregularly distributed throughout the year, that midwinter and midsummer charts are not given. 352 ELEMENTARY METEOROLOGY. The average relative humidity for the year (Fig. 1 14) is distributed over the United States about as follows : On the Atlantic and Pacific coasts there is a relative humidity of abo.ut So% on the northern coast, which decreases to a little below 75 % on the southern coast ; on the coasts of the Great Lakes and the Gulf of Mexico it is about 75 % ; in the low eastern half of the inland United States it is about 70%, except in the upper Ohio River valley and western Virginia, where it decreases to 65 % ; in the elevated western half of the inland United States it is below 60%, and decreases gradually to 40% in southwest Arizona. Degree of Cloudiness. The observations of the degree of cloudiness are made on a scale of from o for clear weather to 10 for an entirely overcast sky, but the aver- age results are as usual expressed in percentage of com- plete cloudiness. Thus 50 % of cloud signifies that the sky is half covered with cloud. The degree of cloudiness, which depends on the relative humidity, varies with the season of the year somewhat irregularly, but is in general greatest in winter and least in summer ; but the maximum occurs in the spring months in the Rocky Mountain region, and over the Great Plains at the center of the continent. Cloudiness in January (Fig. 115). In the extreme north- west of the United States there is a cloudiness of over 70 %, which decreases towards the south to 40% on the south Pacific coast, towards the southeast to between 30 % and 40 % for nearly the whole of the southwestern part of the United States, and towards the east nearly to 40 % on the northern Great Plains. At nearly the longitudinal center of the continent there is a cloudiness of about 50%, and to the eastward of this there is a quite rapid increase in the CLIMATE OF THE UNITED STATES. 353 northern part to over 70% on the Lower Lake region, with a decrease to less than 60 Sfo on the North Atlantic coast ; while in the southeastern part of the United States there is at first an increase to about 60 %, and then a slight decrease again as the coast is approached. In Florida the percentage falls below 50%, except on the eastern coast. The central line of the region of least cloudiness passes from south- ern Arizona in a northeasterly direction near the northwest corner of New Mexico, the western part of Colorado, the southwest corner of FIG. 115. AVERAGE CLOUDINESS IN MIDWINTER, JANUARY (AFTER GKEELY). (Scale: o = clear; 100 = completely cloudy.) South Dakota, and the northeast corner of North Dakota. To the east and west of this line there is a very rapid increase in the cloudiness in the northern part, and very much smaller increase in the southern part, of the country. Cloudiness in August (Fig. 116). --The cloudiness in August is least in the Sierra Nevada region, where it is less than 10%, and is greatest on the southern Pacific coast, the Atlantic coast, and in a small region in north- WALDO METEOR. 21 354 ELEMENTARY METEOROLOGY. western New Mexico, where it varies from 45 % to 50 % of total cloudiness. In the northern half of the United States there is an uninterrupted increase in the cloudiness from 20% in the western part (Idaho) to about 45 % in the region of the Great Lakes, whence it varies little to the Atlantic coast. The increase is slow west of the Rocky Mountains, but east of them is much more rapid. In the southern half of the United States there is a cloudiness of 45 % on the FIG. 116. AVERAGE CLOUDINESS IN MIDSUMMER, AUGUST (AFTER GREELY). (Scale: o = clear; 100 = completely cloudy.) Pacific coast, and a rapid decrease to 10 % in the near inland (central California), a local increase to 45 % or 50% in New Mexico and Colorado, then a decrease to 35% at the looth meridian, and thence eastward a steady in- crease to 50 % on the eastern coast. At the meridional center of the continent, the cloudiness is from 35 % to 40 %, and in general decreases towards the west, and increases towards the east, by about 1 5 %. CLIMATE OF THE UNITED STATED. 355 Winds. Average Wind Velocity for the Year, in Miles per Hour (Fig. 117). -If the average amount of wind which blows during the entire year were evenly distributed through all the hours of the year, it would make a uniform velocity of about 13 miles per hour on the North Atlantic coast, 14 on the central coast, and probably decrease to 11 or 12 on the southern coast. On the shores of the Great Lakes, it FIG. 117. AVERAGE WIND VELOCITIES FOR THE VEAR (IN MILES PER HOUR). would be about 1 1 miles per hour ; and on the Gulf coast, in the western part about 12, but in the eastern part about 10, miles per hour. On the Pacific coast, a velocity of ii miles per hour at the north decreases to probably 8 miles or less per hour at the south. On the Great Plains, extending from northern Dakota to Texas, the velocities would be about 10 miles per hour, with a maximum of ii in central Kansas. To the west of about the looth 356 ELEMENTARY METEOROLOGY. meridian there is a decrease to 6 miles per hour on the Pacific slope along the region of central California, Oregon, and Washington. Towards the east from the looth meridian there is a decrease to 8 miles per hour in the northeastern part of the United States (except directly along the Lakes), and to less than 7 miles per hour for most of the region south of the Ohio and east of the Mississippi River. The velocities are thus the greatest on the coasts and treeless prairies, and least in- land and in the mountainous and wooded regions. Month of Maximum and of Minimum Wind. The month of maximum wind varies somewhat in the different regions of the United States, but usually occurs in March in the eastern half, and in April or a little later in the western half, of the country. The time of minimum is not quite so irregular, and occurs usually in August. At low altitudes in the eastern United States, away from the coast, and as far west as eastern Kansas, the maximum wind occurs in March ; but where the stations have a greater altitude, or are directly on the coast, it occurs in an earlier month, perhaps as early as Decem- ber, or even November. In the region to the west of the looth meridian (central Kansas), and extending to the Pacific slope, the month of maximum is irregular, and at moderate altitudes occurs in April or sometimes in May or June ; but at very high altitudes there is a return to the beginning of the year, which was noticed in the east. On the Pacific slope the months are not the same in different localities. On the southern coast the maxi- mum occurs in April, and the minimum in November ; on the central coast the maximum is in July, and the minimum in November.; on the northern coast the maxi- mum is in January. CLIMATE OF THE UNITED STATES. 357 Amount of Maximum and Minimum Monthly Winds. The average amount of wind for the months of greatest and least wind is about 17 miles per hour for the greatest, and 9 for the least, on the North Atlantic coast ; probably 14 or 15 for the greatest, and 9 to 12 for the least, on the South Atlantic coast ; about 11 or 12 for the greatest, and 6 for the least, on the South Pacific coast ; and 16 for the greatest, and 7 for the least, on the North Pacific coast. On the Great Lakes the average velocity is about 14 miles per hour for the greatest, and 9 for the month of least wind ; while on the Gulf coast it is 15 for the greatest and 10 for the least at the western part (Texas), and only 1 1 for the greatest and 8 for the least at the eastern part (Florida). These numbers apply to the immediate shore ; and even a single mile inland the velocities are much reduced, prob- ably as much as 33 % near the ground even if the inter- vening land is level. Away from the immediate coast the average monthly wind velocities reach about 9 miles per hour for the great- est and 6 for the least in the southeastern part of the United States ; 1 1 for the greatest and 6 for the least in the northeastern part ; but increase with westward progress to 13 miles per hour for the greatest and 9 for the least on the Great Plains from North Dakota to southern Texas, and decrease still farther to the west to 8 for the greatest and 6 for the least in the southwest United States, and 1 1 miles per hour for the greatest and 6 for the least in the northwest United States. The difference between the average amount of wind in the windiest and calmest months of the year, expressed in percentage of the average wind, is about as follows : On the North Atlantic coast region, from 50 % to 60 % ; on the central coast, 40 % ; and on the southern coast, 30%; on the Upper Lakes, 30%, and on the Lower Lakes, 40%; on 358 ELEMENTARY METEOROLOGY. the Gulf of Mexico coast, 40 % ; on the Pacific coast, 70 % at the north, 60% at the center, and probably not over 40 % at the south. Inland, east of the Rocky Mountains, there is an increase from 30 % to 50 % from the northern boundary of the United States towards the south, and towards the southeast as far as the western boundary of North Carolina, but still farther to the southeast there is a decrease to 30 %. To the west of the Great Plains there is an increase to about 60 % along the Rocky Mountain region, but it decreases again to only 20 % in central Nevada, whence it gradually increases again towards the coast. A marked peculiarity of this feature is that the same percentages found on the coast sometimes extend far inland ; and consequently these ratios expressed in percentages hold good for the region in which they occur, irrespective of a land or water surface. Maximum Wind Velocities. The maximum velocities attained in individual wind storms are greatest in regions of tornadoes and severe thunder squalls, but these winds are of short duration. Strong winds which blow for a considerable length of time, say for over five minutes, have velocities of from 50 to 70 miles per hour inland, and from 70 to 90 miles per hour on the coasts of the United States. Inland the velocities are less at the south than farther north ; but on the southern Atlantic and Gulf coasts the hurricane winds are as strong as those ex- perienced on the northern coast. On the Pacific coast the velocities increase from less than 40 miles per hour at the south to over 90 at the north. The southwest United States is the region of least maximum winds. Wind Direction. The direction of the wind is in gen- eral towards the east with the -primary air currents, but it suffers interruption from monsoon effects and the pas- sage across the country of areas of high and low air pres- sure. The average wind directions for any length of time should conform to the distribution of the air pressure dur- ing that time ; and this distribution, in extreme seasons, CLIMATE OF THE UNITED STATES, 359 over the United States is probably about as given in Figs. 1 1 8, 119; but there is considerable uncertainty about the pressures in the elevated western region. There is a strong tendency for winds of a monsoon char- acter to arise, especially in the great central region extend- ing from the Gulf of Mexico to Canada ; but the frequency and magnitude of the cyclonic areas prevent these from becoming so well marked as they would be in lower lati- FIG. 118. NORMAL AIR PRESSURE AT SEA LEVEL, AND NORMAL WIND DIRECTION, FOR JANUARY (U.S. WEATHER BUREAU AND HAZEN). tudes. We can consider in detail only the midwinter month of January and the midsummer month of July. The Average Wind Direction for January is shown on Fig. 1 1 8. The surface wind directions, in the greater part of the United States, are mainly the result of the anticyclonal motion around the region of high air pressure over the interior of the continent, acting on the general cur- rent towards the east. The low pressure at the northeast strongly affects the New England winds. Along the 4Oth 360 ELEMENTARY METEOROLOGY. parallel the wind direction is towards the east all the way across the continent, except on the high plateau of Utah, where it is northerly. To the north of this parallel we find in the extreme northwest in Washington a wind towards the west ; in Idaho it swerves around towards the north, in Montana it has turned towards the east, and in the Dakotas towards the southeast ; and more to the east- ward, in the Great Lake Region, there is a tendency for the winds to turn in a direction to the north of east, but in New England there is a bend again towards the south- east, owing to the area of low pressure at the northeast. Between the 4Oth and 3Oth parallels there is a wind towards the south in the Western States, towards the southwest in the Central States, and towards the southeast in the East- ern States. South of the 3Oth parallel the wind is towards the west in the eastern part, but swerves around towards the south in the Western States. The Average Wind Direction for July is shown in Fig. 1 19. The area of low pressure probably centering over Arizona and Utah controls the direction of the surface wind for the region west of the looth meridian, except at the ex- treme north, where the wind direction is towards the east ; so that in the western United States there is a wind towards the south at the north-central part, towards the west at the eastern part, towards the east at the western part, while at the southern part the wind is towards north and east. The area of low pressure in the northeastern United States (and Lower Canada) gives the easterly direc- tion to the wind in that region. The area of moderately high pressure over the southeastern United States fur- nishes winds blowing outwards towards the areas of low pressures to the west and to the north. In the south- eastern part of the United States the winds are towards CLIMATE OF THE UNITED STATES. 361 the north ; in the eastern section of this region they in- cline towards the east of north (towards the area of low pressure in the northeastern United States), and in the western part towards the west of north (towards the area of low pressure in the western United States). FIG. 119. NORMAL AIR PRESSURE AT SEA LEVEL, AND NORMAL WIND DIRECTION, FOR JULY (U.S. WEATHER BUREAU AND HAZEN). Rain-bearing Winds. The direction of winds most likely to be followed by precipitation, for the whole year, is given on the chart, Fig. 120, the limits being marked by the arrows. These wind directions depend a good deal on the direction of the water supply, and the nature of the intervening country ; and they vary somewhat from month to month. Such charts are very useful in making predic- tions of rain in weather forecasting. Destructive Storms in the United States. The greatest number of destructive storms, which are usually of a thun- der-squall or tornadic character, occurs in the months of June and July. They occur more frequently in the spring 362 ELEMENTARY METEOROLOGY. than in the fall ; and there is a retardation with the advance of the season and with increase of latitude. Thus, in the southern part of the United States, tornadoes are frequent in the spring, but in the northern part they are most fre- quent when summer is well advanced. They seldom occur west of the looth meridian, since the air farther west is FIG. 120. WIND DIRECTION MOST LIKELY TO BE FOLLOWED BY RAIN. too dry for their formation. The greatest number of destructive storms occur in Illinois, Iowa, and Pennsyl- vania, and the least number in California. The chart, Fig. 121, shows the average annual number of thunderstorms in the United States, and the months of their greatest frequency. It is seen that for the great central region, including the Mississippi River and Ohio River valleys, the month of greatest frequency is June ; but there is a retardation of about two months with approach to the eastern and southern coasts, and also with the progress into the region of deficient moisture in the western part of the United States. CLIMATE OF THE UNITED STATES. 363 The region of greatest frequency of thunderstorms is in the southeastern part of the United States near the coast, where, on the average, 40 occur during the year. There is a decrease in all directions, and the number is reduced to about 20 at about the io$th meridian on the west, at the latitude of the northern boundary of the United States on FIG. 121. MONTHS OF MAXIMUM FREQUENCY, AND AVERAGE ANNUAL NUMBER, OF THUN- DERSTORMS (U.S. WEATHER BUREAU). the north, and at the Virginia coast on the northeast. On the New England coast only 10 occur. We have now touched on the main features of the climate of the United States. A study of charts will reveal many details which it has not been possible to mention specifi- cally, and the pupil is recommended to make frequent study of these maps, and thus become familiar with the characteristics of the climate in the various sections of our country c INDEX. Absorption, thermal, 20. of solar rays by atmosphere, 27. Accidental change, 17. Adiabatic, 49. cooling of moist air, 143. expansion, formation of cloud through, 135. heating and cooling of air, 49, 183-185. Africa, climate of, 307. Agricultural rains in United States, 341. Air (see also Atmosphere), 7-14. composition of, 7, 8, 12. Air currents (see also Winds), 101. ascending and descending, 184. breaking of, into eddies, 231. effect of local high temperature on, 232. Air, density of, 75, 234. energy of, 182. expansion and contraction of, 33. heat in, 20, 180. heat of earth retained by, 29 microscopical impurities in, 9. moisture of (see Moisture), 118-142. Air motions (see also Winds), 101, 187-268. between northern and southern hemi- spheres, 205. easterly, at high latitudes, 198, 205. effect on air pressure, 201. effect of earth's rotation on, 195-199. equatorial, 192. horizontal, 101, 180. in whirls, 213. poleward, 191. retardation of, in, 199. velocities of, 205. vertical, 102, 180, 192, 200. westerly, at low latitudes, 198, 205. Air pressure (see Pressure), 73-101. Air, radiation of heat from, 183. temperature of (see Temperature), 19-70. transparency of, 166. waves, 117. weight and pressure of, 10, 13. Altitude, 14, 15. determination of differences in, 79. effect on climate, 301. Alto-cumulus clouds, 130, 131. Alto-stratus clouds, 130, 131. Anemometer, 104. Anticyclones, 214-216, 234-239. classes of, 234. dimensions of, 236, 238. excessive pressure in, 238. in United States, 314-317. in tropics, 239. irregular movement of, 282-285. moisture in, 236. paths of, in United States, 283, 314- 316. permanent, 240. pressure of air in, 236, 238. relation to cyclones, 235, 238-240. ring of, in low latitudes, 235. rotation of, 214-216. temperature in, 237. velocity of, 237. weather conditions of, 273. winds in, 237. Aqueous vapor (see Vapor). Argon, 7, 8. Asia, climate of, 308. Atlantic Ocean, winds of, 113-115. Atmosphere (see also Air), 7. vertical thickness of, 12-14. Atmospheric circulation, general, 102, 187-212. 365 366 ELEMENTARY METEOROLOGY. Atmospheric circulation, cause of, 189. diagram of, 188. secondary, 213-240. viewed as a whole, 204. without rotation of earth, 191, Atmospheric disturbances, irregular local, 241. electricity, 175-179. moisture (see Moisture), 118-142. optics, 166-175. Aurora. 175. Avalanche winds, 103, 267. Average condition, 17. Axis, inclination of earth's, 22, 23. B Bacteria, 9. Barometer, 75-77. aneroid, 77. height or reading, 76. Barometric gradients, 100, 201, 202. maxima, 88. minima, 88. pressure (see under Pressure), 73- 101. Blizzards, 264. Bora, 264, 265. Breezes, land and sea, 262. Brocken specter, 174. Bull's-eye squalls, 261. Buran, 264. Calms, equatorial, 208. Carbon dioxide, 7, 8, 12. Centrifugal force, 181, 201, 202. Centigrade, 32, 33. Central America, climate of, 310. Chinook hot winds, 266. Circulation of air (see under Atmos- pheric). Cirrus clouds, 130. Girro-cumulus clouds, 130, 131. Cirro-stratus clouds, 130, 131. Climate. 269, 293-312. continental, 296. denned, 269, 293. effect of altitude on, 301. Climate, effect of forests on, 299-301. effect of mountain ranges on, 298. effect of oceanic circulation on, 298. effect of wind on, 297, 298. land or continental, 296. oceanic, 296. of Africa, 307. of Asia, 308. of Central America, 310. of Europe, 307. of North America, 311. of South America, 310. of United States (see under United States), 313-363. polar, 303. solar, 295. telluric or physical, 295. temperate, 302. tropical, 302. water or oceanic, 296. Climatic subdivisions of the United States, 317. Climatic zones, 301. Climatological elements, 293, 294. geographical distribution in United States, 321-361 Climatology, 293. Cloud-bursts, 261. Cloudiness, annual change of, 138. degree of. 137. diurnal change of, 137. in United States, 352-354. variation of, with latitude, 138, Clouds, 129-139. amount of, 137-139. dissipation of, 134. formation of, 129, 134-137. forms or kinds of, 130-134, in a thunderstorm, 2580 luminous, 13. Cold-wave frosts, 290. Cold waves, 287. in United States, 332. Cold winds, 264. Colors of the sky, 170. Condensation, 119. causes of, 143. Conditions, meteorological, 17, Conduction, 19. Conductors of heat, 19, 2Oo INDEX. 367 Conservation of areas, 182. Contraction and expansion of air, 33. Continental or land climate, 296. Convection, 20. Corona, 171. Cumulus clouds, 130, 131. Cumulo-cirrus clouds, 130. Cumulo-nimbus clouds, 130, 131. Currents, air (see Air), 101. Cyclones, 214-234, 235, 238-240. calm at center of, 223. classes of, 216. coalescence of, 227. effect of friction on winds in, 233. formation of, 231-234. frequency of, 227-229, 316. in United States, 314-317. irregular movement of, 282-285. locating center of, 274. magnitude of, 216, 223. moisture in, 220. of temperate and colder zones, 216. of torrid zone, 216, 217. paths of, 224-229, 283, 314-316. permanent, 240. pressure of air in, 217-219. rainfall in, 221, 273. region of maximum number of, 227. relation to anticyclones, 235, 238-240. rotation of, 214-216, 233. secondary, 231. separation of, 227. shifting of winds in passage of, 275. temperature in, 219, 220. velocity of, 225. weather conditions of, 273. winds in, 222, 233. Cyclonic motion, source of, 232. Days, length of, 24, 25. Deflecting force of earth's rotation, 195-199. Density of air, 75, 234. Dew, 162. Dew-point, 123. Diathermancy, 20. Diffusion of heat, 19. Diffraction of light, 169. Direction of wind, 103. Doldrums, 208. Droughts in United States, 348. Dry stage, 120. Dust particles, 9, 169. whirlwinds, 261. Dynamical meteorology, 18. Earth, inclination of axis of, 22. revolution of, 21. rotation of (see under Rotation), 22. Eddy winds, 103, 231. Ecliptic, plane of the, 22. Elastic force of a gas, n. Electricity, atmospheric, 175-179. periodic changes in, 178. Elements, meteorological, 16. Energy of the air, 182. Equilibrium of air, 185-187. indifferent, 185, 186. stable, 185. 186. unstable, 186. Europe, climate of, 307. Evaporation, 118,163-165. annual amount of, 165. from the ground, 165. measurement of, 124, 164. periodic variations in, 165. Evaporimeter, 164. Expansion and contraction of air, 33. Expansive force of a gas, n. Eye of the storm, 230. Fahrenheit, 32, 33. Fiducial points, 32. Floods, prediction of, 290. Foehn process, hot waves due to, 288. rapid change of temperature due to 267. winds, 265. Fog image, 174. Fogs, 120, 134-140. formation of, 134-137 ground fog, 134. ocean, 140. 368 ELEMENTARY METEOROLOGY. Force, gradient, 100. of wind, 101, 105. Forests, air temperatures in, 299. effect on climate, 299-301. humidity in, 300. Fracto-cumulus clouds, 134. Fracto-nimbus clouds, 134. Friction, of air layers, in. effects of, on air motions in cyclones, 2 33- decrease of wind velocities through, 112. ' Frosts, cold-wave, 290. dates of earliest and latest, in United States, 333-335- night, 162, 291. prediction of, 290. Frozen earth, region of, 71. Fungi, 9. Gases, arrangement in air, 12. density and volume of, 10. elasticity of, n. pressure of, 10. Glory, or fog image, 174. Glow of sky, 170. Gradient, barometric, zoo, 201, 202. force, loo. Ground, evaporation from, 165. temperature of, 70, 299. H Hail, 159. fall in a thunderstorm, 259. stage, 121. Hailstorms, 259. Halo, 171-174. Halos, classes of, 172. explanation of, 173. Heat (see also Solar, and Temperature). absorbed by air, 27-29, 183. absorbed by land and water, 29, 30, 183, 296. conductors of, 19, 20. denned, 19. diffusion of, 19. in the air, 20, 182. Heat of the earth retained by atmos- phere, 29. radiation of, from air, 183. rays, absorption of, by air, 28. reflected, 20. solar, 20-31. terrestrial, 21. unit of, 27. zones of the earth, 306. Heating of air in descending currents, 184. " High," 278. Horizontal air currents, 101. Hot waves, causes of, 288. prediction of, 288. Hot winds, 264, 265. Humidity, 122-129. absolute, 122. absolute, in United States, 348. annual march of absolute, 127. annual march of relative, 126. diurnal change in absolute, 127. diurnal change in relative, 125. geographical distribution of, 126. in forests, 300. in United States, 348-352. measurement of, 123, 124. relative, 122, 126, 127. relative, in United States, 350. Hurricanes, 216, 229. predictions of, 289. Hygrometer, hair, 124. Hygrometry of the air, 118. I Inclination of earth's axis, 22. Indian ocean, monsoon winds of, 209- 212. Indifferent equilibrium, 185. Is-abnormals, 58-61. Isobar, 85. Isobaric charts, 88. of the world, 88-94. of United States, 359, 361. Isobaric surfaces, 85-88. graphical representation of, 86. inclination of, 85. Isotherms, 51, 278. course of annual, 53. INDEX. 369 Isothermal charts, 51. of the world, 52-57. of United States, 322-324. Isothermal surface, 51. Kahnisin, 265. K Land and sea breezes, 102, 262. Land or continental climate, 296. Land surface, action of heat on a, 29, 30. temperature of, 51, 296. Latitude, 15. Leste, 265. Leveche, 265. Light, diffraction of, 169. rays absorbed by air, 28. reflection of, 168. refraction of, 166. Lightning, 176-178. forms of, 177. tornadoes accompanied by, 254. Longitude, 15, 16. " Low," 278. Luminous phenomena, 166-179. M Maps, weather, 271, 278-282. Mercury, 31. Meridians, effect of convergence of, 193. Meteorology, defined, 7. dynamical, 18. statical, 18. Meteorological conditions, 17. Meteorological elements, 16. distribution of, 18. Meteorological instruments, 17. Microbes, 9. Mirage, 169. Mistral, 264. Mock suns, 172. Moist air, ascending currents of, 184. descending currents of, 184. Moisture of the air, 118-142. amount of, 122-128, 140. as cloud and fog, 129-140. Moisture of the air as precipitation, 142- 165. as vapor, 122-129. in anticyclones, 236. in cyclones, 220. in United States, 348-354. measurement of, 123, 124. on weather maps, 279. relation to temperature, 122. stages of, 120. Molds, 9. Monsoon winds, 102, 116. of Indian Ocean, 209-212. Mountain and valley breeze, 102, 262. Mountain ranges, effect on climate, 298. effect on rainfall, 146. N Neutral planes, 193, 194. Nights, length of, 24, 25. Nimbus clouds, 130, 131. Nitrogen, 7, 8, 12. North America, climate of, 311. Norther, black, 265. Northers, 264. Ocean, temperature of, 71. Oceanic climate, 296. Oceanic circulation.climatic effects of, 298. Optics, atmospheric, 166-175. Oxygen, 7, 8, 12. Pampero, 264. Periodic change, 17. Physical or telluric climate, 295. Plane of the ecliptic, 22. Polar climate, 303. Precipitation (see also under Rain, Rain- fall, Snow, and Snowfall), 142-162. average annual for United States, 335-339- frequency in United States, 346. Prediction of cold waves, 287. of frosts, 290. of hurricanes, 289. 370 ELEMENTARY METEOROLOGY. Prediction of river floods, 290. of thunderstorms, 286. of weather (see also Weather predic- tions), 271. Pressure of the air, 73-100. at various altitudes, 73-75, 77-80, 96. at various latitudes, 95, 96. annual change of, 82. areas of high, 88. areas of low, 88. classified distribution of, 99. diurnal change of, 81. effect of air motions on, 201. excessive, in anticyclones, 238. geographical distribution, 88-95. gradients (barometric), ioo, 201, 202. . graphical representation of, 86. high, 88, 99. hourly, 81. in United States, 359, 361. in anticyclones, 236, 238. in cyclones, 217-219. in tornadoes, 247. irregular oscillations of, 83. latitude of maximum, 203. long-period oscillations, 97. low, 88, 99. monthly, 83. monthly oscillations, 84. observations of, 80. oscillations of, in winter and summer, 84, 238. physiological effects of, 97. variability of, 97. Pressure of wind, 101, 105. Psychrometer, 124. R Radiation, 20. of heat from air, 183. solar (see Solar) , 21-29. Rain (see also Rainfall), 142. agricultural, in United States, 341. artificial production of, 159. causes of, 142-145. probability of, 157. stage, 120. Rainbows, 174. Rainfall (see also Rain), 142-162. amount of, 145. annual, 146. average, in United States, 335-339. diurnal change of, 145, 146. duration of, 156. effect of mountains on, 146. excessive, in United States, 342. frequency of, in United States, 346 gauge, 145. geographical distribution of, 148. greatest daily, 146. in Africa, 153, 156. in Arctic regions, 156. in Australia, 156. in cyclones, 221, 273. in Europe, 155. in North America, 155. in South America, 155. in Siberia, 155. in the Doldrums, 152. in United States, 335-348. intensity of, 156. long-period fluctuations of, 157, 158, probability of, 157. seasonal distribution of, 150. shifting of lines of equal, 158. subtropical, 153, 154. temperate, 154-156. tropical, 152. types in United States, 340. variability of, 156. variation toward interior of continents, 146. variation with altitude, 147. variation with latitude, 147. zones, 150. Reaumur, 33. Reflection of light, 168. from dust particles, 169. Refraction of light, 166. Revolution of the earth, 21. River floods, prediction of, 290. Rotary motion, in anticyclones, 214- 216. in cyclones, 214-216, 233. in tornadoes, 245. Rotation of the earth, 22. deflecting force of, 191, 195-199. influence of, on air currents, 195. INDEX. 371 St. Elmo's fire, 178. Saturation, 118, 119, 122. Sea and land breeze, 102, 262. Sea level, barometric pressures at, 77. Seasons, 40-42. Signals, storm, 291. Sirocco, 265. Sky, aspect of, on weather maps, 279. colors of, 170. glow of, 170. Snow, 159. crystals, 160. density of, 161. eaters, 267. line, 67, 69, 70. stage, 121. Snowfall, 145, 159-162. in United States, 343-346. latitudinal limit of, 159. measurement of, 161. relation to temperature, 121. Soil, effect of forests on temperature of, 299. Solar climate, 295. Solar constant, 27. Solar heat, 20, 21. absolute quantity of, 27. absorbed by atmosphere, 27-29, 183. absorbed by land and water, 29, 3- at earth's surface without atmosphere, 2 5- Solar radiation, 21. on earth's surface, 22-25, 27, 28. on northern and southern hemi- spheres, 29. Solar rays refracted by air, 167, 168. South America, climate of, 310. Spouts, 103, 259, 260. Squalls, 103. bull's-eye, 261. white, 261. Stable equilibrium, 185. Statical meteorology, 18. Stratus clouds, 130, 134. Strato-cirrus (alto-stratus) clouds, 130, 131. Strato-cumulus clouds, 130, 131. Storm (see Tornado, Thunderstorm, Squalls). Storm, eye of a, 230. signals, 291. warnings and weather charts, 276. Straight blows (derechos), 241. Sun (see also Solar). apparent motion of, 25. distance of earth from, 22, visibility of, 24. Sun dogs, 172. Telluric or physical climate, 295. Temperate climate, 302. Temperature, 19-72. abnormal average, 58-61. annual change of, 42-44. anomaly, 47. areas of high or low, 46. average, for year, 45, 46. below freezing, in United States, 329. change with altitude, 48-50, 185. diurnal change of, 35-39. effect of forests on, 299. effect on air motions, 205, 232. extremes, annual, 43, 62-66. extremes, diurnal, 36, 37. extremes in United States, 325. fluctuations of, 34, 35. fluctuations in United States, 328. geographical distribution of, 50. hourly, 39. in anticyclones, 237. in cyclones, 219, 232. in United States, 322-335. irregular changes of, 45. limit of freezing, 65. monthly, 44, 45. normal, 54. observations of, 34. of forests, 299. of ground, 70. of land and water surfaces, 51, 296, of ocean, 71. of small bodies of water, 72. of the hemispheres, 58. on weather maps, 278. range of, 45, 60-63. 372 ELEMENTARY METEOROLOGY. Temperature, relation to moisture, 122. seasonal fluctuations of, 35. variability of, 47. variability in United States, 330. zones of the earth, 304-306. Terrestrial heat, 21. Thermal absorption, 20. Thermal days, 26, 27. Thermometer, 19, 31-33. Centigrade, 32, 33. Fahrenheit, 32, 33. graduation of, 32. mercurial, 31, 33. principle of construction, 31. Reaumur, 33. scales, 31. Thunder, 176, 250. Thunderstorms, 241, 249-259. attendant phenomena, 249-251. changes in meteorological elements during, 255. classes of, 251. clouds in, 250, 258. cyclonic or progressive, 251, 253. growth of, 254. hail fall in, 259. heat or stationary, 251, 252. ideal sketch of a thunderstorm, 257. in United States, 361-363. maintenance of, 253, 254. prediction of, 286. progressive movement of, 253, 254. rain in, 250, 256, 257. region of greatest frequency in United States, 363. time of occurrence of, 255. variation of number of, with latitude, 255- velocity of, 255. wind squall in, 253. winter, 251. Tornadoes, 103, 216, 241-249, 259. air motions in, 243-245. destructive winds in, 247. formation of, 243. frequency of, 249. lightning in, 254. low pressure at center of, 247. moisture in, 245. multiple, 249. Tornadoes, paths of, 246. polar limits of, 246. prediction of, 286. pressure of air in, 247. regions of greatest frequency of, 246. rotation of, 245. safety in, 249. time of occurrence of, 245. velocity of, 247. vertical air motion in, 244, 248. wind force and velocity in, 245, 248. Trade winds, 208. Transparency of the air, 166. Transpiration, 163. Trees, temperature of, 299, 300. Tropical climate, 302. Twilight, length of, 24, 25. Typhoons, 216, 229, 230, 290. U Unstable equilibrium, 186. United States, anticyclones in, 314-317 climate of, 313-363. climate of Atlantic slope, 321. climate of central prairies, 319. climate of Great Plains, 319. climate of Pacific coast, 318. climate of plateau region, 318. climate of western Appalachian slope, 320. climatic types in, 313. cloudiness in, 352-354. cold waves in, 332. cyclones in, 314-317. destructive storms in, 361. droughts in, 348. earliest and latest frosts in, 333. humidity in, 348-352. month of maximum and minimum wind in, 356. rain-bearing winds in, 361. rainfall in, 335-348. rainfall, types in, 340. snowfall in, 343-346. temperature of, 322-335. tornadoes in, 246. weather maps, 279-282. wind direction in, 358-361. wind velocity, 355-358. INDEX. 373 Valley and mountain breeze, 102, 262. Vane, wind, 104. Vapor in air (see Humidity), 7, 118, 122. amount of, 128, 140. Vegetation, climatic effect of, 299. Vertical air movements, 102, 200. Volcanic winds, 103. W Water (see also Moisture). evaporation of, 118, 163-165. or oceanic climate, 296. surface, heat received by, 29, 30. surface, temperature of, 51, 296. temperature of small bodies of, 72. vapor (see Humidity), 7, 118, 122-129. Waterspouts, 241, 259, 260. Weather, 269, 293. and weather predictions, 269-292. Weather charts, 271, 278, 280-282. air pressure on, 278. aspect of sky on, 279. construction of, 278. introduction of, 276. moisture on, 279. temperature on, 278. use of, for weather predictions, 276. variations in, 284. wind on, 279. Weather conditions, absolute and rela- tive, 269. current, 271. in a cyclone and an anticyclone, 272. Weather maps (see Weather charts). Weather predictions, 271. accuracy of, 292. distribution of, 291. for different regions, 284. long-range, 292. methods of making, 274-285. scope of, 282. use of storm charts for, 276. Weather, seasonal, 270. Weight of air (see Pressure) , 10, 13. Whirls, movement of air in, 213. Whirlwinds, 103. dust, 261. White squalls, 201. Wind direction, 103-106, 207. annual change of, 116. daily change of, 113. in United States, 358-361. locating cyclone center by, 274. observation of, 105, 207. resultant or normal, 106, 116. Winds (see also Air motions), 101-117. avalanche, 267. classification of, 102. cold, 264. eddy, 103, 231. effects on climate, 297. foehn, 265. force of, loi, 105. hot, 264, 265. hurricanes, 216, 229. in anticyclones, 237. in cyclones, 222. in tornadoes, 245, 247. in United States, 355-361. land and sea breezes, 262. local and miscellaneous, 241-268. monsoon, 102, 116, 209-212. mountain and valley, 262. of Atlantic Ocean, 113. of Indian Ocean, 209-212. on weather charts, 279. periodic local, 262. shifting in passage of cyclone, 275. trade, 208. volcanic, 103. Wind velocity, 101, 104, 106-113. annual change of, 109, no. average annual, no. average monthly, no. change with altitude, 107, in. change with latitude, 112. diurnal change of, 106-109. in United States, 355-358. in winter, 205. measurement of, 104. retardation of, in, 199. theoretical east and west, 205. Zondas, 265. Zones, 301. THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL. BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO Sl.OO ON THE SEVENTH DAY OVERDUE. SEP 27 t! FEB Q NOV 6 1942 RETURNED TO MATH.-STAT. LIB 1959 24W60MJ j&ttlO ll]an'62KLX 1840 LD DEC 2 11961 JUL 03 1974 21 WAR 1'W" LI) LM