QC 
 
 Af .O. 247. 
 
 AIR MINISTRY. 
 
 METEOROLOGICAL OFFICE, LONDON, 
 
 UC-NRLF 
 
 SHORT COURSE IN 
 
 ELEMENTARY METEOROLOGY. 
 
 BY 
 
 W. H. PICK, B.Sc. 
 
 of tl;e 
 
 w l Committee. 
 
 LONDON: 
 PUBLISHED BY HIS MAJESTY'S STATIONERY OFFICE. 
 
 To be purchased through any Bookseller or directly from 
 H.M. STATIONERY OFFICE at the following addresses: 
 
 IMPERIAL HOUSE, KINQSWA.Y, LONDON, W.C. 2, and 28, ARINODON STREET, LONDON, S.W.I 
 37, PETER STREET, MANCHESTER; 1, ST. ANDREW'S CHKSCENT, CARDIFF: 
 
 2:, FORTH STREET. EDINBURGH; 
 r from EASOX AND SOX, LTD., 40 and 41, LOWER SACKVILLE STREET DUBLIN. 
 
 1921. 
 Price Is. fid. Net. 
 

M.O. 247. 
 
 AIR MINISTRY. : . //, : i 
 
 METEOROLOGICAL OFFICE, LONDON. 
 
 A SHORT COURSE IN 
 ELEMENTARY METEOROLOGY. 
 
 BY 
 
 W. H. PICK, B.Sc. 
 
 fee jfyt gtitfjorttj? of flj* #tet*0r0l0jjtcal 0mmitt, 
 
 LONDON: 
 PUBLISHED BY HIS MAJESTY'S STATIONERY OFFICE. 
 
 To be purchased through any Bookseller or directly from 
 H.M. STATIONERY OFFICE at the following addresses : 
 
 IMPERIAL HOUSE, KINQSWAY, LONDON, W.C. 2, and 28, ABINODON STREET, LONDON, S.W. 1 ; 
 37, PETER STREET, MANCHESTER; l, ST. ANDREW'S CRESCENT, CARDIFF; 
 
 23, FORTH STREET. EDINBURGH; 
 or from EASON AND SON, LTD., 40 and 41, LOWFR SACKVIILE STREET S DUBLIN. 
 
 1921. 
 
 Price Is. Qd. Net. 
 
\ 
 
PREFACE. 
 
 This unimposing little book is a remarkable one, for it 
 marks an epoch ; there are numerous books on elementary 
 meteorology, but this is the first which has been written by a 
 British teacher for a class of students who are learning meteo- 
 rology as a qualification for their profession. 
 
 The British Empire has produced some of the world's 
 foremost meteorologists Halley, Beaufort, Abercromby, Blan- 
 ford, Eliot, and Shaw, to mention only a few but, until recently, 
 no one but a few specialists thought it worth his while to make 
 a serious study of meteorology. But things have now changed, 
 for the war taught us the importance of meteorology in practical 
 life, and the aviator has arrived with his active life in the 
 region of nature which we meteorologists have only looked into 
 from below. 
 
 Meteorology, from being the hobby of a few and the life's 
 work of a still smaller number of despised official " weather 
 prophets," has come into its own, and not only the aviator but 
 the general public is asking for the means to study this 
 neglected branch of science. 
 
 It is, therefore, with the greatest pleasure that I welcome 
 this little book and trust that it is only the first of a long 
 series, for meteorology is still in its infancy, and as it develops 
 text-books must be rewritten or they will become worse than 
 useless. 
 
 This book has been written by a practical meteorologist for 
 practical aviators and by a teacher for his students. This is an 
 ideal combination and the product is not disappointing. I 
 wish, however, to warn the student that there are many unsolved 
 problems in meteorology, and therefore he should not be satisfied 
 with any explanation or theory which the book contains. He 
 has been given the best explanations which it is possible to 
 give in the present state of our knowledge, limited in some 
 cases by the need to use elementary considerations only. All 
 this should, however, give him. a greater interest in his work, 
 for it will afford him plenty of opportunities to use his own 
 brains, to say nothing of the possibility of his making an 
 important discovery. What more could any real student 
 desire ? 
 
 G. C. SIMPSON, 
 
 Director, Meteorological Office. 
 August 1921. 
 
 O (5)15406 Wt 8438 6216 3000 11/21 E&S A 2 
 
 519641 
 
CONTENTS. 
 
 Part I. 
 
 GENERAL METEOROLOGY. 
 
 Chap. I. Introductory. 
 
 II. Winds. 
 
 III. The Trade Winds, 
 IV. Monsoons and Similar Winds. 
 
 V. Some Considerations on Temperature. 
 
 VI. Water Vapour in the Atmosphere. 
 
 VIE. Fog and Mist. 
 VEIL Clouds. 
 
 IX. The Formation of Clouds, Bain, Snow and Hail. 
 
 X. Weather. 
 
 Part H. 
 
 SYNOPTIC METEOROLOGY. 
 
 Chap. XI. Introductory. 
 
 XII. Pressure and its Measurement. 
 
 XIII. The Making of Synoptic Charts and some Lessons derived 
 from them. 
 
 XIV. Types of Pressure Distribution. 
 
 XV. Special Phenomena. 
 
 XVI. Forecasting. 
 
 XVII. Weather Lore. 
 
 Part in. 
 
 THE UPPER AIR. 
 
 Chap. XVIII. The Variation of Wind with Altitude. 
 XIX. The Troposphere and Stratosphere. 
 XX. Pressure, Density, and Humidity in the Upper Air. 
 
 Appendix. Further Beading. 
 Index. 
 
A Short Course in Elementary Meteorology. 
 
 PART I. GENERAL METEOROLOGY. 
 
 CHAPTER L 
 
 Introductory* 
 
 1. The Atmosphere. Man lives at the base of an invisible 
 ocean of air termed the atmosphere. This atmosphere is com- 
 plex in its chemical composition, but in the lower levels near 
 the earth's surface that composition remains practically con- 
 stant. Innumerable chemical analyses of samples from these 
 lower levels taken from various regions of the globe have 
 failed to reveal any very appreciable differences. But, partly 
 by reason of different exposure to the sun and partly by reason 
 of the different properties of the land and sea surfaces in 
 contact with it, the atmosphere near the earth's surface, whilst 
 remaining sensibly constant in chemical composition, reveals 
 striking physical differences from time to time and from place 
 to place. These physical differences may be regarded at first 
 sight as comparatively slight, but their results are amazing ; 
 for example, 200 F. covers the range of atmospheric tempera- 
 tures experienced on the earth, and this is almost negligible 
 compared, say, with the difference between the temperature of 
 the electric arc and that of liquid hydrogen, but, nevertheless, 
 small as it is, it causes the great cold of the arctic regions at 
 one end of the range and the sweltering heat of the tropical, 
 regions at the other. 
 
 The study of atmospheric conditions and changes is called 
 meteorology, and meteorology is a branch of the wider science, 
 physics. The student will find it a great help if he will con- 
 tinually remember this kinship, and will keep always before 
 him the fact that all the conditions and phenomena of the 
 atmosphere are illustrations of the principles of physics, and 
 that the sea of air is a vast physical laboratory in which nature 
 is always showing some experiment or another on the grand 
 scale. It is the function of the meteorologist to note the con- 
 ditions of these experiments and to make an attempt to explain 
 the results which become evident during the course of them. 
 
 2. The Value of Meteorology. It requires no great amount 
 of reflection to realise that the weather is an important matter 
 to most men, no matter what their occupation may be. It is 
 ever present, whether in peace or war, and must be taken into 
 account in most projects. The mariner and the agriculturist 
 find it of supreme importance. The housing engineer, engaged 
 
;:.., 6 
 
 **;>' .,'.'.. 
 
 in .ereGting ; biaildmgg /in' any particular country, must have a 
 knowledge of -the" blfiaaatic 'extremes likely to be experienced in 
 that countiy. Medical men realise that there is a distinct 
 relation between the \veather and the prevalence of certain 
 diseases. The commander-in-chief conducting military opera- 
 tions is brought face to face with the fact that favourable or 
 unfavourable weather is a factor too great to be disregarded ; 
 whilst the gunnery officers have to institute a close liaison with 
 the meteorological ones if they are to secure accuracy of range. 
 And, apart from specialists such as those of the classes already 
 mentioned, even the average citizen dwelling in town or country 
 is often enough made to realise by unfortunate experience of 
 snow or fog or floods how dependent upon the weather is the 
 transport which brings to him the supplies of the necessaries of 
 life. 
 
 3. The Value of Meteorology to the Airman. The coming of 
 aviation has added yet more strongly to the importance of a 
 study of meteorology for, to the airman, a knowledge of it is of 
 peculiar importance inasmuch as the air is the medium in the 
 which he has his flying being and is the medium, the changes 
 and phenomena of which may cause him considerable perturba- 
 tion from time to time. A well-founded forecast of coming 
 weather changes and prediction is part of the work of meteoro- 
 logy is to him a necessity enabling him, first, to avoid the 
 danger which may come to him as a result of those changes ; 
 and second, to put himself in a position to be able to utilise 
 those changes to his own advantage if this be possible. But it 
 cannot be too much emphasised that these changes, important, 
 and even deadly as they may be to him, cannot be foretold by 
 even an expert meteorologist by just a casual glance at the sky 
 and a casual tapping of the " weather glass," but that their 
 prediction requires a wide-flung organisation, daily weather 
 charts, much careful observation and study, and continual 
 appeals both to precedent and to physics. 
 
 It must not be supposed, however, that prediction is the 
 only help that meteorology can give to the airman. Fore- 
 casting is a matter of the vicissitudes of the weather, but for 
 such matters as the floating of an airship or the working of a 
 petrol engine a knowledge of the normal atmospheric conditions 
 at various heights is of more importance than that of the tem- 
 porary divergencies with which forecasting is concerned. 
 Climatic data, that is data dealing with means, either monthly 
 or yearly, of the various meteorological elements is just as 
 important in its own way and for its own uses as is the forecast 
 of to-morrow's weather. Prediction is so attractive a part of 
 the science of meteorology that there is often a tendency to 
 consider it the only part. 
 
CHAPTER II. 
 
 Winds. 
 
 4. General. Wind can neither be measured like tenir 
 perature nor can it be seen and painted like clouds. For 
 these reasons man has always been ready to believe that there 
 is something vague and illusory about it, that it " bloweth 
 where it listeth," and that any attempts to study it are fore- 
 doomed to failure. Going back a little more than a century, 
 it is probably true to say that less was known about wind 
 than about any other common meteorological element. Its 
 direction could be estimated at any moment, but its force was 
 only able to be very vaguely indicated by the loose nomen- 
 clature of airs, breezes, catspaws, squalls, gusts, and the like. 
 
 5. The Beaufort Scale. In order to lend precision to per- 
 sonal estimation of wind force, Admiral Sir Francis Beaufort, 
 then Hydrographer to the Navy, devised in the year 1805 a 
 scale of numbers ranging from to 12, based upon the amount 
 of sail that a well- conditioned man-of-war could carry in winds 
 of various forces. The figure corresponded with a calm and 
 12 with a hurricane " that no canvas could withstand." Sailing 
 ships gave way to ships driven by steam, but still the tradition 
 was carried on until by the end of the century it was realised 
 that a restatement of its specifications was long overdue. The 
 revised scale is given in Table I. It will be seen that separate 
 criteria are given to the observer on coasts or inland. In the 
 inland scale, drawn up by Dr. G. C. Simpson, F.R.S., in 1905, 
 it will be seen that the observer is advised to take into account 
 such points as the movements of twigs and branches and 
 various degrees of structural damage. In the coast scale, also 
 due to Simpson, attention is directed to the behaviour and 
 rigging of smacks. Comparison of the estimates given by 
 experienced sailors with the velocities recorded by instrumental 
 measurements has enabled limits of velocity to be assigned to 
 the various numbers of the scale. 
 
 The scale may seem somewhat crude, but in practice it is 
 found to give good results. Its use is enhanced by the fact 
 that at many meteorological stations no special apparatus for 
 measuring wind is installed. For the use of observers at such 
 stations, and for the individual making a survey of weather 
 conditions away from a station, a scale such as the Beaufort 
 is a necessity. 
 
8 
 
 TABLE I. 
 
 The Beaufort Scale of Wind Force with Specifications and 
 Equivalents. 
 
 LJ 
 i 
 
 
 
 Limits of 
 
 ,Q 
 
 General 
 
 Specification of Beaufort Scale. 
 
 Velocity in 
 Miles per 
 
 5 
 
 Descrip- 
 
 
 Hour at 
 
 " 
 
 tion of 
 
 
 about 30 feet 
 
 1 
 
 Wind. 
 
 For Coast use.* 
 
 For use Inland. 
 
 above level 
 ground. 
 
 
 
 Calm 
 
 Calm - ... 
 
 Smoke rises vertically 
 
 Less than 1 
 
 1 
 
 Light air 
 
 Fishing smack just has 
 
 Wind direction shown by 
 
 1-3 
 
 
 
 steerage way. 
 
 smoke drift but not by 
 
 
 
 
 
 wind vanes. 
 
 
 2 
 
 Slight 
 
 Wind fills the sails of 
 
 Wind felt on face ; leaves 
 
 4-7 
 
 
 breeze. 
 
 smacks, which then move 
 
 rustle ; ordinary vane 
 
 
 
 
 at about 1-2 miles per 
 
 moved by wind. 
 
 
 
 
 hour. 
 
 
 
 
 3 
 
 Gentle 
 
 Smacks begin to careen 
 
 Leaves and small twigs in 
 
 8-12 
 
 
 breeze. 
 
 and travel about 3-4 
 
 constant motion ; wind 
 
 
 
 
 miles per hour. 
 
 extends light flag. 
 
 
 4 
 
 Moderate 
 
 Good working breeze ; 
 
 Raises dust and loose 
 
 13-18 
 
 
 breeze. 
 
 smacks carry all canvas 
 
 paper ; small branches 
 
 
 
 
 with good list. 
 
 are moved. 
 
 
 5 
 
 Fresh 
 
 Smacks shorten sail - 
 
 Small trees in leaf begin 
 
 19-24 
 
 
 breeze. 
 
 
 to sway. 
 
 
 6 
 
 Strong 
 
 Smacks have double reef 
 
 Large branches in motion ; 
 
 25-31 
 
 
 breeze. 
 
 in main sail. 
 
 whistling in telegraph 
 
 
 
 
 
 wires. 
 
 
 7 
 
 High wind 
 
 Smacks at sea lie to 
 
 Whole trees in motion 
 
 32-38 
 
 8 
 
 Gale 
 
 All smacks make for 
 
 Breaks twigs off trees ; 
 
 39-46 
 
 
 
 harbour. 
 
 generally impedes pro- 
 
 
 
 
 
 gress. 
 
 
 9 
 
 Strong gale 
 
 ._.,. . . v . 
 
 Slight structural damage 
 
 47-54 
 
 
 
 
 occurs ; chimney pots 
 
 
 
 
 
 removed. 
 
 
 10 
 
 "W"hnlp era IP 
 
 
 T^VPPG nriTnAi"pr1 oonttirlpT 
 
 55-63 
 
 1 V 
 
 
 
 able structural damage. 
 
 
 
 Storm 
 
 
 ^PTV T'QT'plv PTTiPTlPTlPpH * 
 
 6475 
 
 
 
 
 widespread damage. 
 
 
 12 
 
 Hurricane 
 
 
 
 Above 75 
 
 
 
 * The fishing smack in this column may be taken as representing a trawler of 
 average type and trim. 
 
 6. The Pressure Tube Anemometer. : An instrument for 
 writing a continuous record of the wind from moment to 
 moment with regard to its changes of velocity has been 
 devised by W. H. Dines, F.R.S., and is generally known as 
 the " tube anemometer." Instrumental measurements of wind 
 have obviously very definite advantages over the cruder method 
 of personal estimation. 
 
 The pressure tube anemometer consists of two parts the 
 head and the recording apparatus. 
 
 The construction of the head will be understood by reference 
 to Fig. 1. It is mounted on the top of a long pole which is 
 40 or 50 feet above the ground level and sometimes even 
 more. The upper part of the head is free to rotate, and this 
 
9 
 
 ensures that the open end of the horizontal tube A is kept 
 facing the wind by the vane V, which is attached to^the 
 opposite end of the tube A. The wind, blowing directly into 
 the open end, produces an excess of pressure there which is 
 transmitted through the central tube of the head down to the 
 recording apparatus below by means of flexible tubing attached 
 to the arm B. The fixed part of the central tube is surrounded 
 by another tube S, which is pierced by four rings of small 
 holes. The wind blowing past these holes produces a diminu- 
 tion of pressure within the space between the two tubes, and 
 this diminution of pressure is transmitted also to the recording 
 apparatus by means of a separate length of flexible tubing 
 attached to the arm C. 
 
 FIG. 1. Head of Dines Pressure Tube'Anemometer. 
 
 The recording apparatus is seen in Fig. 2. A vessel F, 
 with a peculiarly shaped inner surface E, floats inverted in a 
 closed tank T partially filled with water. The pressure tube, 
 that is the tube directly attached to the open end of the hori- 
 zontal tube A (see Fig. 1) and in which there is an excess of 
 pressure, opens above the water level inside the inverted 
 floating vessel F, and thus the increase of pressure is directly 
 directed against the base of that inverted float tending to cause 
 it to rise. This pressure tube is marked P in Fig. 2. The 
 other tube, which may be termed the suction tube and which 1 
 transmits the decrease of pressure caused by the wind blowing 
 
10 
 
 past the four rings of small holes in the outer tube of the head, 
 communicates with the space above the float. This suction 
 tube is marked S in Fig. 2. When any wind is blowing there 
 is an increase of pressure caused within the float and a decrease 
 of pressure caused above the float. The float, therefore, rises, 
 the rise being greater the greater the velocity of the wind. A 
 record of the motion of the float is thus a record of the wind 
 blowing past the head. 
 
 WATER LEVEL 
 
 FIG. 2. Recording Apparatus of Dines Pressure Tube 
 Anemometer. 
 
To face page 11. 
 
 .S 
 
 -9 
 
 'c -2 
 
 s* 
 
 e ^ 
 
 c 
 
 g ^ 
 s a 
 
 CO 
 
 O 15406 
 
11 
 
 The arrangement for tracing this record is also shown in 
 Fig. 2. A vertical rod A is attached to the top of the float. 
 The rod passes through an airtight cover in the top of the 
 tank T and carries a pen which marks on a drum D which is 
 kept rotating by clockwork. The float is kept from rotating 
 by the device of a guide B attached to it, this guide working 
 in a vertical slot C. 
 
 G and H in Fig. 2 represent stopcocks which, when opened, 
 allow the spaces above and below the float to be placed in 
 communication with the air of the room. This means that the 
 pressures on either side of the float are made the same and thus 
 the stopcocks are of great value for adjustment purposes. 
 
 The instrument as described gives only a continuous record 
 of the velocity of the wind ; but modifications of it have been 
 devised to enable it to give a continuous record of the direction 
 as well. 
 
 7. The Structure of the Wind. An example of a " trace " 
 given by such a combined direction and velocity anemometer is 
 show r n in Fig. 3. This particular record was taken at Spurn 
 Head ; but the lessons to be derived from it are quite general. t 
 Those lessons relate to the structure of wind. Anemometer! 
 traces are the best means available of showing what wind really ' 
 is like. It is never a perfectly steady " blow," but is a perpetual 
 succession of " gusts" and " lulls " with continual slight varia- 
 tions in direction. The width of the ribbon representing the 
 velocity falls off as the velocity gets less. In other words the 
 range of the gusts diminishes with diminishing velocity. It 
 would seem, therefore, that gustiness represents the eddies set 
 up by mechanical obstacles. This view is supported by the 
 fact that if the anemometer be set up over tree tops the gustiness 
 is found to be largely increased, whilst over the open sea it is 
 considerably diminished. But it is equally true that gustiness 
 is never entirely eliminated, being found even in the exposures 
 most free from mechanical obstacles and also at heights of 
 several thousands of feet. 
 
 8. Squalls. The wind, however, is not only liable to these 
 small but continually recurring variations known as gusts, but 
 also to more considerable fluctuations of velocity known as 
 squalls. nisis_jnust be distinctly distinguished from squallsT^ 
 A squall is a blast of wind, of higher velocity than has been" 
 experienced on the average, occurring suddenly and lasting 
 for some minutes at least and then dying away as suddenly 
 as it arose. A gust is also a sudden increase of wind, but it is 
 of very short duration. Squalls are probably due to definite 
 meteorological causes, though this cannot be stated with 
 certainty ; /gusts are due to the mechanical interference of 
 obstacles in the way of the wind. 
 
 9. The Exposure of a Wind Station. From what has been 
 already said, it will be realised that the exposure of a wind 
 
12 
 
 station is of the greatest importance. Just as running water 
 forms eddies and ripples around every obstacle in its path, so 
 will wind which is only running air ; and these wind eddies 
 once formed seem to possess the power of persisting for a while 
 with an independent existence of their own. Wind eddies, 
 then, must be formed at the corners of every house, over every 
 little hillock, on the faces of cliffs, around every tree or shrub, 
 or in the immediate neighbourhood of anything rising above 
 the level of the ground. A wind measurement made in close 
 proximity to any such obstacles cannot be a true measure of the 
 general flow of the air ; and for general meteorological purposes 
 it is this general flow which is required. An observer, taking 
 a wind observation, should be diligent in taking it in the most 
 open exposure possible. The vitiating power of buildings, 
 trees, and the like cannot be over-estimated. A wind vane, for 
 example, placed just above the roof of a house is thoroughly 
 bad ; this exposure is especially commented upon as it is so 
 common. 
 
 CHAPTER III. 
 
 The Trade Winds. 
 
 10. The Trade Winds. Winds may either form part of great 
 atmospheric movements, of a scheme of planetary circulation, or 
 they may be local in character. Of the former class the trade 
 winds are the outstanding example ; of the latter those accom- 
 panying cyclones and anti-cyclones (which winds will be treated 
 later), or land and sea breezes are illustrations. 
 
 Just as water will flow from places of higher level to those 
 of lower, and as electricity will flow from points of higher 
 potential to those of lower, so air will flow from places of higher 
 barometric pressure to places of lower. 
 
 This consideration is of great importance when the general 
 arrangement of barometric pressure over the earth is taken into 
 account. The general arrangement is as follows. Two belts of 
 air at high pressure encircle the earth, the belts being enclosed 
 approximately between the lines of latitude 30 N. and 40 N., 
 and between the lines 30 S. and 40 S., whilst a belt of lower 
 pressure lies along the equator, and great areas of low pressure 
 surround both of the geographical poles. Remembering, then, 
 that a wind's direction is that from which it blows, and also 
 that air flows from places of higher pressure to places of lower, 
 it would appear that, if no other factor operated, there should be 
 northerly winds blowing from 30 N. latitude towards the equator, 
 and southerly winds blowing from 30 S. latitude also towards 
 the equator. Similarly, southerly winds should blow from 40 N. 
 latitude towards the North Pole, and northerly winds from 
 
13 
 
 40 S. latitude towards the South Pole. These expected winds 
 are shown in Fig. 4, which also shows the general distribution 
 of pressure over the earth. The figure is shown in plan ; the 
 winds would actually, in. the conditions assumed, blow along 
 meridians. $ ^ #- * 
 
 ' $&& 
 
 / 
 
 LOW PKtSSUKEL 
 
 t t T 
 
 HIGH FRE.SSUR.e. 
 
 Mil 
 
 LOW PRESSURE. 
 
 M r t 
 
 EQUATOR 
 
 V 
 
 HtCrH 
 
 v V y 
 
 J-OW 
 
 30$, 
 
 FIG. 4. General Scheme of Pressure over the Earth with 
 Winds as they would be if the Earth did not rotate. 
 
 But a most important other factor does operate. The earth 
 rotates about its axis once in every 24 hours approximately. 
 The speed of rotation of any point on the equator is about 
 1,000 miles per hour, and this speed falls off with increasing 
 latitude until at either pole it is zero. 
 
 Foucault, the great French physicist, showed that a large 
 pendulum, hung suspended in such a way that it could move in 
 any plane, once set swinging apparently changed the plane of 
 its swing slowly but regularly. If the pendulum were started 
 swinging from north to south, it was found after a certain time 
 to be swinging from north-east to south-west, and after a further 
 time from east to west, and so on around in a clockwise direction . 
 The student is referred to any good text-book of physics for the 
 actual details of this fundamental experiment. The great point 
 to be realised is that any pendulum so hung freely suspended 
 does not really change its direction of swing in space ; it remains 
 in a state of uniform motion in a uniform direction, and the 
 apparent twisting of its plane of swing is due to the rotation of 
 the earth. 
 
 An analogy may be drawn from a passenger on board a ship. 
 If the ship be at rest the passenger has no difficulty in walking 
 
14 
 
 parallel to the planking of the deck. If, however, the ship be 
 suddenly swung rapidly round to the right the passengers finds 
 it an exceedingly difficult matter to maintain his walk parallel 
 to that planking, but tends to deviate to the left. If the ship 
 be swung rapidly towards the left his tendency is to deviate to 
 the right. What really happens is that the passenger tends to 
 maintain his direction of motion unchanged with regard to 
 objects outside the ship, but that the rotation of the ship gives 
 him an apparent motion to the left or right as the case may be. 
 Every moving body on the earth, whether it be a bullet fired 
 from a rifle, a steamship travelling over^ the seas, or a particle of 
 air moving to form wind, is subject to the same truth, for the 
 earth is continually rotating. The moving body keeps on in its 
 original direction ; the rotation of the earth causes it to have an 
 apparent deviation. 
 
 The general law has been stated thus. Every body moving 
 on the surface of the earth is deflected to the right in the 
 northern hemisphere, and to the left in the southern hemisphere 
 because of the earth's rotation. 
 
 Let particles of air be imagined just above the surface in 
 30 N. latitude. They are impelled by the difference of 
 pressure to move equatorward to form a northerly wind, but the 
 law j ust enunciated shows that they will be deflected towards the 
 right as they are in the northern hemisphere, and hence the 
 northerly wind will become a north-easterly one, the north-east 
 " trade." 
 
 A similar application of the law of deviation will reveal how 
 the southerly wind blowing from 30 S. latitude becomes 
 converted into a south-easterly one, forming the south-east 
 " trade." Care must be taken to apply that part of the law 
 which has reference to the southern hemisphere. 
 
 It will be noticed, too, that the winds blowing from the 
 south from 40 N. latitude towards the North Pole, and the 
 northerly ones blowing from 40 S. latitude towards the South 
 Pole become westerly or south-westerly in each case by similar 
 applications of the law. 
 
 The scheme of the winds as they actually blow is shown in 
 Fig. 5. This figure is complementary to Fig. 4 and should be 
 studied in conjunction with it. 
 
 It is interesting to notice that the name " trade " was not 
 given to these winds because of their benefit to commerce, 
 great as this was in the days of sailing ships, but because they 
 are " track " winds, winds keeping to one track. The north- 
 east and south-east trades blow at sea with remarkable steadiness. 
 This uniformity must not, however, be exaggerated. Sometimes 
 they weaken or shift and their direction is apt to be subject to 
 considerable variation when they encounter islands. This 
 latter variation is most marked about the larger island groups 
 of the Pacific Ocean, especially the Fiji and Samoa ones. 
 Occasionally, too, they are invaded by revolving storms which 
 
15 
 
 control the winds temporarily ; and near coasts they are 
 interrupted by the phenomena of land and sea breezes. 
 
 LOW PRESSURE: 
 jrr 
 
 /VfeSTERUE.3 
 
 / /Nfc. 
 
 O-OW PR.EL55URE. 
 
 U 
 
 \.v\ 
 
 EQUATOP 
 
 HIGH PRESSURE. 
 
 '30*5. 
 
 FIG. 5. Scheme of Winds as they actually blow taking 
 into account the Rotation of the Earth. 
 
 The great streams of air constituting the trade winds 
 average nearly 2 miles in depth. The north-east trade over the 
 Atlantic has a mean velocity for the whole year at the surface 
 of about 10 ' 5 miles per hour, varying from a mean of about 
 7 '5 miles per hour in October to a mean of about 13 '5 miles 
 per hour in April. The south-east trade over the Atlantic has 
 a higher mean velocity of about 14 miles per hour for the 
 whole year, varying from a mean of about 13 miles per hour 
 in January to a mean of about 15 miles per hour in April, June, 
 and August. 
 
 11. The Doldrums. The Doldrums lie between the steady 
 trades and in the immediate neighbourhood of the equator. 
 The Doldrum belt is a region of light and variable winds with 
 many calms and is characterised by cloudy skies, sultry 
 weather, much rain and exceptionally violent thunderstorms. 
 In the days of sailing ships this region was much dreaded by 
 mariners who were liable to lie becalmed there for considerable 
 periods with dire results in regard to their temper and health. 
 The reason for the name is sufficiently clear. 
 
 12. The " Roaring Forties.' 1 This name is given by sailors 
 to the region between 40 S. and 50 S. latitude. Here there 
 is a more uninterrupted surface of ocean than is to be found in 
 
16 
 
 any other region, and the westerly winds blow from the outer 
 margin of the south-east trade belt with great regularity and. 
 with greater strength than do the trade winds. The constancy 
 of these winds, the " brave west winds," is so great that even 
 now sailing ships are enabled to compete by their aid with 
 steam ships in trade with New Zealand via the Cape of Good 
 Hope, the return journey being made around Cape Horn. 
 
 13. The Westerly Winds of the Northern Hemisphere. The 
 westerly winds blowing from the outer margin of the north-east 
 trade belt are by no means so steady as are the " brave west 
 winds " of the southern hemisphere. They are very frequently 
 interrupted by storms, and, indeed, the northern temperate 
 zone on land is characterised by such a succession of shifting 
 winds that the prevailing direction of wind movement from the 
 south-west or west is not very apparent until careful records 
 are made and examined. 
 
 CHAPTER IV. 
 
 Monsoons and Similar Winds. 
 
 14. The Unequal Heating of Land and Water. The sun, 
 whilst above the horizon, is pouring down a continuous stream 
 of radiant energy, or, as it is alternatively termed, " insolation." 
 It cannot be too clearly realised, however, that whilst on its 
 way from the sun to the earth, this insolation is not heat. It 
 was excited by the heat of the sun and it produces heat when 
 its energy is absorbed by any material particles on which it 
 may- fall ; it must be regarded as only a particular form of 
 energy until it is so absorbed, a form of energy having a 
 definite velocity of propagation which is equal to that of light. 
 Even at the cost of repetition it is again stated that only when 
 this radiant energy is absorbed by material particles does it 
 manifest itself as heat, and consequently as a rise in tempera- 
 ture. But the nature of the material on which the radiant 
 energy falls is of great importance in determining how much 
 absorption shall take place and consequently how much rise 
 of temperature shall ensue. Every different substance has to 
 absorb a different amount of heat to produce the same rise in 
 temperature. 
 
 The different behaviour of land and water surfaces subjected 
 to the same amount of insolation from the sun has far-reaching 
 effects in nature. Only the surface of the land is heated and 
 it is practically the same surface all the time. In the case of 
 water this is transparent to some of the radiant energy, which 
 therefore passes through to be absorbed by the lower layers. 
 The heat in the case of water is thus more widely distributed 
 
17 
 
 than in the case of land and consequently the rise of tem- 
 perature in the water surface is proportionately diminished. 
 But there is a more important factor than this : the specific 
 heat of land, that is the heat required to raise a unit mass 
 of it through a unit rise in temperature, is much less than 
 that of water. There is the further point, too, that some of 
 the insolation absorbed by the water is used in bringing about 
 evaporation and insolation so used is not available to bring 
 about any rise in temperature. 
 
 A consideration of the three factors mentioned will reveal 
 the very important result that the surface of land is warmed to 
 a considerably higher temperature than the surface of water 
 when the two surfaces are exposed to equal amounts of 
 insolation. 
 
 But when the sun, the source of the radiant energy, goes 
 below the horizon, the earth begins to pay back to space the 
 energy it has received. If there is income of energy there is 
 also expenditure of energy ; absorption is countered by radia- 
 tion ; and it is one of the most important laws in the science 
 of heat that a good absorber is also a good radiator. The 
 result of this is that as a land surface is a better absorber than 
 a water surface, so also it is a better radiator than a water 
 surface. The total effect of this is that a land surface gets 
 heated quicker and rises to a higher temperature than a sea 
 surface during the hours that the sun is above the horizon, but 
 loses 1 its heat quicker and falls to a lower temperature than does 
 the sea surface during the hours of night. In the last sentence 
 it must be understood that the land and sea surfaces are 
 subjected to the same amount of insolation. 
 
 Clouds prevent the full effects of both absorption and 
 radiation of radiant energy so far as the surface of the earth 
 (including both land and water) is concerned. During the day 
 time, if the sky be cloudy, the insolation from the sun has to 
 pass through the clouds before it can reach the earth. It does 
 not accomplish this passage without loss, for the clouds are 
 composed of material particles, and it is a property of such 
 particles that they absorb some of the radiant energy falling 
 on them. Some of the energy, too, is reflected from the top 
 surfaces of the clouds. Consequently, on a cloudy day, less 
 radiant energy actually reaches the surface of the earth than on 
 a day of clear sky ; hence there is less heating of both the land 
 and water of the earth. At night, too, clouds radiate heat and 
 also reflect some of the radiation emitted by the earth's surface 
 back to that surface, so that on a cloudy night the land and 
 water surfaces do not cool so much as on a clear sky night. 
 For the maximum effects of both absorption and radiation to be 
 made manifest a clear sky is necessary. Fog or mist acts in a 
 similar way to cloud ; this would be expected, us fog or mist 
 is only cloud at the surface of the earth. 
 
 O 15406 B 
 
18 
 
 15. Land and Sea Breezes. The foregoing considerations 
 are of great importance in explaining land and sea breezes. 
 As the day progresses the land gets rapidly warmer than the 
 sea. Consequently, in its turn, the air over the land becomes 
 heated more than does the air over the sea. The warmer air 
 then sets up upward convection currents on account of its 
 lessened density and a cooler breeze flows in from the sea to 
 take the place of the air which has risen. During the night 
 the reverse process takes place. The sea, through losing its 
 heat by radiation at a much less rapid rate than does the land, 
 becomes warmer than the land. The upward convection currents 
 now occur over the sea and a breeze off the land sets in. 
 
 It is clear that the land and sea breezes introduce a levelling 
 effect upon the temperatures experienced in coastal regions. 
 They prevent great extremes as, compared with temperatures 
 inland in the same latitude, the land temperature is made lower 
 by day and warmer by night. 
 
 In such a region as the British Isles the phenomena of land 
 and sea breezes are often masked by cyclonic disturbances (see 
 Part II.), but in tropical regions they are much more important. 
 In those regions the diurnal temperature variation is very 
 marked, and cyclonic storms are by no means so frequent. In 
 some tropical places where the configuration of the land favours 
 their development, the land and sea breezes attain to great 
 strength. This is particularly the case at Port Royal, in 
 Jamaica. 
 
 Normally, in nearly all latitudes, the land breeze sets in at 
 about 8 o'clock in the evening and reaches its maximum, 
 strength at about 3 o'clock in the morning. Conversely, the 
 sea breeze sets in at about 9 or 10 o'clock in the morning, 
 attaining its maximum strength at about 3 o'clock in the 
 afternoon. 
 
 Some doubt exists as to the depth to which the land and 
 sea breezes extend landwards or seawards as the case may be. 
 Along the coasts of the United Kingdom the sea breezes 
 probably extend about 10 miles inland, whilst the land breezes 
 reach a slightly less distance seawards. In tropical regions 
 these extensions are greater. 
 
 16. Monsoons. Well marked differences of temperature 
 between land and sea surfaces may arise either as a result of 
 rapid diurnal heating and cooling, which is the cause of the land 
 and sea breezes, or as a result of rapid seasonal heating and 
 cooling. Over the centre of large continents far removed from 
 the sea the seasonal variation of temperature is greatest ; 
 summer being characterised by a very high mean temperature 
 and winter by a very low one comparatively. As a result, the 
 pressure over the land in summer is much reduced, becoming 
 less than that over the surrounding less warmed seas, whilst in 
 winter the reverse is the case. Consequently winds blow 
 
19 
 
 outwards from the land towards the seas in winter, and inwards 
 from the seas to the land in summer. 
 
 These winds changing with the seasons of summer and 
 winter are the monsoons. 
 
 The monsoons of India are the best known winds of their 
 class. The hot weather in India usually commences about 
 February and continues on until June. The heating process 
 goes on for a considerable period before the actual reversal of 
 wind direction takes place. When the reversal does occur, 
 it comes suddenly and is generally accompanied by stormy 
 weather. The wind then begins its " blow ''from the south- 
 west, mainly directed towards the high lands of the Himalayas 
 with an appreciable turn to the right in accordance with the 
 law of deviation due to the rotation of the earth, which has 
 already been enunciated. The south-west monsoon blows up to 
 about September or October and the whole season is one of rain. 
 The winds of the summer monsoon attain to great strength. 
 
 In winter, the summer monsoon from the south-west is 
 replaced by one from the north-east flowing out from the main 
 land masses towards the sea. This monsoon, flowing from such 
 a huge area of land, is a dry wind, rain on]y occurring lightly 
 and irregularly. The wind force in this monsoon is consider- 
 ably less than that in the south-west one. 
 
 For most continents the monsoon effect is comparatively 
 insignificant. Asia, especially in India and China, shows it 
 most distinctly, and there, in fact, it is the predominating 
 feature of the climate. Australia, too, shows it in some degree, 
 but the reversals are hardly abrupt enough to justify the use of 
 the special term in its usual connotation. The equatorial regions 
 of Africa show it to some slight extent ; but it hardly appears at 
 all on the American continent. So far as Europe is concerned 
 there seems to be a slight monsoonal effect observable on the 
 Iberian Peninsula and that is all. 
 
 17. Katabatic and Anabatic Winds. Claiming some sort 
 of kinship with land and sea breezes and with monsoons are 
 what are termed katabatic and anabatic winds. 
 
 At night there is greater radiation from high ground than 
 from valleys, greater radiation because at higher altitudes there 
 are less dust particles or water vapour particles in the air to 
 intercept the radiation and thus to absorb or reflect it. The 
 colder land means colder air in contact with it. Colder air is 
 denser air, which, therefore, by gravitational action slides down 
 the slopes. These winds caused by the gravitation of cold air 
 off high ground are katabatic winds. Such winds are local but 
 are very frequently observed, especially at night. 
 
 An example on the large scale is found in the Northern 
 Adriatic in the cold and very violent wind which sometimes 
 blows down the slopes of the plateau to the north and is termed 
 the " Bora." The plateau becomes very cold in the clear 
 
 B 2 
 
20 
 
 weather which often occurs in winter, much colder than the 
 valleys below it, and, consequently, a stormy descending 
 current of air is started which persists, when conditions are 
 favourable, for considerable periods. 
 
 The converse of katabatic winds is seen in the daytime. 
 Then, the land in valleys becomes warmed which warms the 
 air in contact with it, and this causes convection currents to rise 
 up the slopes of the surrounding high lands. Such up currents 
 constitute anabatic winds. They do not approach the violence 
 which characterises some examples of katabatic winds. 
 
 Note. It should be noticed that all the winds treated up to 
 the present are quite distinct from those non-periodic winds 
 due to travelling cyclones and anti -cyclones. A consideration of 
 the winds accompanying those systems is deferred until those 
 systems themselves are treated. 
 
 18. The Lull of Wind at Night. An extended investigation 
 of wind records kept at any land station will reveal the note- 
 worthy fact that, generally, the wind drops in force after sunset 
 and remains lulled until after sunrise. This must not be taken 
 .as true for every night, for travelling cyclonic systems quite 
 often cause this, the normal lulling, to be entirely masked. 
 Eliminating these travelling systems, however, the statement 
 remains true. The nightly lull is not so marked far oat at sea 
 nor is it at even the modest height of the Eiffel Tower in Paris. 
 
 The cause of the 1 all ing appears to be found in the 
 following considerations : 
 
 In day time, especially in the summer, the surface layers of 
 air get considerably warmed through contact with the rapidly 
 absorbing and hence rapidly warming surface of the land. The 
 warmed surface air then rises to mingle with the currents a 
 little higher up. The upper air currents have more energy 
 because they are not subjected to the friction of the earth's 
 surface. It will be recognised, then, that the rising of some of 
 the less energetic surface air will tend to slow down the 
 currents just above the surface layers while the coming down 
 of some of the more energetic air of these upper layers (coming- 
 down because colder and thus denser) to take the place of the 
 risen air will tend to increase the velocity of the currents in 
 the surface layers. 
 
 At night, no air rises and hence the velocity of the currents 
 just above the surface layers tends to increase upon what it was 
 during the day, as it is no longer subjected to any " brake " 
 effect, whilst no more energetic upper air comes down to 
 increase the velocity of the surface currents. 
 
 Under normal conditions, then, the wind at the surface 
 lulls at night whilst the wind just above the surface layers, say, 
 up to 1,000 feet, tends to freshen. During the day, the reverse 
 happens ; the wind in the surface layers tends to increase in 
 velocity and the wind up to 1,000 feet to decrease. 
 
21 
 
 As would be expected from what has gone before, the 
 surface lulling of wind is more marked during a night of clear 
 sky than during one which is cloudy or overcast, for it is on 
 such clear nights that the radiation from the earth is greatest 
 and hence that cooling is greatest ; and the greater the cooling 
 of the earth's surface the greater will be the cooling, and thus 
 the density, of the air in contact with that surface ; and such a 
 greater density means a greater lulling of the wind. 
 
 CHAPTER V. 
 
 Some Considerations on Temperature. 
 
 19. Heating due to Compression. The absorption of insolation 
 is not the only means by which temperature may be raised. 
 There is the heating due to mechanical compression. Tyncfall 
 showed this effect by his device of a glass syringe in which a 
 piece of tinder or some other combustible substance was placed. 
 By pushing in smartly the piston of the syringe, the combustible 
 substance could be set alight. The same phenomenon is within 
 the experience of any one who has pumped up the tyres of a 
 bicycle. The pump soon becomes warm. In. each case the 
 compression of air has produced heat. In each case, too, the 
 conditions are such that practically no heat can be transferred to, 
 or taken from, the air compressed, either by conduction or 
 radiation. Such conditions are said to be "adiabatic" and the 
 heating produced is purely dynamical. 
 
 20. Cooling due to Expansion. The effect considered above 
 is entirely reciprocal. If compression causes heating, expansion 
 causes cooling. The expanding gases have to do work in pushing 
 away the surrounding air. To do this work they must derive 
 energy from themselves if the conditions are assumed to be 
 approximately adiabatic. This energy must come from the 
 store contained in the vibration of the molecules of which the 
 gases are composed. Consequently, the molecules vibrate a 
 little less quickly and this, following the kinetic theory of 
 gases, means a fall in temperature. 
 
 21. Adiabatic Changes in the Atmosphere. As has already 
 been stated, the essential matter in. an adiabatic change is that 
 no heat must either get to, or be taken from, the substance under 
 investigation. Such a condition of things is realised in the 
 atmosphere in the interior either of ascending or descending 
 masses of air. Ascent or descent means entry into regions either 
 of less or greater pressure with consequent expansion or com- 
 pression and resulting dynamical cooling or heating. Such dyna- 
 mical changes are of the greatest importance in meteorology as 
 
22 
 
 they are vital, as will be seen later, in explaining the formation 
 of clouds and rain. 
 
 22. The Vertical Tem.perature Gradient. In ordinary still 
 air, that is, air without vertical components of motion, the 
 temperature falls off generally at the rate of approximately 1 F. 
 for every 300 feet of ascent. 
 
 The rate of fall of temperature of ascending air due to dyna- 
 mical cooling is also fairly accurately known. It is about 1 F. 
 for every 185 feet of ascent. This is termed the adiabatic gradient. 
 
 Considering air rising, then, the ascending air falls in 
 temperature at the rate of 1 ' 6 F, in 300 feet, whilst the 
 surrounding air only falls at the rate of 1 F. in the same 
 range. Thus the rising air is falling in temperature at the rate 
 of ' 6 F. for every 300 feet compared with the surrounding air. 
 But the rising air probably started to rise owing to the fact 
 that, in the first place, it was slightly higher in temperature, 
 through some cause or another, than the air which surrounded 
 it near the ground. As it is falling in temperature compared 
 with its surroundings during the ascent at the rate of 0'6 F. 
 for every 300 feet it will be seen that it will soon arrive at such 
 a height that its temperature will become equal to the tempera- 
 ture of its surroundings. When this point is reached, no 
 further ascent will occur. 
 
 23. Temperature Inversions. It has been said that tem- 
 perature decreases in still air with ascent above ground level. 
 Though this is generally the case, it is not invariably so. 
 Sometimes the air increases in temperature as one ascends. 
 Such a condition is known as an inversion of temperature. 
 Such inversions are common during clear sky nights in winter 
 when radiation from the earth is able to act longest and to 
 exert its maximum effect in lowering the temperature of the 
 surface. The great cooling of the surface chills the air in 
 contact with it to such an extent that its temperature falls 
 considerably below that prevailing in the atmosphere 200 or 
 300 feet up. 
 
 Nor do inversions invariably occur in the neighbourhood of 
 the surface. Quite often they are encountered in isolated layers 
 in the upper air well above the surface. Inversions occurring 
 thus generally mark the margins of cloud layers. 
 
 24. Distribution of Temperature over the Globe. Little need 
 be said under this heading as any good book on Physical Geo- 
 graphy will give the details required. The maps used to 
 demonstrate the distribution are constructed by the use of 
 isotherms, that is, lines drawn through places having the same 
 temperature. Isotherms on such maps refer to temperatures at 
 sea level, and hence in order to find from them the actual 
 temperature at a given place, the height of the place must be 
 
23 
 
 noted and an allowance of 1 F. decrease per 300 feet of eleva- 
 tion must be made. It should be recognised, too, that the 
 temperatures plotted are mean temperatures over, generally 
 speaking, a considerable number of years. 
 
 The isothermal charts for July and January are especially 
 worthy of study. In the July chart the maxima in the centres 
 of the great land masses like North West India and the Sahara 
 should be noticed. Those maxima would be expected to occur 
 where they do because the sun is nearly overhead there and 
 thus is pouring out a great stream of insolation which the land 
 surfaces greedily absorb with a resulting great rise in tempera- 
 ture. The comparatively low temperatures of the oceans in the 
 same latitudes as the ]and masses in question are also note- 
 worthy as illustrating the different degrees of heating of land 
 and water when subjected to the same amount of insolation. 
 The bending of the isotherms as they strike the coasts is also 
 interesting in the same connection. 
 
 In the January chart the maxima occur in the land masses 
 of the southern hemisphere as would be expected as a result of 
 the migration of the sun across the equator towards them. The 
 minima in the land masses of the northern hemisphere, for 
 example, in Asiatic Siberia, should also be noticed. These 
 northern districts receive little insolation during their winter 
 and, hence, the good radiating power of land comes fully into 
 play with a consequent great fall in temperature. 
 
 25. Annual Range of Temperature. Facts concerning the 
 annual range of temperature in various places may also be learned 
 from a consideration of isothermal charts of the globe. The 
 area of moderate annual range, that is, of less than 10 F. 
 extends over nearly all the torrid zone because there the 
 annual variation of insolation is small. It also extends over 
 large parts of the oceans of the southern hemisphere even up 
 to quite high latitudes because the waters are only able to change 
 their temperature with difficulty. 
 
 The areas of the most extreme ranges, over 70 F., are only 
 found in the large land masses remote from the equator. Thus 
 they are limited to the northern hemisphere as no great land 
 masses are so situated in the southern. The greatest annual 
 range occurs in the centre of Asiatic Siberia, where it approxi- 
 mates to 120 F. This is all in accord with what would be 
 expected from what has been previously said with regard to 
 absorption and radiation. 
 
24 
 CHAPTER VI. 
 
 Water Vapour in the Atmosphere. 
 
 26. General.* Of all the constituents of the lower atmos- 
 phere water vapour is the most variable. At any time the 
 total mass of it is very small compared with the mass of dry 
 air, but the results accruing from that small amount are, from 
 the meteorological point of view, amazingly great. 
 
 27. Evaporation. The Molecular Theory of Matter states 
 that matter is discrete, and not continuous, being made up 
 of minute particles termed molecules. Further, the molcules 
 are assumed to be in a state of incessant motion. 
 
 The results of the incessant motion vary, however, according 
 to whether the matter is in the gaseous, liquid, or solid state. 
 The molcules in the gaseous state are much more sparsely 
 scattered than in either the liquid or solid state ; collisions 
 are much less frequent, and, consequently, the particles are 
 much more mobile. Their impact on the walls of the contain- 
 ing vessel produces pressure on those walls. En the liquid 
 state, the molecules, whilst by no means so mobile as in the 
 gaseous state owing to the fact that they are closer together 
 and that considerable forces are exerted between them, are 
 yet sufficiently mobile to allow the liquid quickly to take up 
 the shape of the vessel into which it is put. Whilst most 
 of the molecules in the mass are not free to move far from 
 their mean positions owing to the attractions of the surrounding 
 molecules, yet some will attain to such a momentum that they 
 will, when near the surface, break away from their neighbours 
 and escape into space, thus reducing the volume of the original 
 liquid. In course of time, continued repetitions of such escapes 
 lead to the complete evaporation of the liquid. In thus 
 breaking away from their neighbours, the escaping molecules 
 do work which can only be performed at the expense of the 
 heat energy in the remaining liquid. In consequence of this 
 drawing out of heat, the remaining liquid is cooled and it is 
 possible to find experimentally how much heat energy is 
 necessary to evaporate 1 gram of the liquid. This quantity 
 of heat is termed the latent heat of vaporisation and, in the 
 case of water, it is 600 gram-calories. It is to be noted that 
 raising the temperature of the liquid accelerates the motions 
 of the molecules and thus increases the rate of evaporation. In 
 the solid state, the molecules are much less mobile than in the 
 case of either the gaseous or liquid state. This is shown by 
 
 * Much of the data in this, and the succeeding chapter, has been 
 derived from the manuscript of a lecture on " The Water in the Atmos- 
 phere," by Dr. G. C. Simpson, F.R.S., Director of the Meteorological Office, 
 London, who very kindly lent that manuscript to the present writer. 
 
25 
 
 the low compressibility of solids as well as by the slow rate 
 at which one solid will diffuse into another. It is a mistake 
 to suppose, however, that there is no such phenomenon as 
 evaporation in the case of solids. The smell of camphor, for 
 example, is due to the direct breaking away of some of its 
 molecules, and ice is also known to evaporate directly into the 
 vapour state. The heat required to bring about this latter 
 change, however, is exactly the same as though .the ice were 
 first changed into liquid water and then the liquid water into 
 water vapour. The latent heat of vaporisation of ice is equal 
 to the sum of the latent heat of liquefaction of ice and the 
 latent heat of vaporisation of water. 
 
 But whether a solid or a liquid be evaporating, the number of 
 molecules which leave each square unit of its surface per unit of 
 time depends only upon the temperature, the number increasing 
 with rise of temperature. 
 
 28. Vapour Pressure. Let it be supposed that water is 
 evaporating in a closed, vacuous space. The molecules that 
 escape from the surface of the water accumulate in the space 
 above the water as water vapour. The concentration of this 
 water vapour goes on increasing but the molecules of it are also 
 in a state of incessant motion, and some of them will plunge 
 back into the liquid water from which they originally escaped. 
 The number of the molecules which thus return to the liquid 
 from the vapour increases as the concentration of the latter 
 increases. After a certain interval of time, a point will be 
 reached at which the number of molecules which return to the 
 liquid in any given time is exactly equal to the number of 
 molecules which leave the liquid in the same given time. The 
 system is then in equilibrum and the vapour is said to be 
 saturated. It is to be recognised, however, that the equilibrium 
 is kinetic and not static. There is no state of rest ; molecules 
 continue to leave the liquid and molecules continue to return to 
 the liquid ; but as the number of each is equal for any given 
 time, no visible result, one way or the other, can be produced. 
 
 Though in what has been said the space above the liquid 
 water has been taken as being originally vacuous, it could be 
 filled with air or any other gas and evaporation would continue 
 just as before, and at the point of equilibrium, provided the 
 temperature had not been altered, the same number of molecules 
 would still leave the surface per unit of time. 
 
 When equilibrium is attained for any particular temperature, 
 that is, when the vapour is in a saturated condition, there is 
 a definite quantity of water vapour in each unit volume of the 
 space above the liquid water, and this quantity is the same 
 whether there be air in the space or not. This water vapour 
 will exercise a definite pressure upon the walls of the vessel 
 containing it by reason of the impacts of its molecules. 
 
26 
 
 The same holds true for any other temperature ; there will 
 be at the point of equilibrium a definite quantity of water 
 vapour, different for each temperature, in each unit volume of 
 the space above the water, and this water vapour will exert a 
 definite pressure entirely irrespective of the pressure of the 
 surrounding air. 
 
 For each temperature, then, the actual amount of water 
 vapour in the air above a flat water surface when that air is 
 saturated with water vapour may be expressed either as a mass 
 per unit volume or, and this is more convenient in practice, 
 by the pressure it exerts. Tables have been prepared and are 
 generally available giving these data. The lower the tempera- 
 ture the smaller the mass of water vapour per unit of volume 
 and the lower the pressure exerted by that vapour. 
 
 29. The Dew Point. From what has been already said, it 
 will be seen that if air, only partially saturated with water 
 vapour, be cooled, a temperature will be reached sooner or later 
 at which the water vapour actually present will be sufficient to 
 saturate that air. Any cooling beyond that point, called the 
 " dew point," will cause some of the water vapour to condense 
 in the form of liquid water. 
 
 30. Absolute and Relative Humidity. At any moment, the 
 air contains a certain quantity of water per unit volume, a 
 quantity which varies from place to place. It may, or may not, 
 be saturated. The actual amount of water vapour in the air is 
 termed the absolute humidity of that air. It is most correctly 
 expressed as a mass per unit volume, but, in practice, is more 
 conveniently expressed by the pressure it exerts. 
 
 The important meteorological matter, however, is the relative 
 humidity which may be defined as the ratio of the actual 
 amount of water vapour present in the air to the amount which 
 the same volume of air would hold if it were saturated. It is 
 usually expressed as a percentage. At the dew point, the 
 percentage is 100. 
 
 Relative humidity is a very variable meteorological element 
 inasmuch as it depends not only upon the amount of water 
 vapour present, but also upon the temperature of the air, which 
 temperature is continually varying from place to place and from 
 time to time. 
 
 31. The Measurement of Relative Humidity. The instru- 
 ments usually employed at meteorological stations for 
 determining the relative humidity are the dry- and wet-bulb 
 thermometers. These consist of two exactly similar ther- 
 mometers placed in a Stevenson screen, which consists of a 
 large wooden box with louvred sides. The screen protects the 
 thermometers within from direct sunshine and the louvred sides 
 permit of sufficient ventilation to ensure that the air within is a 
 fair sample of the air without. Further, to obtain the true air 
 
conditions, it is erected 4 feet above the ground in the most 
 open, exposed position possible. The dry bulb shows the 
 temperature of the air. The wet bulb is exactly similar to the 
 dry-bulb thermometer except that its bulb is covered by a 
 piece of muslin which is kept wet by means of conducting 
 threads leading into a small vessel containing water. 
 
 If the outside air is not saturated, evaporation will take 
 place from the muslin. Energy is needed to break down the 
 inter-attractions of the particles during the evaporation. This 
 energy is obtained at the expense of the heat of the adjacent 
 surface, that is, the thermometer bulb which records a fall in 
 temperature accordingly. 
 
 If the outside air is saturated already with water vapour, no 
 evaporation takes place from the muslin ; consequently, there 
 is no fall of temperature recorded by the wet bulb and the two 
 thermometers read the same. If the surrounding air is com- 
 paratively dry, that is, if it has a low relative humidity, much 
 evaporation will take place from the wet bulb which will cause 
 considerable cooling of that bulb, and hence there will be a 
 large difference between the readings given by the two 
 thermometers. 
 
 Theoretical formulas have been proposed to obtain the 
 relative humidity of the air from the respective temperatures 
 indicated by the dry and wet bulb thermometers, but the 
 phenomena occurring during the experiment are too complicated 
 to lead to a simple yet adequate expression. Tables have, 
 however, been constructed by other experimental means to 
 show the relationship between the two temperatures and the 
 humidity of the air and these tables are ahvays used in 
 practice. 
 
 Various instruments for automatically recording the relative 
 humidity of the air from moment to moment have been devised. 
 The best depend for their action upon the fact that the length 
 of a human hair, which has been freed from fat by boiling in 
 caustic soda or caustic potash, varies with the relative humidity 
 of the air in which it is placed, getting longer with increase of 
 humidity and shorter w r ith decrease. A similar phenomenon is 
 observed with spiders' webs which hang much more loosely 
 when the relative humidity of the air is high than when it is 
 low, and also with the catgut strings of musical instruments. 
 Other materials contract when they absorb moisture, as, for 
 example, rope ; with which fact negligent campers in tents are 
 sometimes made painfully aware. 
 
 If the human hair a bundle is found more convenient in 
 practice be suitably supported, a pen attached to the middle 
 of it may be made to mark the relative humidity upon a 
 properly calibrated chart placed upon a drum which is kept 
 revolving by clockwork. Such self-recording hair hygrometers 
 are very useful, but require great attention, as the properties of 
 hair are subject to slow changes, so that ihe same relative 
 
28 
 
 humidity is not always shown with the same change in the- 
 length of the hair. Frequent settings against the relative 
 humidities determined by a dry-and-wet bulb thermometer are, 
 therefore, necessary. 
 
 32. Other Considerations. In Sections 27 and 28 it was 
 assumed that the surface of the water from which molecules 
 were escaping was flat. This point is important, because the 
 number of molecules which leave a surface depends upon 
 whether that surface is flat or not. More molecules leave a 
 convex surface than a flat one, and more molecules leave a flat 
 surface than a concave one. The point has an important 
 meteorological bearing inasmuch as water often appears in the 
 atmosphere as drops either as rain drops or cloud particles. 
 
 Let it be imagined that in a closed vessel water vapour is 
 in equilibrium with a flat water surface. Let it be imagined 
 further that a drop of water be introduced into the space above 
 the flat water surface. The drop is convex, and hence more 
 molecules will leave each unit area of its surface per unit of 
 time than leave each unit area of the flat water surface in the 
 same time. But in the same unit of time the drop receives per 
 unit area of its surface from the water vapour surrounding it 
 just as many molecules as does each unit area of the flat water 
 surface. That is to say, the drop in each unit of time receives 
 less molecules than leave it. In other words, the drop is 
 evaporating. The water evaporated from it will condense into 
 the flat water surface, because it increases the number of mole- 
 cules above that surface and thus destroys the state of kinetic 
 equilibrium originally existing between the ingoing and out- 
 going particles passing through that surface. 
 
 If, however, there is no flat water surface present, and if 
 condensation on to any body is not possible, the vapour pressure 
 will rise above its normal saturation value until the vapour- 
 returns to the drop the same number of molecules per unit of 
 time as it receives from it. Equilibrium- will then be esta- 
 blished between the drop and the vapour, but the pressure of 
 the latter will be above its normal saturation value. Therefore, 
 in order for drops to be in equilibrium with surrounding water 
 vapour, super-saturation of the latter is necessary. 
 
 The smaller the radius of the drop the greater must be the 
 amount of super-saturation of the surrounding vapour to pre- 
 vent the drop from evaporating. It follows from this, too, that 
 the smaller the drop the greater must be the amount of the 
 super-saturation of the surrounding vapour if that vapour is to 
 condense at all upon the drop. Calculations made by Wegener* 
 show that the drops must be extremely small if the super- 
 saturation of the surrounding vapour necessary to preserve 
 equilibrium is to be appreciable. The drops constituting 
 ordinary cloud particles have a radius of O'OOl cms., and for 
 
 * A. Wegener. Thermodynamik der Atmosphare, p. 71. 
 
29 
 
 this size the super-saturation necessary for equilibrium is 
 almost negligible, being only 0' 00012 per cent. ; but for 
 smaller drops the super-saturation if equilibrium is to be pre- 
 served is comparatively large, a factor which must weigh 
 considerably in the initial stages of rain and cloud formation. 
 
 33. Dew. About sunset the temperature of the ground, 
 and hence of the air in contact with the ground, begins to fall 
 owing to the increasing effect of radiation. The falling tem- 
 perature of the air means that its capacity for water vapour 
 decreases, that its relative humidity increases, and that soon the 
 air near the ground is brought very near to its saturation point. 
 From this time onwards the further cooling of the ground by 
 radiation lowers the temperature of the air gradually, and the 
 dew point, the point of saturation, is soon reached. But not 
 only is the air cooling, but also all bodies which can radiate to 
 the clear sky. The loss of heat from all bodies is practically 
 the same, but those bodies which are good conductors of heat 
 and which are in good thermal contact with the ground are 
 able to draw from the supplies of heat contained in the latter, 
 and thus, despite the radiation, their temperatures will not fall 
 very much. On the other hand, bodies which are bad con- 
 ductors of heat are not able to draw appreciably upon those 
 supplies of heat, and hence, not being able materially to com- 
 pensate for the loss of heat due to radiation, their temperatures 
 will fall. Rocks are a good example of the former class, and 
 grasses of the latter. 
 
 Further, compared with air, all solid bodies are good 
 radiators of heat. Consequently the temperature of grasses, 
 small shrubs, and the like, may fall below the dew point of the 
 air before the air itself reaches that dew point. Hence these 
 solid bodies will chill the air immediately in contact with them 
 to below its dew point, and water will be condensed from that 
 air upon them. This theory of the formation of dew was put 
 forward by Wells in 1818. Aitken has amplified this theory 
 by showing that much of the water deposited as dew rises from 
 the damp ground by capillary action, and that some is actually 
 exuded by the plants themselves. 
 
 Dr. G. C. Simpson,' 1 ' in the following, states the conditions 
 which affect the formation of dew. He says : " Any condition 
 which prevents the cooling of the air or of the bodies on 
 which dew usually deposits is detrimental to the formation 
 of dew. For example, a cloudy sky prevents radiation ; a 
 high wind equalises the temperature of the air and of the 
 bodies over which it passes, and by mixing up large quan- 
 tities of air prevents a local fall of temperature near the 
 ground. There are four conditions absolutely necessary to 
 
 * Dr. Gr. C. Simpson, F.R.S. Manuscript of lecture on " The Water 
 in the Atmosphere " previously referred to in section 26. 
 
30 
 
 " the formation of dew, (1) a good radiating surface, (2) a still 
 " atmosphere, (3) a clear sky, and (4) thermal insulation of the 
 " radiating surface. It is also necessary if there is to be a 
 " copious deposition of dew that the ground should be warm 
 " and moist or there should be some other source of water 
 " vapour to maintain the supply of moisture to the surface 
 " layers of air. Dew is generally formed during all clear calm 
 " nights in situations where the air is naturally moist, for, 
 11 example, in marshy country and along the banks of rivers." 
 
 34. Frost. When the dew point is lower than 32 F., the 
 water vapour condenses as ice in the form of hoar frost. In 
 the British Isles the dew point is not often likely to be below 
 32 F. in the late spring, the summer, or the early autumn. 
 In the remaining part of the year the main factors which an 
 observer, anxious to know whether frost is to be expected, has 
 to consider are the following : 
 
 (a) The likely state of the sky face between sunset and 
 
 sunrise. 
 (6) The likely velocity of the wind during the same 
 
 period, 
 (c) Local peculiarities of situation. 
 
 It is assumed that the observer is propounding his question 
 just before the time of sunset. 
 
 A clearing sky and a falling wind are helpful to the forma- 
 tion of frost, as would be expected from what has already been 
 said about dew formation, because a clear sky greatly favours 
 the cooling of the earth's surface by radiation as has been 
 previously explained, and because calms or very light winds 
 keep any particular sample of the air in contact with the 
 cooling ground for a considerable period, and hence that air 
 is given the best opportunity to become cooled itself. 
 
 The effect of situation is that air on the hill-tops becomes 
 cooled more quickly than does the air in the valleys owing to 
 the greater radiations from the hill-tops (see section 17). This 
 means that cold air trickles down the hill slopes by gravitation 
 and settles in the valleys. These pools of cold air thus formed 
 in the valleys are all in favour of frost development with the 
 result that, as a rule, frost becomes more severe as one goes 
 down from the hills into the valleys, and that hollows in the 
 hill sides are colder than the unindented slopes. 
 
 A big difference in the dry- and wet-bulb readings at sunset 
 is also a herald of frost provided that the other conditions are 
 favourable, inasmuch as that means that the air is very dry, 
 and that, consequently, it will need to be cooled considerably, 
 probably to below 32 F., to reach its saturation point. The 
 other conditions, the clearing sky and the falling wind, will 
 bring about the maximum cooling ; this maximum cooling, in 
 conjunction with the original very low relative humidity as 
 shown by the big difference noticed between the dry- and wet- 
 
31 
 
 bulb readings, ensures that when the saturation point is reached 
 and passed, it will be frost deposited and not dew. 
 
 The American practice of " smudging " to protect crops 
 from the damage caused by frost is interesting, because it is a 
 practical application of a general principle already stated. It 
 has been said (see section 14) that clouds prevent the earth's 
 surface from coolmgflso much as otherwise it would because 
 they reflect back some of the radiation to it. "Smudging" 
 consists in building a smoky fire on the windward side of the 
 crops to be protected so that a thick layer of smoke drifts over 
 them. The layer of smoke acts just as a cloud in a similar 
 position would do ; it reflects back much of the radiation to 
 the earth, and thus maintains the surface temperature at a 
 higher level than that of the non-smoke screened surfaces in 
 the neighbourhood. 
 
 CHAPTER Vll. 
 
 Fog and Mist. 
 
 35. General. In section 33 it was shown that dew is a direct 
 deposition from the water vapour in the air upon solid bodies 
 due to the fact that the temperature of the latter falls below the 
 dew point of the air, whilst the air itself has a temperature 
 above that dew point. If the temperature of the air itself falls 
 below the dew point, then an entirely different phenomenon, 
 the formation of mist or fog, occurs. 
 
 36. The Necessity of a Nucleus for Condensation. To form 
 a water drop in the atmosphere, the basis of mist or fog forma- 
 tion, there must already be present some nucleus upon which 
 the water may form. It was formerly thought that dust 
 particles formed the nuclei for condensation, but very careful 
 work by Assmann,* Wegener,f and WegandJ has shown that 
 dust in the ordinary sense of the term is not sufficient. The 
 results of their work point to the fact that the nuclei, whatever 
 they are, begin to attract water to themselves before the air 
 around them is saturated, begin to attract water, in fact, when 
 that air is relatively dry. The nuclei seem to have some 
 natural affinit} 7 " for water and are nuclei for condensation just 
 because of that property. 
 
 Substances are known which have such a natural affinity 
 for water. They are termed " hygroscopic," and examples of 
 
 * Assmann. Meteorologische Zeitschrift, 1885, p. 41. 
 t "Wegener, A. Meteorologische Zeitschrift, 1910, p. 357. 
 J "Wegand. Meteorologische Zeitschrift, 1913, p. 10. 
 
32 
 
 them are calcium chloride, sodium chloride (common salt), and 
 sulphuric acid. 
 
 Meteorological opinion now holds that condensation does 
 not begin upon dust particles in the ordinary sense but upon 
 particles of hygroscopic substances occurring in the air. The 
 nature of these particles is still open to some doubt, but it seems 
 certain that salt, derived from sea spray or from the drying 
 sand of the shore, constitutes a large proportion of them. 
 Household and factory chimneys, too, belch forth great 
 quantities of such nuclei-forming material, the chief of which 
 is sulphur dioxide which, when illuminated by sunlight, 
 becomes a highly hygroscopic substance capable of causing 
 condensation in unsaturated air. The activities of volcanoes 
 constitute another means whereby the atmosphere is supplied 
 with nuclei-forming material. 
 
 Nuclei-free air can be prepared experimentally in the 
 laboratory and C. T. 11. Wilson has shown that air in this 
 condition can be dynamically cooled until a state of fourfold 
 super-saturation is reached without condensation of the water 
 vapour in it occurring ; in other words, the air may be cooled a 
 long way below its dew point without liquid water being 
 deposited. But this hardly concerns the student dealing with 
 the free atmosphere for no sample of that has yet been found, 
 even at heights, to be free from nuclei, which is not surprising 
 when what has been said about the sources of those nuclei is 
 remembered. 
 
 37. Some Lessons from Smoke Aggregations. It is instructive 
 to consider some questions arising with regard to aggregations 
 of smoke, especially with regard to the atmospheric conditions 
 which aid in the maintenance of the smoke aggregations once 
 they are formed, for similar conditions may apply to the 
 maintenance of fogs, at least in part. 
 
 Imagining, then, that a smoke aggregation is formed in any 
 way, say by the belching forth of several neighbouring factory 
 chimneys, there are two ways in which dissipation of it may be 
 brought about. Those two ways are 
 
 (a) A wind, or 
 
 (6) Upward convection currents. 
 
 It is surprising how small a wind velocity is sufficient to 
 disperse smoke, and it would appear that for the maintenance of 
 any considerable smoke in the neighbourhood of the chimneys 
 a calm is practically necessary. 
 
 The question of upward convection currents presents 
 points of interest. Let it be assumed that a mass of air 
 is carried upwards from one layer A to a higher layer B. 
 As it rises, it gets into regions of less pressure. Conse- 
 quently it expands and the expansion causes a fall in 
 temperature. This fall in temperature has already been stated 
 (see section 22) to be equal to 1 ' 6 F. in 300 feet. It is clear, 
 
33 
 
 therefore, that if the upward temperature gradient of the 
 surrounding air be less than 1 ' 6 F. per 300 feet, the rising 
 mass of air will, when it arrives at layer B, find itself surrounded 
 by air which is warmer than itself. Consequently, the mass of air 
 will have no further tendency to rise, but, on the contrary, a 
 tendency to descend. Therefore, an atmosphere in which the 
 vertical temperature gradient is less than 1 ' 6 F. per 300 feet 
 will tend to stop upward convection currents. The smaller the 
 vertical temperature gradient is, the more effective does this 
 tendency to stop upward convection currents become. If the 
 vertical temperature gradient be actually reversed, that is, if an 
 inversion of temperature exists, the tendency reaches its 
 maximum effectiveness and no currents are likely to rise at all. 
 The atmospheric conditions, then, favouring the maintenance 
 of a smoke aggregation once it is formed are (1) the absence of 
 wind and (2) a vertical temperature gradient in the air of value 
 less than 1 ' 6 F. per 300 feet, with this latter condition 
 reaching its maximum efficiency if an actual inversion of 
 temperature exist. 
 
 38. Fogs. Fogs are due to the condensation of the water 
 vapour of the atmosphere upon nuclei existing in that atmosphere. 
 The condensation is due in most cases to cooling of the air 
 below its dew point by reason of it radiating either to the night 
 sky or to the cold ground. 
 
 One factor, then, in the formation of fogs appears to be the 
 cooling of the surface underneath the air in which the fog 
 appears. But the differences of land and sea in regard to such 
 cooling need to be borne in mind. The land surface is subject 
 to great changes of temperature due to its great powers both of 
 absorption and radiation of radiant energy. The extent of 
 these changes depends, as has already been said, upon the 
 cloudiness of the sky. Sea surfaces, on the other hand, have a 
 very much more limited range of temperature ; and, in fact, 
 the diurnal variation is sometimes almost inappreciable. 
 
 The thick " pea-soup " fogs, characteristic of London and 
 other large industrial centres, require some mention. The 
 atmosphere in those centres is full of actual particles of carbon 
 and the like emitted in the smoke from the many chimneys. 
 The water forming the basis of the fog is deposited upon the 
 small hygroscopic particles, the nuclei, already discussed. 
 When the drops are once formed they absorb the carbon and 
 dust particles and, in consequence, become dark. The fog 
 would form without the black smoke particles but it would be 
 a white fog and not a black one. Smoke abatement would 
 therefore not do away with fogs, but it would do much to 
 make the fogs occurring white and not black, and to this extent 
 is worthy of much consideration. 
 
 39. Fog at Sea. The famous fogs of the Newfoundland 
 Banks are a typical example of the formation of fog at sea. 
 
 O 15406 C 
 
34 
 
 In the region of that island, the temperature of the sea water 
 is kept low by reason of cold currents from the Arctic flowing 
 into it. In the summer time, it is surrounded on all sides 
 except the north by considerably warmer regions. Air blowing 
 across the Banks, then, from south, east, or west, is cooled by 
 contact with the cold waters. If the air blowing in is nearly 
 saturated with water vapour, the cooling produced is sufficient 
 to bring it down to its saturation point and water condenses 
 around the nuclei in the air and a fog is formed. This is 
 especially the case with southerly winds. 
 
 40. Fog on Land. Water-drop fogs are formed on land 
 under those conditions which promote great cooling of the land 
 surface with which the air is in contact. Those conditions 
 are 
 
 (a) a light wind or a calm ; 
 (6) a mainly clear night sky. 
 
 The first condition acts by keeping the mass of air long- 
 enough in close proximity to the surface of the ground to enable 
 it to become thoroughly cooled, and the second by permitting 
 radiation to space to be abundant. 
 
 41. Mist. It is a matter of some difficulty to distinguish 
 between fog and mist. The practical method of differentiation 
 is that based on how far one can see ; if the distance which can 
 be seen is under a certain limit the obscurity is termed fog ; 
 if over, mist. 
 
 Mists and fogs are similar in many respects to clouds, but 
 there is one fundamental difference which makes it very 
 undesirable to describe them as " clouds near the ground." As 
 will be seen later (Chapter IX.), clouds are nearly always caused 
 by temperature changes due to adiabatic cooling when air 
 ascends, This means that clouds are due to pressure changes 
 in the mass of air itself. Pressure changes, however, play very 
 little part in the formation of mists and fogs. The lowering of 
 temperature which causes these latter phenomena is due, not 
 to pressure changes, but. in the main, to actual radiation from 
 the ground. 
 
 42. Haze. Further difficulty is introduced by haze. Haze 
 may be due to smoke or dust due to the neighbourhood of a 
 town, or to dust raised from the ground by wind, or to 
 irregularities of density in the atmosphere and consequent 
 irregular refraction of the light such as occurs over the flame 
 of a bunsen burner. It seems well to limit the use of the 
 term " haze " to those occasions when the air is dry, and to use 
 "mist" when the air is wet. The dividing line is generally 
 taken as 80 per cent, relative humidity, but a skilled observer 
 is usually able to see if the obscurity is caused by dust particles 
 or irregular refraction, or by condensed water vapour. 
 
35 
 
 43. Visibility. Visibility, with which haze, mist, and fog 
 are 'intimately connected, is measured at meteorological stations 
 by locating a number of objects, such as church towers, buildings, 
 and the like, within definite limits of distance and determining 
 at the time of observation which object can be seen whilst the 
 one at a greater distance just beyond it cannot be seen. At 
 night, distant lights are similarly used. 
 
 The whole subject is, however, an exceedingly difficult one. 
 The state of visibility, of good or bad seeing, depends upon 
 a great number of factors. It depends not only upon the 
 meteorological conditions prevailing at the time but also on the 
 conditions ^which have gone before. It depends, also, upon 
 the geographical location of the place with regard to sources 
 of dust. The structure of the upper air, too, is important, 
 especially with regard to temperature. Much, too, depends 
 upon the direction and velocity of the surface wind. 
 
 Inversions of temperature near the ground are a cause of 
 bad seeing as they prevent upward convection currents (see 
 section 37). Attempts, too, have been made to correlate the 
 various types of pressure distribution (see Part II.) with 
 visibility, but it cannot be said that these have met with any 
 conspicuous success. 
 
 CHAPTER VIII. 
 
 Clouds. 
 
 44. Classification. Cloud phenomena are " infinite in their 
 variety." Certain types may be selected as standard, but even 
 a short period of observation reveals that the types selected do 
 not exhaust the clouds actually appearing. All manner of 
 intermediate types occur. The series of clouds is continuous. 
 Nevertheless it is necessary to have some classification. 
 
 Luke Howard in 1803 was the first to propose some such 
 classification. He distinguished three principal cloud forms 
 
 (1) Cirrus cloud (which included all types of fibrous or 
 
 feathery appearance). 
 
 (2) Cumulus cloud, having rounded tops and being generally 
 
 lumpy in appearance. 
 
 (3) Stratus cloud, applying to all clouds lying in level 
 
 sheets. 
 
 This classification suited the purposes of those comparatively 
 early times when accurate observation was in its infancy. In 
 course of time, however, attention was directed to smaller 
 degrees of difference, and the need for a more detailed classifi- 
 cation became apparent. 
 
 C 2 
 
36 
 
 45. The International Classification. Matters came to a 
 head in 1894, and at a meeting of the International Meteoro- 
 logical Committee held in that year at Upsala the publication of 
 a cloud atlas, which should at the same time be a classification, 
 was entrusted to MM. H. H. Hildebrandsson, A. Riggenbach, and 
 L. Teisserenc de Bort. The first edition of the atlas was 
 published in 1895, and a new edition, incorporating the 
 improvements suggested at the International Meteorological 
 Conference at Innsbruck in 1905, appeared in 1910. 
 
 This classification, consisting of 10 main types backed by 
 such international authority, must be regarded as the standard 
 until some similar gathering introduces alterations. 
 
 46. International Definitions and Descriptions of Cloud 
 Forms. The international definitions and descriptions of the 
 10 types of cloud forms* are now given. Those which follow 
 are those appearing in the " Observer's Handbook" issued by 
 the Meteorological Office, London, 1919 edition. In that 
 publication the translation into English has been altered in 
 certain respects from that which appears in the English version 
 of the introduction to the " International Cloud Atlas " in order 
 to represent more closely the original French. The abbrevia- 
 tions in brackets after the names of the clouds are the contractions 
 agreed upon internationally, and no others should be used. 
 
 " (1) tCurus (Ci.). Detached clouds of deli- 
 cate appearance, fibrous (threadlike) structure 
 and feather-like form, generally white in colour. 
 
 Cirrus clouds take the most varied shapes, such as isolated 
 tufts of hair, i.e., thin filaments on a blue sky, branched 
 filaments in feathery form, straight or curved filaments 
 ending in tufts (called cirrus uncinus), and others. 
 Occasionally cirrus clouds are arranged in bands, which 
 traverse part of the sky as arcs of great circles, and as an 
 effect of perspective appear to converge at a point on the 
 horizon, and at the opposite point also if they are 
 sufficiently extended. Cirro-stratus and cirro-cumulus 
 also are sometimes similarly arranged in long bands. 
 
 " (2) ^Cirro-Stratus (Ci.-St.). A thin sheet 
 of whitish-cloud ; sometimes covering the sky 
 completely and merely giving it a milky 
 appearance; it is then called cirro-nebula or 
 cirrus haze; at other times presenting more 
 
 * " Atlas International des Nuages " (The International Cloud Atlas). 
 Public conformement aux decisions du Comite, par A. Hildebrandsson et 
 L. Teisserenc de Bort, Membres de la Commission des Nuages. Seco/ide 
 edition. Gauthier-Villars. Paris, 1910. The definitions and descriptions 
 are published in the atlas in French, English, and German. 
 
 f It may be noted that the outline of the sun is visible, and his rays 
 cast a shadow in spite of the presence of clouds of these types, unless the 
 clouds and the sun are both low down on the horizon. 
 
37 
 
 or less distinctly a fibrous structure like a 
 tangled web. This sheet often produces halos around 
 the sun or moon. 
 
 " (3) *Cirro-Cumulus (Ci.-Cu.) (Mackerel Sky). 
 Small rounded masses or white flakes with- 
 out shadows, or showing very slight shadow; 
 arranged in groups and often in lines. French, 
 Moutons G erman, Schdfchen-wolken. 
 
 "(4) Alto-Stratus (A.-St.). A dense sheet of 
 a grey or bluish colour, sometimes forming a 
 compact mass of dull grey colour and fibrous 
 structure. At other times the sheet is thin like the 
 denser forms cirro-stratus, and through it the sun and 
 the moon may be seen dimly gleaming as through 
 ground glass. This form exhibits all stages of transition 
 between alto-stratus and cirro-stratus, but according to 
 the measurements its normal altitude is about one-half 
 of that of cirro-stratus. 
 
 " (5) Alto-Cumulus (A.-Cu.). Larger rounded 
 masses, white or greyish, partially shaded, 
 arranged in groups or lines, and often so 
 crowded together in the middle region that 
 the cloudlets join. The separate masses are generally 
 larger and more compact (resembling strato-cumulus) in 
 the middle region of the group, but the denseness of the 
 layer varies and sometimes is so attenuated that the 
 individual masses assume the appearance of sheets or 
 thin flakes of considerable extent with hardly any 
 shading. At the margin of the group they form, smaller 
 cloudlets resembling those of cirro-cumulus. The cloud- 
 lets often group themselves in parallel lines, arranged in 
 one or more directions. 
 
 " (6) Strato-Cumulus (St.-Cu.). Large lumpy 
 masses or rolls of dull grey cloud, frequently 
 covering the whole sky, especially in winter. 
 Generally strato-cumulus presents the appearance of a 
 grey layer broken up into irregular masses and having 
 on the margin smaller masses grouped in flocks like 
 alto - cumulus. Sometimes this cloud -form has the 
 characteristic appearance of great rolls of cloud arranged 
 in parallel lines close together. (Roll-cumulus in 
 England, Wulst-cumulus in Germany.) The rolls 
 themselves are dense and dark, but in the intervening 
 spaces the cloud is much lighter and blue sky may 
 sometimes be seen through them. Strato-cumulus may 
 be distinguished from nimbus by its lumpy or rolling 
 
 * It may be noted that the outline of the sun is visible, and his rays 
 cast a shadow in spite of the presence of clouds of these types, unless the 
 clouds and the sun are both low down on the horizon. 
 
38 
 
 appearance, and by the fact that it does not generally 
 tend to bring rain. 
 
 "(7) Nimbus (Nb.). A dense layer of dark, 
 shapeless cloud with ragged edges from which 
 steady rain or snow usually falls. If there 
 are openings in the cloud an upper layer of 
 cirro-stratus or alto-stratus may almost in- 
 variably be seen through them. If a layer of 
 nimbus separates in strong wind into ragged cloud, or 
 if small detached clouds are seen drifting underneath 
 a large nimbus (the " Scud," of sailors), either may 
 be specified as fracto-nimbus (Fr.-Nb.). 
 
 "(8) Cumulus (Cu.) (Wool-pack or Cauli- 
 flower Cloud). Thick cloud of which the 
 upper surface is dome-shaped and exhibits 
 protuberances while the base is generally 
 horizontal. These clouds appear to be formed by 
 ascensional movement of air in the daytime which is 
 almost always observable. When the cloud and the 
 sun are on opposite sides of the observer the surfaces 
 facing the observer are more brilliant than the margins 
 of the protuberances. When, on the contrary, it is on 
 the same side of the observer as the sun it appears 
 dark with bright edges. When the light falls side- 
 ways, as is usually the case, cumulus clouds show deep 
 shadows. 
 
 " True cumulus has well-defined upper and lower 
 margins ; but one may sometimes see ragged clouds 
 like cumulus torn by strong wind of which the 
 detached portions are continually changing ; to this 
 form of cloud the name Practo-Cumulus may be 
 given. 
 
 "(9) Cumulo-Nimbus (Cu.-Nb.). The Thun- 
 der Cloud; Shower Cloud. Great masses of 
 cloud rising in the form of mountains or 
 towers or anvils, generally having a veil or 
 screen of fibrous texture (false cirrus) at the 
 top and at its base a cloud-mass similar to 
 nimbus. From the base local showers of rain or of 
 snow, occasionally of hail or soft hail, usually fall. 
 Sometimes the upper margins have the compact shape 
 of cumulus, or form massive heaps round which floats 
 delicate false cirrus. At other times the margins them- 
 selves are fringed with filaments similar to cirrus clouds. 
 This last form is particularly common with spring 
 showers. The front of a thunderstorm of wide extent 
 is frequently in the form of a large low arch above 
 a region of uniformly lighter sky. 
 
39 
 
 Stratus (St.). A uniform layer of 
 cloud like fog but not lying on the ground. 
 
 Plate I. The cloud layer of stratus is always very 
 low. If it is divided into ragged masses in a wind or 
 by mountain tops, it may be called Practo-Stratus. 
 The complete absence of detail of structure differentiates 
 stratus from other aggregated forms of cloud. 
 
 " The following remarks are added in the inter- 
 national atlas as instructions to observers : 
 
 " (a) In the daytime in summer all the lower clouds 
 assume, as a rule, special forms more or less resembling 
 cumulus. In such cases the observer may enter in his' 
 notes ' Stratus- or nimbus-cumuliformis.' 
 
 " (6) Sometimes a cloud will show a mammillated 
 surface and the appearance should be noted under the 
 name mammato-cumulus. 
 
 " (c) The form taken by certain clouds, particularly 
 on days of sirocco, mistral, fohn, etc., which show an 
 ovoid form with clean outlines and sometimes irisation, 
 will be indicated by the name lenticular, for example : 
 cumulus lenticularis, stratus lenticularis (Cu.- 
 lent., St.-lent.). 
 
 " (a) Notice should always be taken when the clouds 
 seem motionless or if they move with very great velocity. 
 
 " A definition of the lenticular form of cloud is 
 necessary. It may be put into the following words : 
 
 "(11) Lenticular Cloud Banks. Banks of 
 clouds of an almond or airship shape, with 
 sharp general outlines, but showing, on close 
 examination, fretted edges, formed of an 
 ordered structure of cloudlets similar to alto- 
 cumulus or cirro-cumulus, which is also seen 
 in the bank itself when the illumination is 
 favourable. Sometimes the body of the cloud bank 
 is dense, and the almond shape is complete, fore and 
 aft, but sometimes the bank thins away from the 
 forward edge to clear sky within, so that the bank 
 presents the appearance of a horse-shoe seen in per- 
 spective from below at a great distance. The bank 
 appears nearly or quite stationary, while the cloudlets 
 move rapidly into it at one side and away from it at 
 the other." 
 
 47. Other Considerations. The 10 main types of inter- 
 national classification are subdivisible into 
 
 (a) Cloud sheets. 
 (6) Heap clouds. 
 
 Cumulus and cumulo-nimbus are heap clouds and 
 the remaining types cloud sheets. The heap clouds are 
 
40 
 
 characterised by considerable vertical structure but do not 
 form horizontal layers. The cause of their formation, ascen- 
 sional movement of air, will be treated in the next chapter. 
 The height of heap clouds is very variable. It has been 
 estimated that the mean height of their bases is approximately 
 4,500 feet, but they are often found much lower, and the 
 height of their summits varies from about 6,000 feet up to 
 even 25,000 feet. 
 
 Cloud sheets, on the other hand, are characterised more by 
 horizontal extension than by considerable vertical structure. 
 Sometimes the sheets have breaks in them, sometimes they 
 have not. Quite often, indeed, the sheets are represented by 
 just a few isolated clouds in a mainly blue sky. The sheets 
 vary considerably in thickness and, often, the sky shows several 
 cloud sheets or layers existing at different levels at the same 
 time. 
 
 The height of cloud sheets is also very variable. Cirrus, 
 cirro-stratus and cirro-cumulus are generally above 25,000 feet ; 
 alto-stratus and alto-cumulus are generally between 10,000 and 
 25,000 feet ; and strato-cumulus, nimbus, and stratus are 
 generally below 7,000 feet. 
 
 Cirrus, cirro-stratus, and cirro-cumulus clouds are composed 
 of ice crj^stals. It is important to notice that halos only occur 
 in these very high clouds, inasmuch as halos are an optical 
 phenomenon due to regular refraction of the light of the sun or 
 moon as it passes through ice crystals. This point will occur 
 again later in connection with the forecasting of weather and 
 with weather lore. 
 
 48. General. There is no royal road to a knowledge of 
 cloud forms except the strenuous one of diligent observation of 
 the sky face on as many occasions as possible, together with a 
 careful comparison of what is observed with the descriptions of 
 cloud forms already given. An atlas of cloud forms, ready at 
 hand, is a very useful aid. 
 
 Such a study will be found at once delightful and useful : 
 delightful because clouds are the main component of the variety 
 and exquisite beauty of the sky face, and useful because, as will 
 be seen later, it affords most valuable information to the fore- 
 caster anxious to know what the weather is likely to be. 
 Clouds are alike beautiful and significant, and a study of them 
 is confidently recommended to all. 
 
CHAPTER IX.* 
 
 The Formation of Clouds, Rain, Hail and Snow. 
 
 49. Condensation by Dynamical Cooling. This is the chief, 
 and perhaps, the only cause | of cloud making. In all cases 
 where a vertical or obliquely ascending motion is given to air, 
 the air rises into regions of diminishing pressure. Such 
 diminishing pressure upon the rising mass of air means that 
 that air expands. As has been already stated, this expansion 
 takes place adiabatically ; that is, that considering the interior 
 of the rising mass of air, no heat is either given to that interior 
 or taken from it. Such adiabatic expansion is necessarily 
 accompanied by cooling inasmuch as expansion means that 
 work must be done in pushing back the surrounding air and 
 that the energy necessary for the accomplishment of this work 
 must come from the internal supplies held by the air in the 
 rising mass. Such subtraction of energy from the molecules 
 of the air in the mass means that the amplitude of vibration of 
 those molecules is diminished. This, in its turn, means that 
 cooling takes place in the interior of the ascending air. 
 
 The ascending air, then, merely because of its elevation, 
 cools. The rate of the cooling (see section 22) due to this 
 cause is 1'6 F. in every 300 feet of ascent. The temperature 
 gradient upward of the surrounding air is, generally, about 
 1 F. in the same number of feet. This means that the rising 
 air loses 0'6 F. for every 300 feet rise compared with the 
 surrounding air. In a longer or shorter time, then, that rising 
 air will arrive at such a height that its temperature, originally 
 greater than its surroundings, becomes equal to that of its 
 surroundings and no further ascent will then take place through 
 the ordinary processes of convection. 
 
 Before this position of comparative equilibrium is reached, 
 the ascending air may have cooled to its dew point. The 
 nuclei present, however, may be so small that condensation 
 cannot at once take place as super-saturation would be required 
 to maintain the drops formed. The air, then, continues to rise 
 
 * Once again in this chapter as in Chapters VI. and VII. the writer is 
 very greatly indebted to the manuscript of a lecture on " The Water in 
 the Atmosphere," by Dr. G. C. Simpson, F.B.S., Director of the Meteoro- 
 logical Office, London, kindly lent to him by the author. 
 
 f Wegener, in Chapter 16 of his " Thermodynamik der Atmosphare," 
 gives the reason for this view. 
 
42 
 
 until it is sufficiently super-saturated to allow of water being 
 deposited on the small drops. The moment, however, that 
 water is thus deposited, the drops, by reason of their increase 
 in size, require a less super-saturation. A small drop in super- 
 saturated air is, in fact, in a state of unstable equilibrium ; the 
 moment its size is increased at all, water is deposited upon it 
 until it reaches such dimensions that there is practically no 
 super-saturation in the surrounding air for the amount of super- 
 saturation necessary for equilibrium decreases rapidly with 
 increasing size. It has already been stated (see section 32) that 
 cloud particles, as shown by direct measurement, have a radius 
 of about O'OOl cms. and that drops of this size only require 
 ' 00012 per cent, super-saturation of the surrounding air for 
 equilibrium. This super-saturation is practically negligible 
 and it may be said therefore that the air in a cloud is not in a 
 state of super-saturation ; but the initial super-saturation often 
 necessary before condensation actually begins is noteworthy. 
 It is not likely that there is ever much more than 1 per cent, of 
 super-saturation . 
 
 As soon as condensation has actually begun, further deposi- 
 tion will continue upon the drops so long as the air current 
 continues to ascend. The drops will thus increase in size until 
 they fall as rain. 
 
 50. The Fall of Water Drops through the Air. Some con- 
 sideration of the laws governing the fall of water drops through 
 the air is necessary before the actual formation of rainfall is 
 treated. 
 
 In a vacuum, all bodies fall at the same rate with a constant 
 acceleration. When, however, bodies fall through a resisting 
 medium, such as air, they do not continually increase in velocity 
 but only increase until the air resistance is exactly equal to the 
 weight of the falling body under consideration, when the latter 
 thenceforward continues to fall at a constant speed. This 
 speed is not strictly uniform but decreases with the increasing 
 density of the air in the lower layers. 
 
 The rate at which a rain-drop falls depends upon its size. 
 The steady speed ultimately attained is called the " terminal 
 velocity " and Lenard has prepared the following table* (Table II) 
 showing how the terminal velocity in still air varies with the 
 size of the drops. 
 
 * Extracted from the " Meteorological Glossary " (4th issue) : Meteoro- 
 logical Office, London. 
 
43 
 
 TABLE II. 
 Terminal Velocities of Water Drops falling in Air. 
 
 Diameter 
 of Drop. 
 
 Terminal Telocity. 
 
 Diameter 
 of Drop. 
 
 Terminal Velocity. 
 
 
 
 
 miles/ 
 
 
 
 
 miles/ 
 
 mms. 
 
 ins. 
 
 metres/sec. 
 
 hour. 
 
 mms. 
 
 ins. 
 
 metres/sec. 
 
 hour. 
 
 o-oi 
 
 0-0004 
 
 0-0032 
 
 0-007 
 
 3-0 
 
 0-118 
 
 6-9 
 
 15-4 
 
 o-i 
 
 0-0039 
 
 0-32 
 
 0-71 
 
 3-5 
 
 0-138 
 
 7-4 
 
 16-5 
 
 0-5 
 
 0-020 
 
 3-5 
 
 7-9 
 
 4-0 
 
 0-157 
 
 7-7 
 
 17-2 
 
 1-0 
 
 0-039 
 
 4-4 
 
 0-8 
 
 4-5 
 
 0-177 
 
 8-0 
 
 17-9 
 
 1-5 
 
 0-059 
 
 5-7 
 
 12-6 
 
 5-0 
 
 0-200 
 
 8-0 
 
 17-9 
 
 2-0 
 
 0-079 
 
 5-9 
 
 13-2 
 
 5-5 
 
 0-216 
 
 8-0 
 
 17-9 
 
 2-5 
 
 0-098 
 
 6-4 
 
 14-3 
 
 
 
 
 
 The frictional resistance offered by air to a drop depends 
 upon the relative motion of the two. It does not matter 
 whether the air be still and the drop moving, or vice versa, or 
 whether both are moving with different velocities. The 
 terminal velocities given in Tables II may be looked upon, 
 therefore, as the velocities with which air must be moving 
 vertically upward to keep drops of the diameters stated floating, 
 without rising or falling. Lenard, in fact, actually obtained 
 his results from experiments with vertical air currents and 
 drops of known size. 
 
 The important fact to notice is that the terminal velocities 
 do not increase indefinitely with the size of the drops, but that 
 when the drops attain to a certain diameter the terminal 
 velocity remains constant at 8 metres per second (25 feet per 
 second or 17 '9 miles per hour) for further increases in size. 
 
 This is due to the fact that the drops become deformed, 
 flattening out horizontally so that the air resistance is increased. 
 For diameters greater than 5'5 mms. (0'216 ins.) the deforma- 
 tion becomes so great that the drops break up before the 
 terminal velocity of 8 metres per second is reached. 
 
 It follows, therefore, that no rain can fall through a current 
 of air ascending at a greater velocity than 8 metres per second. 
 In upward currents of velocity greater than that limit, the drops 
 are either carried up intact or after being broken up into 
 smaller drops. 
 
 It is known that upward currents having such velocities do 
 exist from time to time. The drops formed in such currents 
 are afforded ample opportunity to increase in size, and it is 
 therefore probable that such currents carry up with them great 
 quantities of water. The large drops so formed have two ways 
 of reaching the earth ; first, by being carried on into the outflow 
 of air above the region of violent convection, or second, by a 
 sudden stoppage of the upward convection. Such sudden 
 
44 
 
 stoppages are probably the cause of the very heavy rains 
 experienced in what are termed, " cloud bursts." 
 
 51. The Formation of Rain. It is pertinent at this point to 
 outline briefly the process which causes rain. When the wind 
 circulation is such that the air currents flow towards a common 
 centre, as is the case in cyclones (see Part II.), the only way by 
 means of which the air can escape is by upward motion. Its 
 ascent is accompanied by the phenomena of expansion and 
 cooling. When the dew point is reached condensation 
 commences upon the nuclei present. The drops formed at 
 first are very small, probably of the order diameter ' 002 mms., 
 and thus they are carried upward with the rising air. The 
 further ascent means further cooling and more condensation of 
 water upon the drops, thus increasing their size. When the 
 diameter reaches 0*2 mms., the drops are falling relatively 
 to the air at the rate of about 1 metre per second, which is 
 more rapid that air currents normally rise and thus the cloud 
 begins to rain. 
 
 52. Condensation at Temperatures below the Freezing 
 Point. In a consideration of condensation occurring at tempera- 
 tures below the freezing point, there are two cases to examine : 
 
 (a) when the condensation commences before the tempera- 
 ture falls below the freezing point ; 
 
 (6) when condensation only commences after the freezing 
 point has been passed. 
 
 The first of these two cases comes into operation when the 
 ascending currents continue until the temperature of 32 F. 
 (the freezing point) is reached and passed, the dew point being 
 above 32 F. The point arises as to whether the water drops 
 freeze after the freezing point is passed and, if they do, 
 what form do they take ? 
 
 Now, observations have been made in balloons and on 
 mountain tops, and these have revealed the fact that water 
 drops, certainly in the liquid form, can co-exist with temperatures, 
 as low as F. It is clear, then, that small drops can be super- 
 cooled a long way without solidifying. Exactly how far such 
 super-cooling can be carried is not known but, in any case, the 
 state is one of instability. If, for any reason, a few drops do 
 solidify, they find themselves at once in a greatly super-saturated 
 atmosphere and condensation in the solid form rapidly takes 
 place upon them. The equilibrium over the other drops is, at 
 the same time, destroyed and the water drops rapidly evaporate, 
 the water distilling from the drops on to the ice particles. 
 
 Every big cloud formed as a result of an upward current of 
 air is divisible into three parts. First, there is the lowest layer 
 where the cloud particles are in the form of liquid water drops ; 
 next, there is a region where the cloud particles are still water 
 drops but in which the latter are more or less super-cooled ; 
 
45 
 
 and thirdly, there is the highest region of the cloud where the 
 drops are frozen into ice, this region being usually called the 
 " snow region." 
 
 53. Hail. There is no sharp line of demarcation between 
 the snow region and the region of super-cooled water. For 
 quite a distance ice crystals and super-cooled water are co- 
 existing. If a super-cooled water drop impinge against an ice 
 crystal it at once solidifies and, at the same time, imprisons a 
 little air. Consequently, the ice crystal increases in size and 
 will begin to fall if the velocity of the ascending air be not too 
 great. But as it falls it passes through the entire length of the 
 region of super-cooled water with the result that it grows still 
 more, and imprisons more air, arriving at the bottom of the 
 super-cooled region as a ball of soft ice without any definite 
 crystalline structure. 
 
 If the velocity of fall of the ball through the super-cooled 
 region has been at all great, it arrives in the region where the 
 temperature is just 32 F. with its own temperature lower than 
 32 F. Consequently any further falling means that water is 
 directly deposited upon it, which water covers the surface -of 
 the ball with a uniform layer of liquid which quickly freezes as 
 a hard, transparent covering with little or no air imprisoned in 
 it. Finally the ball falls out of the base of the cloud as a 
 hailstone. 
 
 Hailstones cut open to reveal their structure show the soft 
 core of white ice and the hard transparent layer of clear ice 
 which are in conformity with the theory of formation which 
 has been outlined. 
 
 But there is a further addition to be made to the life story 
 of large hailstones. An ascending current of air is not steady 
 but composed of increases and decreases, of gusts and com- 
 parative lulls. A hailstone which has reached the lower part 
 of the cloud may meet one of the gusts and be blown upwards 
 and thus be forced to go through the whole process again, a 
 layer of white, soft ice being formed around the previous 
 transparent layer, and another transparent layer being added 
 around that. Further repetitions of the carrying up add 
 additional layers. A hailstone, if cut open when it finally falls 
 after such a protracted history, would show several concentric 
 layers of white and clear ice. Such hailstones have frequently 
 been found. 
 
 Two conditions, then, seem essential for the production of 
 large hailstones : 
 
 (a) the clouds must possess great vertical structure so that 
 the three regions the water region, the super-cooled 
 region, and the snow region are well developed. 
 
 (6) there must be present ascending currents of air of 
 considerable violence so that the hailstones, once 
 formed, may be whirled up and down several times. 
 
46 
 
 Warm regions best favour the production of these conditions 
 .for there the high temperature reached by the ground during 
 the hours the sun is well above the horizon favours the forma- 
 tion of violent ascending currents and, at the same time, 
 condensation starts at comparatively high temperatures, thus 
 making the water and the super-cooled regions of the clouds 
 to be of extensive depth. 
 
 54. Soft Hail. If the point where the temperature is just 
 32 F. (the freezing point) be near to the ground, the hailstones 
 in falling from that point to the ground do not have much 
 water deposited on them and so the hard, transparent layers 
 (see section 53) covering the soft, white core are more or less 
 absent and the hail arrives at the ground as just the original 
 soft white ice. It is then known as " soft hail." It is assumed 
 that there are no violent upward currents present so that the 
 hail is not subjected to rapid whirlings up and down. 
 
 55. Further Consideration of Condensation at Temperatures 
 beloic the Freezing Point. In section 52 it was stated that there 
 are two cases of condensation occurring at temperatures below 
 the freezing point. The first of these two cases has already 
 been considered in sections 52-54. The second case is when 
 the condensation only commences after the freezing point has 
 been passed. 
 
 Dr. Simpson approaches the subject from a study of fogs 
 and iridescent clouds seen in the Antarctic.* He describes one 
 occasion, for example, on which he was enveloped in a fog during 
 sunshine. A white bow was seen on the fog opposite the sun 
 in the position usually occupied by a rainbow. The formation 
 of such a bow, however, could only be explained on the 
 assumption that the fog was composed of small drops. The 
 temperature of the air was 21 F. Experiments on iridescent 
 clouds also led him to the conclusion that the effects observed 
 in them, as well as in the fog, could only be produced by the 
 refraction of light by small spheres. Iridescent clouds, how- 
 ever, are only formed in very low temperatures in polar regions 
 or at very high elevations in temperate ones. His general 
 conclusion then was, that super-cooled water exists at much 
 lower temperatures than is usually thought possible. 
 
 It is clear that the air in which the water drops formed to 
 give the effects already referred to was far below the freezing 
 point in temperature. The problem is how water drops can 
 form at such a temperature. 
 
 Dr. Simpson"!" states that the solution of the problem is to be 
 found in the chemical fact enunciated in the following 
 quotation J : " As has been shown time and again, when a new 
 
 * See Quarterly Journal of the Royal Meteorological Society. Yol. 38. 
 p. 291. 1912. 
 
 t Manuscript of lecture on " The Water in the Atmosphere." 
 Taylor, " The Chemistry of Colloids," p. 18. 
 
47 
 
 phase, which is finally solid, makes its appearance suddenly, 
 whether from vapour or solution, it appears first as a liquid ; 
 it may run through many intermediate (labile) forms before 
 reaching its final solid form. Sulphur, for instance, forms 
 globules, which crystallise later. Crystallisation as spherolites 
 
 " is well known." 
 
 When, then, water vapour condenses at low temperatures it 
 
 forms, first of all, in liquid drops which crystallise afterwards. 
 
 56. Snow. Once the first crystals are formed, further 
 condensation takes place, but now the vapour condenses directly 
 into the solid form without going through the intermediate 
 liquid stage. Water crystallises in the hexagonal form. Snow 
 flakes are formed of one or more ice crystals and the variety of 
 patterns formed as the result of the agglomeration of the crystals 
 is almost infinite. When snow falls with comparatively high 
 temperatures, large, wet flakes are often the result ; with lower 
 temperatures, the flakes are smaller ; and, if the temperature 
 be much below the freezing point (32 F.) "snow dust" results, 
 that is, the crystals are so small as to appear like dust. 
 
 57. Sleet. Sleet, which is such a common phenomenon in 
 such countries as the British Isles, may be defined as the 
 precipitation of snow and rain together, or of partially melted 
 snow. It is, perhaps, snow which has fallen through an 
 inversion of temperature. The layer of comparatively warm 
 air thus encountered causes a partial liquefaction, especially if 
 the temperature be much over 32 F. as it often is in an inversion 
 comparatively near the ground in temperate latitudes. 
 
 58. Glazed Frost. Glazed frost is, in England, a compara- 
 tively rare phenomenon!. It is important, however, inasmuch 
 as when it occurs it greatly impedes transport, even rail 
 locomotion being rendered difficult. In the previous section 
 it was stated that sleet is caused by snow passing through 
 an inversion of temperature, in which passage it is partially 
 liquefied, the temperature experienced being over the freezing 
 point. If however, the surface temperature is below the 
 freezing point, the partially melted snow will re-freeze when 
 it reaches that surface so that a layer of ice is formed on all 
 objects on the ground exposed to the precipitation. It is this 
 layer of ice which consitutes glazed frost. 
 
CHAPTER X. 
 
 Weather. 
 
 59. The Beaufort Notation. It is found extremely con- 
 venient to have some notation in which, the weather, either 
 prevailing at the time or which has been experienced, can 
 be briefly but comprehensively expressed. Such a notation 
 has a great value for use on synoptic charts, a consideration 
 of which will be found in Part II of this book. 
 
 For use in the United Kingdom, Admiral Sir Francis 
 Beaufort who, it will be remembered, has already been mentioned 
 in connection with the Beaufort Scale for wind, introduced 
 a system of notation consisting as a rule of the first letter of 
 the phenomenon to be indicated. Various additions to his 
 notation have been made from time to time. 
 
 The Beaufort Weather Code, however, makes no reference 
 to several important matters such as gales or the various optical 
 phenomena which frequently occur. Moreover, it is not inter- 
 national in character ; and internationalism is a very desirable 
 asset in such a study of world-wide phenomena as is meteoro- 
 logy. Consequently, by international agreement, a code of 
 symbols in place of letters has been arranged. Neither the 
 Beaufort Scale nor the International Code of Symbols is 
 quite comprehensive and, naturally, overlapping occurs. The 
 complete scheme of both is given for purposes of reference. 
 See Table III). 
 
 TABLE III. 
 
 The Beaufort Weather Code and the International 
 Weather Symbols. 
 
 Beaufort 
 Letter. 
 
 International 
 Symbol. 
 
 
 
 
 b 
 be 
 
 ... 
 
 Blue sky, cloudless, or not more than 
 covered. 
 A combination of blue sky with detached 
 
 one quarter 
 clouds. 
 
 c 
 o 
 
 
 Sky mainly cloudy, but with openings 
 clouds. 
 Completely overcast. 
 
 between the 
 
 g 
 
 ... 
 
 Gloom. 
 
 
 u 
 
 ... 
 
 Ugly, threatening sky. 
 
 
 e 
 
 ... 
 
 Wet air, without rain falling. 
 
 
International 
 Symbol. 
 
 or 
 
 
 
 A 
 
 A 
 
 E 
 4* 
 
 T 
 K 
 
 00 
 
 CL 
 
 V 
 
 CD 
 ID 
 
 Continuous rain. 
 
 Drizzle. 
 
 Snow. 
 
 Passing showers. 
 
 Hail. 
 
 Soft hail. 
 
 Ice crystals. 
 
 Snow on ground. 
 
 Snowdrift. 
 
 Gale. 
 
 Squalls. 
 
 Lightning. 
 
 Thunder. 
 
 Thunderstorm. 
 
 Fog. 
 
 Wet fog. 
 
 Mist. 
 
 Ground fog. 
 
 Dust haze. 
 
 Dew. 
 
 Hoar frost. 
 
 Rime. 
 
 Glazed frost. 
 
 Unusual visibility of distant objects. 
 
 Solar halo. 
 
 Solar corona. 
 
 Lunar halo. 
 
 Lunar corona. 
 
 Rainbow. 
 
 Aurora. 
 
 Zodiacal light. 
 
 O 15406 
 
50 
 
 60. The Control of Weather Changes. The chief factors 
 controlling the weather of any area may be briefly summarised 
 as follows ; 
 
 (a) The diurnal variation of insolation from day to night. 
 
 (6) The annual change of insolation from summer to 
 winter. 
 
 (c) The passage of cyclonic and anti-cyclonic areas and theii 
 attendant smaller storms. (An extended consideration 
 of these systems is given in Part II. of this book.) 
 
 61. The General Weather of the Various Zones. In order to 
 gain a broad view of weather in its general relation to latitude, 
 it is interesting to consider very briefly the general weather 
 experienced in the various zones into which the globe may be 
 divided. 
 
 62. The Torrid Zone. The Torrid Zone, which embraces 
 about half of the surface of the globe and has a large oceanic 
 area, is chiefly characterised by a regular sequence of diurnal 
 changes, with a comparatively regular and steady change from 
 season to season. The days are very much alike ; clouds increase 
 in the morning and decrease in the evening, and the night skies 
 are mostly clear. Rain occurs chiefly in the afternoon when it 
 occurs at all as, at that period, convectional up currents have 
 
 -reached their maximum. This regular daily routine of weather 
 changes is seldom interrupted by cyclonic storms. 
 
 The Torrid Zone should, however, be divided into two 
 sections ; first, the " Doldrums," and second, the trade wind 
 belts. 
 
 The weather of the " Doldrums," the region between the 
 trade wind belts, is hot, moist and cloudy with frequent rains 
 returning day after day with much regularity. Calms are 
 frequent, and the wind generally does not exceed a few miles 
 per hour in velocity. Thunderstorms of terrific violence occur 
 from time to time. As would be expected, that part of the 
 region which has land as the greater part of its surface has 
 much more pronounced weather changes than does the part 
 which is mostly sea. 
 
 The trade wind belts at sea are characterised by the 
 constancy of the winds which blow from approximately the 
 same direction and with the same speed both day and night. 
 Moderate amounts of cloud occur in the daytime, but are 
 dissipated as night approaches, and rain is not frequent. 
 Cyclones occasionally enter these belts and control both weather 
 and wind. The annual and diurnal ranges of temperature are 
 small. 
 
 In the trade wind belts on land, the dry ness of the air and 
 the increase in the diurnal range of temperature compared with 
 the belts at sea are the most remarkable features. In the hot 
 seasons the temperature during the day may reach to a very 
 high maximum, but the nights are cool. 
 
51 
 
 
 
 The monsoon regions, especially of India, provide the 
 greatest extremes experienced in the Torrid Zone. The weather 
 in one season is comparatively cold and dry ; that in the other 
 warm and sultry with much rain. 
 
 63. The Temperate Zones. The Temperate Zones differ from 
 the Torrid Zones in that in them chief control of the weather 
 is taken by the non-periodic phenomena of cyclones and anti- 
 cyclones. Consequently, the winds and weather do not show 
 the regularity which characterises the zone already considered. 
 Weather is not diurnal, and the changes are more frequent and 
 more capricious. A study of the weather and wind sequences 
 associated with the travelling systems of cyclones and anti- 
 cyclones, which will be given later, will reveal more clearly 
 what these changes are. 
 
 64. The Frigid Zones. It must be remembered that in very 
 high latitudes the diurnal control of weather is completely, or 
 almost completely, lacking. For two or three months in the 
 depth of winter the sun does not rise above the horizon ; for 
 two or three months in the middle of summer it does not sink 
 below it. The weather changes much as it does in the lower 
 latitudes of the Temperate Zones according to the control 
 of passing cyclones and anti-cyclones ; and in areas, like 
 North Russia, which are inhabited the main features of the 
 meteorological year are the times of the thawing and freezing of 
 the seas and rivers. 
 
 PART IL-SYNOPHC METEOROLOGY. 
 
 CHAPTER XI. 
 
 Introductory. 
 
 65. General. Modern meteorology must be regarded as the 
 work of an organisation and not of any individual. One of the 
 chief duties of a Meteorological Office now-a-days is the making 
 of forecasts of weather, and this involves the co-operation of 
 hundreds of people. The making of synoptic charts is an 
 essential part of the process of forecasting. By a synoptic 
 chart is understood a map of the geographical area under 
 consideration showing the distribution of the various meteoro- 
 logical elements such as wind, pressure, temperature and the like 
 over that region for a given point of time. Such maps are 
 usually prepared four times daily for fixed hours of observation, 
 which, in the British Isles, are 1 o'clock and 7 o'clock in the 
 morning, and 1 o'clock ^ and 6 o'clock in the afternoon, times 
 being Greenwich Mean* Time. Their preparation connotes a 
 corps of observers and a well arranged distribution of meteoro- 
 logical stations. It is the duty of such stations to take, at the 
 
 D 2 
 
52 
 
 prescribed hours, careful observation of the meteorological 
 elements required and to transmit the results of their observations 
 to the Central Office, to reach that Office as soon as possible. 
 
 The speed of transmission of these messages is obviously 
 very .important. The synoptic charts have to be prepared, and 
 the forecasts of coming weather based upon them have to be 
 issued as soon as possible, for speed in issuing forecasts is a 
 most commendable meteorological virtue, inasmuch as weather 
 changes in such a region as the British Isles occur, sometimes, 
 very rapidly, and these changes, as in the case of gales, may 
 involve much damage and even loss of life. Forecasts issued 
 promptly may often enable the authorities concerned to take 
 such precautions as may minimise both the damage to property 
 and the danger to life. 
 
 Here, then, in the question of speed of transmission enters a 
 third factor ; the factor of adequate and rapid means of com- 
 munication between the observers and the Central Office, where 
 the maps are prepared and the forecasts issued. Ordinary 
 telegraphy, wireless telegraphy, and the telephone have done 
 much to secure the efficiency of modern forecasting. Each link 
 in the chain is very important ; the observers at the outlying 
 places of observation must be thoroughly competent, the means 
 of communication they have with the Central Office must be 
 rapid and exact, which is a matter of thoroughly competent 
 telegraphists, and the forecaster drawing the charts and fore- 
 casting from them must also be thoroughly competent. 
 
 But it cannot be too much emphasised that even an expert 
 meteorologist is not able to do much in the way of good 
 forecasting if he has no current synoptic chart. No amount of 
 knowledge of weather saws and maxims is of much avail. 
 Even a thorough knowledge of cloud forms is not sufficient, 
 and the " weather glass " with the legends around its dial 
 is a snare and a delusion. Synoptic charts, properly inter- 
 preted, are the essential matter in sound forecasting ; and the 
 interpretation involves much appeal both to physics and to 
 precedent. 
 
 CHAPTER XII. 
 
 Pressure and its Measurement. 
 
 66. The Pressure of the Atmosphere. Perhaps the most 
 important meteorological element entered upon synoptic charts 
 is the pressure of the atmosphere as determined at the various 
 meteorological stations contributing to the preparation of those 
 charts. 
 
 Pressure is the most important of the meteorological elements 
 because all the other elements wind, humidity, and the rest 
 seem to depend upon it, or, rather, upon its changes. The 
 characteristic of fluid pressure which must be borne in mind 
 
53 
 
 and air is a fluid is that of transmission. If water be considered 
 that water will flow through any crevice, no matter how small, 
 in order " to find its own level." The same is true of air, except 
 that air does not take so long. In any room there are some 
 crevices and through these crevices the air will flow in the 
 endeavour to settle down in equilibrium, and such equilibrium 
 is attained very rapidly. Therefore, the atmosphere is regarded 
 as having the same pressure at the same levels so long as the 
 air is not moving ; and movement of the air is regarded as a 
 component in the equalising process. 
 
 67. The Measurement of Pressure. The pressure of the 
 atmosphere is measured by means of barometers, and barometers 
 are of two kinds ; first, those in which the pressure is balanced 
 by the weight of a column of liquid, and second, those in which 
 movements of the flexible lid of a box which is nearly exhausted 
 of air are recorded. 
 
 The barometers in which a column of liquid is the main 
 feature will be considered first. The making of a simple 
 barometer of this type is an easy matter. A glass tube, about 
 33 inches long and closed at one end, is taken and completely 
 filled with mercury. It is next inverted with a finger pressed 
 tightly against its open end. The next step is to immerse the 
 finger, and the end it is closing, under the surface of mercury 
 contained in a bowl. The tube is brought into the vertical 
 position and the finger withdrawn. On withdrawing the finger 
 the mercury drops. When things have become steady it is 
 found that a column of mercury, about 30 inches in height 
 above the level of the mercury in the bowl is still maintained 
 in position in the tube. It is so supported by the pressure of 
 the air acting downwards on the surface of the mercury in the 
 bowl. The space above the mercury in the tube is a vacuum, 
 that is, it is practically entirely free from air, and is termed the 
 " Toricellian vacuum " after Torricelli, the Italian, who first 
 performed the experiment which has been described in 1643. 
 Repetitions of the above experiment performed in rapid 
 succession 011 the same day may show no difference in the 
 height of mercury supported. A daily experiment done for 
 several days in succession, however, will reveal that the height 
 supported varies appreciably from day to day. 
 
 The experiment could be repeated using other liquids than 
 mercury. Water, for example, could be used but as mercury 
 is, approximately, 13 \ times as heavy as water, it is clear that, 
 if the atmosphere supports about 30 inches of 'mercury, it would 
 support about 405 inches of water, or nearly 34 feet, which 
 would make the length of tube required extremely long and 
 cumbersome in working. Nearly all liquids suffer from the 
 same practical disadvantage and mercury, on account of its high 
 density, remains clearly marked out as the most suitable liquid 
 for barometric purposes. Moreover, its opacity makes it 
 peculiarly easy to read. A great additional advantage is that 
 
54 
 
 its vapour lias a very low pressure at ordinary temperatures and 
 this makes possible the existence of a nearly perfect vacuum 
 above it. 
 
 The standard mercurial barometers, such as the Fortin 
 pattern and the Kew pattern, used in metereological stations, 
 are made on the same principle as the simple barometer whose 
 construction has already been described. For details of the 
 refinements introduced to obtain a high degree of accuracy in 
 the Fortin and Kew patterns the student is referred to " The 
 Observer's Handbook," issued by the Meteorological Office, Air 
 Ministry, London, and published by His Majesty's Stationery 
 Office. 
 
 68. Corrections to be applied to Barometer Readings. The 
 height of the mercury as read needs to be corrected for 
 
 (a) The temperature of the instrument ; 
 
 (b) The latitude of the place where the instrument is set up ; 
 
 (c) The height of the cistern of the instrument above sea 
 
 level. 
 
 These corrections are necessary in order that the pressures 
 when plotted on the synoptic charts shall be strictly comparable. 
 The correction for the temperature of the instrument is due 
 to the errors introduced by the expansion and contraction of 
 both the mercury and the metal scales. If two barometers were 
 placed at exactly the same level, one in the warm air of a room, 
 the other in the colder air outside in the open, the one in the 
 warm room would read the higher, but the actual pressure 
 exerted by the atmosphere as the levels are equal would be the 
 same. The correction is carried out and strict comparison is 
 secured by reducing all readings to what they would be if the 
 temperature of the instrument were some fixed temperature. 
 This fixed temperature is taken to be 32 F. 
 
 The curvature of the earth is not perfect and the force of 
 gravity varies according to latitude, reaching a maximum at the 
 poles and a minimum at the equator. Thus, if there were 
 uniform atmospheric pressure over the whole globe, a mercurial 
 barometer at the equator would read higher than one at either 
 pole. To avoid errors due to this variation of gravity with 
 latitude all barometric readings are corrected to what they 
 would be if the instrument were in some standard latitude, 
 which latitude is taken to be 45. 
 
 The correction for height above sea level is clearly necessary 
 if readings all over an area are to be comparable in view of the 
 fact that pressure decreases with height. What is needed in 
 synoptic charts is a knowledge of the changes of pressure due 
 to general causes and not to such ones as are due to local con- 
 figuration. This error is avoided by correcting the readings to 
 what they would have been if situated at a standard height. 
 The standard height selected is that of mean sea level, which 
 differs slightly from country to country, but in the British Isles 
 
55 
 
 is an arbitrary level at Liverpool. The differences of the level 
 taken as sea level in different countries are not great enough 
 seriously to vitiate the use of the readings from those various 
 countries as comparable upon synoptic charts. 
 
 The barometric pressures, then, appearing on the synoptic 
 charts are those which would have been read if all the 
 barometers supplying readings were at a temperature of 32 F., 
 at a height equal to mean sea level, and situated in latitude 
 45 N. or S. 
 
 69. Barometric Units. The purpose of the barometer is to 
 give the value of the pressure of the air. Owing to the 
 convenience of the mercury barometer, it became the custom to 
 express the pressure in terms of the height of the column of 
 mercury, inches being used in England and millimetres on the 
 continent of Europe. If a water barometer had been in general 
 use, the value would no doubt have been expressed in feet or 
 metres of water. In an aneroid barometer neither inches of 
 mercury nor feet of water are intrinsically suitable, but never- 
 theless inches of mercury formed the standard in this country. 
 
 Pressure is, however, not a length but a force per unit area, 
 and it is usually expressed in this way except in the particular 
 case of the pressure of the atmosphere. For example, the 
 pressure in a steam engine is expressed in Ibs. per square inch. 
 
 By a fortunate coincidence the pressure of the atmosphere 
 near the earth's surface is very nearly one million times the 
 fundamental unit of pressure in the metric system, that is, 
 1 dyne per square centimetre, and it is therefore not merely of 
 scientific value but also of practical convenience to use this 
 system. The millibar is 1,000 of these fundamental units and 
 the pressure of the atmosphere near the earth's surface is 
 approximately 1,000 millibars. The millibar is a conveniently 
 sized unit and mercury barometer can be graduated in such a 
 way that the pressure in millibars is read off with great ease. 
 
 The derivation of the millibar is now given more com- 
 pletely. 
 
 The unit of force on the C. G. S. (centimetre -gramme- 
 second) system, now used universally for most scientific work, 
 is the force which produces an acceleration of 1 centimetre per 
 second in a mass of 1 gramme, and is termed a dyne. 
 
 The unit of pressure in the C. G. S. system, as already 
 stated, is 1 dyne per square centimetre. 
 
 The dyne is, however, a very small unit and so a mega- 
 dyne, that is, one million dynes, called a bar, is adopted as the 
 practical unit of force, and this leads to the use of the 
 millibar. 
 
 1 bar is equal to the pressure of 1 mega-dyne per square 
 centimetre, that is, to 1,000,000 dynes per square 
 centimetre; 1 millibar = rtiW of a bar and is therefore 
 equal to the pressure of 1,000 dynes per square 
 centimetre. 
 
56 
 
 1,000 millibars, that is, 1 bar, are equivalent to the pressure 
 of a column of mercury 29 '531 inches (750 '1 millimetres) high 
 at a temperature of 32 F. (0 C) under the conditions of 
 gravitation existing in latitude 45 N. or S. 
 
 In Table IV. will be found values facilitating conversion 
 from one system of barometric units to another. The values 
 refer to standard conditions of temperature and latitude. 
 
 TABLE IV. 
 Conversion Table for Pressure. 
 
 Inch. 
 
 Millimetre. 
 
 Millibars. 
 
 1 
 
 25-4 
 
 33-9 
 
 0-0394 
 
 1 
 
 1-33 
 
 0-0295 
 
 0-75 
 
 1 
 
 The principal advantage of using the millibar as a unit is 
 that a pressure quoted as a number of millibars is completely 
 definite ; millibars can be used just for pressure and for 
 nothing else. 
 
 As the reading of a barometer is not independent of either 
 the temperature of the instrument or of the latitude of the place 
 where it is set up, it follows that the 1,000 millibar graduation 
 can only be correct for one particular temperature at each 
 latitude. The temperature appropriate to latitude 45 is 
 termed "the standard temperature" of the barometer. The 
 temperature at which the 1,000 millibar graduation reads 
 correctly in any other latitude is termed " the fiducial 
 temperature" for that latitude. Both the c< standard" and the 
 " fiducial" temperatures refer to sea level, and it should be noted 
 that the "standard" temperature in latitude 45 is also the 
 " fiducial " temperature for sea level in latitude 45. 
 
 It needs to be recognised, however, that the conditions of 
 temperature and latitude which ensure that the 1,000 millibar 
 graduation is correct are not necessarily such as make gradua- 
 tions in other parts of the scale correct. The " standard " 
 temperatures applicable to latitude 45 for several readings of 
 the scale are usually supplied with a good barometer graduated 
 in this system, and from these "standard" temperatures the 
 corresponding "fiducial " temperatures for the various readings 
 of the scale given can be computed. 
 
 In any particular case, then, the barometer graduated in 
 the millibar system is read as well as the thermometer attached 
 to the instrument giving the temperature of that instrument. 
 The " fiducial " temperature for that part of the scale where 
 the^mercury level stands is then read from previously computed 
 Jbables. The difference between this '"fiducial" temperature 
 
57 
 
 and the attached thermometer reading is noted and the cor- 
 rection to be added or subtracted to correct the reading down 
 for temperature is read from prepared tables. The remaining 
 correction is for height above sea level. Once again this is 
 done by reference to tables and here it is the temperature of the 
 outside air which counts. 
 
 For a fuller explanation of the methods of correction by 
 "fiducial" temperatures and for the necessary formulae and 
 tables used in that correction, the reader is referred to the 
 " Observer's Handbook " issued by the Meteorological Office, 
 Air Ministry, London. 
 
 70. Aneroid Barometers. The second type of barometer in 
 general use is that in which the compression or recovery of a 
 box which is nearly exhausted of air and has a flexible lid is 
 recorded. Such barometers are termed aneroids. An aneroid 
 consists generally of a circular metallic chamber hermetically 
 sealed which has been rendered practically vacuous. The lid 
 is made of very thin elastic metal, and by an arrangement of 
 levers and springs the movements of this elastic lid are recorded 
 by a needle moving over a dial. 
 
 The pressure within the box is very small, whilst the lid is 
 carrying the weight of the outside air upon it. Any increase 
 in outside pressure pushes in the lid slightly and any decrease 
 allows the lid to recover. If the dial be properly calibrated 
 the needle recording the movements of the lid gives the pressure 
 of the outside air. 
 
 Aneroids do not require correction for latitude or tempera- 
 ture, but only for height above sea level and for index error. 
 Whilst extremely suitable for explorers and for mountaineers 
 on account of their portability and comparative lack of fragility 
 as compared with mercurial barometers, aneroids are not 
 accepted as being sufficiently accurate for use at meteorological 
 stations as their mechanism introduces errors beyond the limits 
 allowed. 
 
 Aneroids are very useful in the rapid determination of 
 heights. As a rough rule it may be said that the barometer 
 falls 30 millibars per 1,000 feet of elevation for low altitudes. 
 Starting, then, with a normal pressure at sea level and marking 
 that as feet, the dial of the aneroid may be calibrated in 
 heights instead of pressures. Arrangements can be made to 
 enable the zero height mark to. be adjusted to the actual 
 pressure existing at any place, so that heights above that place 
 can be read off to a fair degree of accuracy directly from the 
 dial. 
 
 71. Barographs. The ordinary barograph is really an 
 aneroid barometer fitted with a lever recording on a chart or a 
 drum, the drum being kept revolving by clockwork. By using 
 such an instrument a continuous record of the pressure from 
 moment to moment is obtained. The great use of a barograph 
 is to show how the pressure is varying. It must be standardised 
 
58 
 
 daily by comparison with, readings given by a mercury 
 barometer. 
 
 72. The Diurnal Variation of the Barometer. An exami- 
 nation of barograph traces taken in quiet weather over a long 
 period reveals one striking fact, and ' that is a definite diurnal 
 variation of pressure. There is a slight rise of pressure 
 between about 4 o'clock in the morning and 10 o'clock in the 
 morning, and also between about 4 o'clock in the afternoon and 
 10 o'clock in the afternoon, with corresponding falls in between. 
 This diurnal variation is most marked in the Torrid Zone and 
 becomes less distinct with increasing latitude. This would 
 be expected as the cause of the diurnal pressure variation is to 
 be found in the diurnal variation of temperature. Fig. 6 
 shows the variation in diagrammatic form. The variations in 
 the figure are exaggerated. 
 
 MlBHIOKT 
 
 FIG. 6. DIURNAL VARIATION OF PRESSURE. 
 
 CHAPTER XIIL 
 
 The Making of Synoptic Charts and some Lessons derived 
 
 from them. 
 
 73. The Foundations of a Synoptic Chart. The forecaster 
 receives from several stations distributed over the region for 
 which he is concerned observations taken at precisely the same 
 hour. These observations include the wind direction and also 
 its force estimated on the Beaufort Scale (see section 5), the 
 temperature of the dry-bulb thermometer, the pressure corrected 
 down to 32 F., and to sea level in latitude 45, the weather 
 and other special data. He sets the details down beside the 
 location of the particular station on a blank map, showing the 
 wind direction by an arrow with its point actually on the spot 
 marking the station and with feathers equal in number to the 
 Beaufort Scale figure indicating the force, calms being indicated 
 thus : 0. The arrows are drawn so as to be travelling in the 
 wind. The pressure is added in plain figures as is the tempera- 
 ture, whilst the weather existing at the time is either denoted 
 

 
 CO 
 
 O5 
 
 IJ 
 1 
 
59 
 
 by the letters of the Beaufort Notation of Weather (see section 59) 
 or by a conventional picture code. Whether the pressure at the 
 place is rising or falling or steady as denoted by the barograph 
 there is denoted by means of a conventional sign (see Table V.), 
 and the amount that the barograph has varied within the three 
 hours preceding the hour of observation is added, also in a 
 conventional code, generally by a figure showing the number of 
 half-millibars of change. 
 
 TABLE V. 
 
 Conventional Signs to illustrate Barograph behaviour. 
 
 Barograph behaviour. 
 
 Conventional 
 Sign. 
 
 Code 
 Figure. 
 
 Steady or rising - 
 
 O The barometer f 
 
 O or + 
 
 
 
 Rising, then steady <'. 
 
 is now higher 
 
 + o 
 
 1 
 
 Rising, then falling . , 
 
 - Y than, or the { 
 
 + 
 
 2 
 
 Falling or steady, then rising | same as, three 
 
 - + or + 
 
 3 
 
 Unsteady, but rising - 
 
 -J hours ago. |^ 
 
 u + 
 
 4 
 
 Falling 
 Falling, then steady 
 Falling, then rising 
 
 j The barometer r 
 is now lower J 
 
 1 fViQ-n fli-roA 1 
 
 - O 
 
 5 
 6 
 
 7 
 
 ?T e \ dy r t^ir** faUing 1 ^"8 ago" ~ I 
 Unsteady but falling - - J 
 
 O or H 
 U - 
 
 8 
 9 
 
 The actual change in the pressure at the station within the 
 past 3 hours is termed the " barometric tendency " and, as has 
 already been stated, is usually given in the United Kingdom in 
 half millibars. The conventional sign indicating the type of 
 variation is termed the " characteristic " of the tendency. The 
 code figures used in practice to report the variation depend, it 
 will be noticed, on the position of the pen on the chart with 
 regard to its position 3 hours before. 
 
 74. Isobars. The most important thing on a synoptic chart 
 is the pictorial representation of pressure distribution at sea level 
 which is secured by drawing what are termed isobars. Isobars 
 are lines drawn through places at which the pressures corrected 
 for temperature and latitude and reduced to sea level are the 
 same. The isobars are usually drawn for definite pressures 
 with fixed intervals of pressure between them. As it is generally 
 the case that no stations supplying observations actually have 
 these definite pressures, the isobars are drawn by a process of 
 interpolation. 
 
 Figures 7, 8, 9, and 10 should now be carefully studied. 
 They all refer to the conditions over the British Isles at 6 p.m. 
 on the 2nd of August, 1915, and they serve to illustrate the 
 making of a synoptic chart. Figure 7 illustrates the distribution 
 of weather, the Beaufort Scale of Weather Notation (see section 59) 
 being used. 
 
 Figure 8 gives the distribution of wind. Figure 9 shows the 
 distribution of pressure and the isobars are drawn in ; the 
 
60 
 
 isobars drawn are those for 1,000 millibars, ] ,005 millibars and 
 so on at intervals of 5 millibars, and the interpolation involved 
 should be carefully noticed. Figure 10 gives a chart combining 
 all the foregoing particulars, but the weather is now expressed 
 pictorially instead of in the Beaufort Notation, and the area 
 considered is somewhat larger and more isobars become 
 necessary. 
 
 Isobars are almost exactly analogous to the contour lines 
 appearing on an ordinary geographical map ; but whereas 
 contour lines apply to heights, isobars apply to pressures. 
 
 75. Buys Ballot's Law. A study of synoptic charts caused 
 Professor Buys Ballot of Utrecht to enunciate in 1857 the famous 
 law which bears his name. Buys Ballot's law is as follows : 
 If, in the northern hemisphere, you stand with your back to the 
 wind, pressure is lower on your left hand than on your right. 
 In the southern hemisphere the reverse is true for there, if you. 
 stand with your back to the wind, pressure is lower on your 
 right hand than on your left. 
 
 The student should demonstrate the truth of this law to 
 himself by reference to as many synoptic charts as possible, 
 including the one shown in Figure 10. 
 
 A further point to be noticed is, that the winds seem closely 
 to follow the lie of the isobars but shows, generally, a tendency 
 to turn inwards towards the low pressure. This deviation from the 
 true lie of the isobars is generally between and 45, but is 
 sometimes greater. The attention paid by the wind direction 
 to the isobar direction is greatest in polar regions. As the 
 equator is approached the attention paid is less ; it is, however, 
 still quite noticeable at 20 N. or S. latitude. Nearer the 
 equator still the attention paid is very little and, actually, at the 
 equator itself is non-existent. Crossing the equator means that 
 the directions of the wind with regard to pressure are exactly 
 reversed in accordance with Buys Ballot's Law. The change 
 over is not noticeable, however, in practice owing to the fact that 
 the equatorial regions constitute a belt of low pressure, the 
 Doldrums, in which only light and variable winds blow as a 
 rule and in which calms are frequent. 
 
 76. The Strength of the Wind. Another lesson to be 
 derived from a consideration of a collection of synoptic charts 
 relates to the strength of the wind. Hardly any map will fail 
 to reveal to the observant student that where the isobars lie close 
 together the wind is strong, and, where far apart, weak. This 
 is quite what would be expected, remembering the analogy 
 between isobars and contour lines. On a geographical map, 
 contour lines close together indicate steep gradients or slopes 
 down which it may be imagined that, if a stream existed, the 
 water would rush tumultuously ; contour lines far apart, on the 
 other hand, indicate gentle gradients or slopes down which 
 the water in the stream would flow slowly. What applies to 
 
CD 
 
 ill 
 
 I 
 
 
 
 I 
 
 3 C> 
 
 I - 
 ^ ^ 
 '* 
 
 ^J 2 
 
 si 
 
 II, 
 
 H H I 
 
61 
 
 contour lines applies to isobars. Isobars close together, that is, 
 steep pressure gradients, mean that the air flow is rapid, that is, 
 that winds are strong ; isobars far apart, that is, slack pressure 
 gradients, mean that the air flow is gentle, that is, that winds 
 are light. Isobars very far apart mean that the air flow is 
 inappreciable, and calms result. 
 
 77. The Gradient Wind. It will be seen then, that it should 
 be possible to calculate the wind speed at any place from the 
 pressure gradient at that place as shown by a carefully drawn 
 synoptic chart. This can be done and the result obtained is 
 termed the gradient wind. In practice, however, the wind 
 existing at the surface is found to be, as a rule, less than the 
 calculated gradient wind. This is due to the friction of the 
 surface of the earth ; the calculated gradient wind agrees with, in 
 most cases, the actual wind met at about 1,500 feet above the 
 surface. 
 
 CHAPTER XIV. 
 
 Types of Pressure Distribution. 
 
 78. Introductory. Many thousands of synoptic charts of the 
 British Isles have been prepared for various hours, but no two 
 charts have ever been found to be identical ; and what applies 
 to the British Isles applies to any other region. Similar the 
 charts may be with regard to some features, but that is all. 
 The variety of distribution is endless. 
 
 But emerging from the many variations, seven main types*" 
 of isobaric distribution may be recognised. Those seven types 
 are as follow : 
 
 (a) Depression or low ; 
 
 (6) Anti-cyclone or high ; 
 
 (c) Col; 
 
 (d) Secondary depression ; 
 
 (e) V-shaped depression ; 
 
 (/) Wedge of high pressure ; 
 (g) Straight isobars. 
 
 The main features of each of these types will now be 
 considered. 
 
 79. Depressions. Depressions are parts of the atmosphere 
 where the pressure is low. They appear on synoptic charts 
 
 * For a classification of pressure distribution into 28 types, the student 
 is referred to "Aids to Forecasting" (Geophysical Memoirs, No. 16), by 
 E. Gold, F.R.S., issued by the Meteorological Office, Air Ministry, London, 
 1920. This is the best extended classification of types of pressure distribution 
 known to the present author. 
 
62 
 
 as series of closed isobars, approximately oval or circular in 
 shape, with, the minimum pressures occurring in the centres. 
 An example of a depression, with its centre marked L, is shown 
 in Fig. 10. 
 
 The sizes of depressions vary enormously. One example 
 may have a diameter of 1,000 miles or more ; another, a diameter 
 of much less ; others, too, which constitute the tornadoes of the 
 United States of America or the typhoons of the China Sea, have 
 very small diameters, indeed, sometimes under 10 miles. A 
 depression in which the barometer is very much lower in the 
 centre than it is on the margins is said to be deep ; one in which 
 there is comparatively little difference of pressure between the 
 centre and the margins is said to be shallow. 
 
 Alternative names for a depression are " low " and " cyclone." 
 The use of the term "low" has now become popular and little 
 can be said against it ; but care needs to be exercised in using 
 the term "cyclone." In America, especially in newspaper 
 reports, the term '' cyclone " is applied to what is more properly 
 called a "tornado," that is, a depression of extremely small 
 diameter which is accompanied by exceptionally steep pressure 
 gradients, and, consequently, by exceptionally violent winds 
 which cause great damage. In the British Isles, the connotation 
 of "cyclone" is that of a large area of low pressure accom- 
 panied by broad sheets of "cloud, comparatively large areas of 
 rain and winds which may, or may not, be violent. The two 
 uses of the term " cyclone " need to be carefully differentiated. 
 The English use is distinctly the more generally accepted 
 scientific one, and even American meteorologists, as apart from 
 the American general public, so use it, calling their smaller 
 whirling storms by the correct name of " tornadoes." Any use 
 of the word " cyclone " in this book must be taken as having 
 the scientific connotation. 
 
 The winds in a depression obey Buys Ballot's Law (see 
 section 75). One very convenient way of stating the relation 
 of wind direction to depressions is to say that winds blow in an 
 anti-clockwise direction around the centre of low pressure. The 
 winds, too, generally exhibit quite markedly the tendency to 
 turn inwards towards the low centre ; and their force depends 
 upon the steepness of the pressure gradient, that is, upon the 
 distance apart of the isobars. 
 
 80. The Nomenclature of a Depression. There is a very 
 definite nomenclature, due to Abercromby, attaching to depres- 
 sions. As the terms employed will be frequently used in what 
 follows, a diagram (Fig. 11) is given illustrating these terms. 
 Reference is made more fully in section 82 to the travel of 
 depressions, and hence the " path " marked in the figure. The 
 " trough " marks the period of lowest barometer during the 
 passage of a depression over a place. 
 
63 
 
 FIG. 11. NOMENCLATURE OF A DEPRESSION. 
 
 th of 
 )lone. 
 
 - Wind direction at the earth's surface. 
 
 . -> "VVmd direction in the region of the cirrus cloud. 
 
 FIG. 12. DISTRIBUTION OF A CLOUD IN A DEPRESSION. 
 
 81. The Weather in a Depression. An examination of the 
 weather occurring in several examples of depressions as seen on 
 
64 
 
 synoptic charts reveals the fact that there seems to be a fairly 
 definite distribution of cloud and rainfall connected with them. 
 
 The distribution of cloud is shown in Fig. 12, which figure 
 also shows the wind direction both at the earth's surface and at 
 the height of the cirrus clouds. It will be noticed that the 
 whole depression is ringed around with clouds of the cirrus 
 types which are especially prominent in the extreme front of 
 the system. It will be noticed, too, that the cloud lowers from 
 the margins towards the centre, and that the area approximately 
 surrounding the centre is covered with nimbus clouds. This 
 area of nimbus cloud is generally the scene of fairly continuous 
 rain. 
 
 It should be noted, however, that the distribution of rainfall 
 within a depression presents more difficulty in the way of 
 general statement than does the distribution of cloud. Each 
 depression seems to have its own peculiarities of rainfall. 
 Generally, in fact, the rain seems more marked in the left front 
 near the centre than elsewhere. This is to be explained in the 
 following way : The left front of the depression is a region of 
 easterly winds ; the right front, of southerly or south-westerly 
 winds. The easterly winds are colder, and hence denser, than 
 the southerly or south-westerly ones, which, however, so far as 
 the British Isles are concerned, differ from the easterly ones in 
 being saturated or nearly so. This means that the southerly 
 and south-westerly winds are forced to rise over the colder and 
 denser easterly ones, and the ascent is accompanied by dynamical 
 cooling and, as the rising air is of high relative humidity 
 already, much low cloud formation and rain (see section 49). 
 The rain is carried on in the wind a little further and falls, 
 mainly in or near the left front, through the surface easterly 
 currents. 
 
 It should be clearly recognised that the statements already 
 made about the distribution of weather in cyclonic systems 
 can only be taken as average statements and not as definitely 
 applicable to any given depression appearing on a synoptic 
 chart. The depressions, however, which do riot approximate 
 fairly closely to those general statements are, comparatively, few 
 in number. 
 
 82. The Travel of Depressions. An examination of synoptic 
 maps for a succession of days reveals the fact that depressions 
 are not stationary, but travel. Generally speaking, considering 
 their centres, they travel from a westerly point to an easterly 
 one, with a tendency towards north, so that a large number 
 move from south-westerly to north-easterly. 
 
 The result of researches, however, especially for North- 
 West Europe, gives the very disappointing result which has 
 been expressed* as follows: "No path can be drawn so 
 
 * " Forecasting Weather." Sir Napier Shaw, F.R.S. (Messrs. Constable 
 and Co., Ltd.). Chapter XV. 
 
65 
 
 " tortuous and so out of tlie way that we can fairly say it must 
 " be ruled out as a possible path for the centre of a cyclonic 
 " depression." 
 
 Clearly the direction of travel of a depression is a matter cf 
 great importance to the weather of any particular region, inas- 
 much, as has been shown in section 81, definite weather is 
 generally associated with the various parts of the system. The 
 wind, too, varies in the different positions with regard to the 
 centre of low pressure. A change of even a few degrees in 
 the direction of travel of a depression moving towards the 
 particular region under consideration may cause the wind and 
 the weather in that region to be entirely different from what 
 would have been anticipated if the depression had kept on its 
 expected path. 
 
 The forecaster, then, confronted by a depression coming 
 towards the area with which he is concerned, has to predict, as 
 exactly as he possibly can, the course that that depression will 
 take, for everything will depend upon that. It is of little avail 
 to rely upon the average of the behaviour of past depressions : 
 he must deal with the one actually advancing and deal with it 
 as an individual possessing its own characteristics. 
 
 83. To determine the Direction of Travel of any Particular 
 Depression. Without doubt, depressions are the main factor in 
 the weather of such a region as North-West Europe. Conse- 
 quently given one upon, or one affecting, the current synoptic 
 chart, it is the forecaster's first duty to predict the direction 
 of its further travel. 
 
 The first step to accomplish this end is for him to acquaint 
 himself with the past history of the depression so far as he is 
 able so to do from a study of the three or four synoptic 
 charts immediately preceding the actually current one. If the 
 depression has maintained one steady course it will need strong 
 reasons if he predict a change in that course. 
 
 He next, in English practice, considers the barometric 
 tendencies, that is, the changes in the pressures within the 
 three hours immediately preceding the fixed hour of observa- 
 tion at the various stations contributing to the synoptic chart. 
 He then assumes, in default of strong suspicion to the contrary, 
 that the depression will travel towards the place of greatest fall. 
 The objection to be urged against this method is, that the 
 tendencies are past history and that it does not follow that, 
 because the barometer has fallen considerably at a place within 
 the past three hours, it will continue to fall considerably there 
 within the ensuing three hours. In practice, however, the 
 method of predicting travel towards the places where greatest 
 barometric falls have occurred is found to be fairly successful. 
 
 Dr. Ekholm, of Stockholm, 15 or 20 years ago, advocated 
 the use of a new aid to the prediction of the advance of depres- 
 sions. He draws charts in which it is the pressure differences 
 from the last observation which are plotted, and not the actual 
 
 O 35406 E 
 
66 
 
 pressures observed at the time. Thus, lines of equal pressure 
 difference, called isallobars, are obtained. Examination of 
 successive isallobaric charts reveals that the groups of isallobars 
 travel similarly to the groups of isobars ; and Ekholm main- 
 tains that the isallobaric systems travel more regularly than do 
 the isobaric ones. The British Isles suffer in using Ekholm 's 
 method because of their extreme western position ; before the 
 isallobaric groups are well denned, they have travelled past. 
 Isallobaric charts are, however, useful aids in determining where 
 the greatest barometric falls are likely to occur. 
 
 But a more useful aid in practice is to be found in a study 
 of the wind directions at the various stations. By Buys Ballot's 
 Law it is recognised that when the wind direction is changing, 
 the distribution of pressure is changing also. A change of 
 force alone, without any change in direction, just means that 
 the isobars are maintaining their general lie unaltered, but are 
 either crowding more together or becoming wider apart. A 
 change of direction means that the relative distribution of the 
 high and low pressures is changing. The "backing" of a 
 wind, that is, its changing in an anti-clockwise direction, is a 
 sign of the coming of a depression. That this is so is evident 
 from the consideration that one depression follows another 
 depression with a greater or less interval in between. By Buys 
 ^Ballot's Law, the winds in the rear of the depression which has 
 -passed are north-westerly ; those in front of the depression 
 which is approaching are south-easterly. A station, then, in 
 4;he line of advance of the coming depression often reveals the 
 coming of that depression by a backing of its wind long before 
 the centre appears on the chart ; and it is a fair assumption to 
 .say that the coming depression will advance more or less 
 directly towards the station or stations showing the backing. 
 The importance of the meteorological observations taken at 
 ^Valencia, Blacksod Point, and Malin Head, on the western coast 
 of Ireland, can hardly be over-estimated in this respect with 
 regard to depressions from the Atlantic Ocean advancing on to 
 the British Isles. 
 
 Wind study is necessary for yet another method of antici- 
 pating the approach, and predicting the line of advance, of 
 depressions, which method is due to M. Gabriel Guilbert.* An 
 examination of synoptic charts reveals the fact that the wind 
 forces at the various stations are often not in accord with what 
 .would be expected from the pressure gradients at those stations 
 as shown, by the charts, and that sometimes the wind directions 
 differ also from what they should be considering the lie of the 
 isobars. So far as direction is concerned, Guilbert considers 
 ,that the wind is an abnormal one if it deviates more than 40 
 from the run of the isobar. He treats these abnormalities of 
 force or direction as significant, and has formulated a set of 
 
 * " Nbuvelle Methode de Prevision du Temps," par Gabriel Guilbert. 
 (Gaultier Villars, 1900.) 
 
67 
 
 empirical rules in which he employs them as means to .predict 
 the progress of isobaric groups. It is obvious that the method 
 implies that the wind exposures of the stations considered are 
 perfect. This is so seldom the case and abnormalities of wind 
 are produced so often at many stations because of some 
 unavoidable imperfection or another of exposure that the 
 method needs using with the utmost care. 
 
 A selection of Guilbert's empirical rules is now given : 
 
 (a) A depression with winds above normal force (as 
 compared with the pressure gradient) on all sides 
 will fill up ; one with winds below normal force 
 will become deeper, and, hence, depressions which 
 are, apparently, weak will change into cyclonic 
 storms. 
 
 (6) These parts of the depression in which the winds are 
 below normal force, as shown by the pressure 
 gradient, indicate the direction in which the 
 depression will move ; thus, when a depression 
 consists partly of winds that are above normal and 
 partly of winds that are below normal, it will move 
 in the direction of least resistance, that is, in the 
 direction where the winds are below normal. 
 
 (c) Pressure increases from right to left across the line of 
 
 winds too strong for the gradient, In other words, 
 a wind above normal transfers pressure from the 
 right to the left. 
 
 (d) Depressions are attracted towards the regions of least 
 
 resistance ; towards regions of light airs or calms, 
 and, above all, of divergent winds (divergent, if 
 blowing in opposition to the normal direction 
 corresponding to the run of the isobars). 
 
 (e) The existence of several anomalies in the directions 
 
 and forces of winds at various stations in a limited 
 area at the same time decides the sudden formation 
 of an important depression over that area. 
 
 It must be confessed, however, that, whilst in Guilbert's own 
 hands his method has given some astonishingly successful 
 forecasts, other forecasters have found it difficult to use with 
 much success. As has been already said, it needs to be used 
 with extreme care, and with a real knowledge of the wind 
 exposures of the stations where the abnormalities are occurring. 
 
 To sum up, it cannot be said that the line of advance of a 
 depression can yet be predicted with certainty ; but a 
 forecaster, diligent in watching the various points already 
 mentioned, and diligent also in appealing to experience, can be 
 very successful. 
 
 It is interesting to notice that the centres of depressions 
 appear to have particular likes and dislikes, and these are 
 useful, sometimes, in predicting their courses. They seem, for 
 example, to prefer to travel over water in preference to land. 
 
 E 2 
 
68 
 
 Quite often, a depression in North West Europe travels down the 
 North Sea. This particular track is responsible for bringing 
 Eastern England very disagreeable weather in winter, snow and 
 sleet abounding with squalls in the north-westerly or northerly 
 winds in the rear of the cyclonic centre. The Bristol Channel, 
 too, is a very favourite depression track. Another peculiarity is, 
 that on land, depression centres seem to prefer to keep to the 
 low land, if possible. 
 
 84. The Rate of Advance of Depressions. Depressions 
 advance at very variable speeds. Across the British Isles, 
 a speed of 20 to 30 miles per hour is very usual : sometimes, 
 however, this speed is considerably exceeded, and sometimes 
 depressions stay quite stationary for appreciable periods. 
 
 85. The Sequence of Wind and Weather in a Depression. 
 It is clear that the sequence of wind and weather experienced 
 by a particular observer at a particular place as a depression 
 passes across him depends entirely upon how he is situated with 
 regard to the path of advance of the centre. 
 
 If he be actually on the track of the advancing centre, he 
 will notice that the barometer begins to fall, that high cirrus 
 cloud appears in the sky, and that the wind becomes light 
 southerly. If it be in winter, the temperature will be above 
 the seasonal normal. Soon the sky becomes overcast with cirro- 
 stratus and cirro-nebula, and halos are probably seen around 
 the sun or moon as the case may be. The wind, too, freshens. 
 As the centre approaches, the clouds lower to alto-cumulus 
 or alto-stratus, then to cumulo-stratus, and, finally, to nimbus 
 when rain begins to fall. As the centre passes the rain gets 
 heavier and the wind stronger. Next, the barometer begins to 
 rise after the actual passage of the trough line, and gusts or 
 squalls of wind are experienced. The wind changes to north- 
 west or north, and, often, blows more strongly than before. As 
 the centre moves away the rain ceases, or only occurs in showers, 
 the cloud shows signs of breaking, and long rolls of strato- 
 cumulus appear with patches of blue sky. Soon, the showers, 
 stop altogether, the barometer rises rapidly, the cloud gets higher, 
 some cirrus appears and the wind, whilst still remaining north- 
 westerly, moderates. Very soon, now, the depression entirely 
 passes out of range and blue sky, with light winds, prevails. 
 
 If the observer be situated to the south of the track of the 
 centre, instead of actually upon the track of the centre, he will 
 experience much the same sequence of cloud and rain as has 
 been described, but the winds will show a continual veering 
 Irom south-east to west as the centre passes north of him, and 
 to north-west as it passes away. If lie be to the north of the 
 centre's track, the wind will back from south-easterly at the 
 commencement, through east as the centre passes actually due 
 south of him, and, finally, to north-easterly and northerly as it 
 passes away. 
 
69 
 
 The time occupied in this cycle of changes clearly depends 
 upon the size of the depression and upon its rate of travel, but a 
 period of 24 hours is quite normal. 
 
 It should be noted that the winds on the north side of a 
 depression are, generally, less strong than those on the south 
 side so far as the northern hemisphere is concerned, it being 
 usually the case that the isobars on the north side are less 
 crowded than those on the south, A depression passing to the 
 south of the observer generally brings less wind, but more rain 
 and greater gloom, than does one passing to the north. 
 
 If high clouds occur, they are generally travelling outwards 
 from the centre of the depression. Thus a southerly wind on 
 the surface, with high cirrus moving from a westerly point, is 
 almost a certain sign of a depression situated to the westward 
 of the observer. 
 
 86. Highs or Anti-Cyclones. Anti-cyclones are the contrary 
 to depressions. Here, the maximum pressure occurs in the 
 centre of the closed isobars which are approximately oval or 
 circular in shape. An example of an anti-cyclone will be seen 
 in the lower left-hand corner of Fig. 10. The great distinction 
 between the winds and weather experienced in an anti-cyclone 
 to those experienced in a depression is that the latter are often 
 violent whilst those in the former are very rarely so. Moreover, 
 an ti- cyclones travel much slower and more irregularly than do 
 depressions, and, in fact, seem more stable, often remaining 
 more or less stationery for several days. It will be recognised 
 from an application of Buys Ballot's Law that the winds in an 
 anti-cyclone blow clockwise round the centre of high-pressure. 
 This should be compared with the wind-blow round a depression. 
 
 In summer, anti-cyclones generally bring fine, calm weather 
 during the day with much sunshine and a high temperature, 
 whilst at night the skies are clear but there is much ground 
 mist or fog. 
 
 In winter, however, there seem to be two distinct types of 
 anti-cyclones : 
 
 (a) those in which the weather is fine during the day and 
 
 very cold, with frost or fog, during the night. 
 
 (b) those in which the sky is persistently covered with low 
 
 cloud sheets, causing a dullness which is often 
 referred to as "anti-cyclonic gloom." 
 
 It used to be thought that the central regions of anti-cyclone 
 were regions of descending currents of air. In the face of the 
 evidence of the persistent low cloud, so characteristic of some 
 '"highs," this view cannot be regarded as a correct one. In fact, 
 it seems more true to regard anti-cyclones as comparatively 
 isolated and inert masses of air taking little part in the circulation- 
 going on about them ; and the fact emerges that almost any kind 
 of weather, except that of a violent nature may be got from them. 
 
70 
 
 It is a fallacy to regard anti-cyclones as, necessarily, areas of 
 fine, balmy weather, a fallacy, however, that seems to die hard. 
 It seems quite impossible to forecast whether any given anti- 
 cyclone, especially in winter, will be of the blue sky, or the 
 gloomy, type until an actual sample of the particular system 
 becomes available for examination. 
 
 87, Cols. A col is the region between two anti-cyclones. 
 An example is shown in Fig. 13. 
 
 An examination of the example given will reveal that a col 
 is the meeting place of two sets of winds of different directions ; 
 one set attached to one of the anti-cyclones, and the other set to 
 the other an ti -cyclone. These winds are all anti-cyclonic and 
 therefore light. They may, also, be of entirely different temper- 
 ature and relative humidity. Condensation by mixing is 
 therefore given full opportunity to manifest itself, and fogs 
 and mists are frequent. These fogs and mists are especially 
 marked in cold weather ; in warm weather, the mixture of the 
 air streams, with their different physical conditions, makes 
 the development of thunderstorms likely. 
 
 88. Secondary Depressions. Often, on a synoptic chart, a 
 big depression is seen to have one or more smaller depressions 
 on its periphery. Fig. 14 shows an example of a primary 
 depression with a secondary satellite. Sometimes, the 
 secondary grows whilst the primary diminishes, so that from 
 a study of one chart only, without reference to the preceding 
 three or four, it is occasionally difficult to say which is the 
 actual primary and which the secondary. This is the case in 
 the system shown in Fig. 14. 
 
 When the secondary depression is fully developed, that is. 
 has its own system of closed isobars, it carries its own wind and 
 weather systems with it. The wind and weather experienced 
 in secondaries are, then, the same in kind as occur in ordinary 
 depressions. 
 
 A secondary depression usually travels in an anti-clockwise 
 direction around its primary, and, generally, at a faster rate 
 than does the primary. This means that, as a rule, it pro- 
 gressively alters its relative position with regard to the centre 
 of the primary. 
 
 It is interesting to notice the wind behaviour in the region 
 between the secondary and the primary. In Fig. 14, taking the 
 centre over France as the secondary, and the centre over Scot- 
 land as the primary, it is obvious that, at some point between 
 the two depressions, the easterly winds due to the northern 
 part of the secondary must balance the westerly ones due to the 
 southern part of the primary, and that, in consequence, a calm 
 should result. This calm in the figure is seen to be in the 
 London area. Such a region of calms or of light variable airs 
 is a feature of the area between a primary and its secondary or 
 secondaries. 
 
to 
 
 I*! K 
 
 !^i! 
 
 * -i K<S 
 
 1*11; 
 
 L Jli ! if 
 
 1 5oe*in 
 
 i 
 
Y-SHAPED DEPRESSION, 
 
 DISTRIBUTION OF WEATHER, WIND, AND PRESSURE, 
 7 A.M. 8th OCTOBER, T915. 
 
 20V 
 
 ISOBARS are drawn for intervals of ten inilll- WEATHER. Shown by the following symbols 
 V>ars. . O clear sky. <T) sky i clouded. 
 
 WIN*- Direction is shown by arrow? flying )=( i n . a< i /m\ -tv a )mi<ii>ri 
 
 with the wind. QD y * cloude(I - 5iD " y * c 
 
 Force, on the scaie 0-12, U iudf- xtTX ovarc*at sky . rain failing 
 
 eated by the number of feathers. W 8now J hailT S 'og. 
 
 Calm 
 
 s=*oaist. T thunder. "K thunderstorm 
 
 Fig. 15. 
 
 7LM.S.O. Press, Kinjsway, W.C.2, 
 
71 
 
 In summer time, secondary depressions are apt to form 
 in comparatively large numbers. Being shallow, they do not 
 occasion much wind, but usually bring heavy rain and are often 
 the scene of thunderstorms. 
 
 Up to the present, only fully developed secondaries, those 
 with their own systems of closed isobars, have been considered. 
 Very frequently, however, the secondary depression is only 
 represented by a sinuosity in the isobars, the sinuosity taking the 
 form of a bulge, the bulge being outwards from the centre of 
 the main "low." These sinuosities travel forward in the same 
 direction as the main depression, sometimes with the same 
 speed, sometimes faster, and carry with them their own altera- 
 tions of wind and weather, which alterations are, often, very 
 marked. Such " bulges " are usually accompanied by much low 
 cloud and some rain. The important alterations produced by 
 even comparatively slight bulges call for great care in the 
 drawing of the isobars on synoptic charts. These bulges often 
 being otherwise unexpected gales on the side remote from the 
 primary depression. 
 
 89. V-shaped Depressions. V-shaped depressions may be 
 regarded as half-way between " bulges " and fully developed 
 secondaries. An example is shown in Fig. 15. 
 
 The point of the V in such a depression is generally directed 
 towards the south-east. The eastern side of the depression is 
 marked by southerly winds, much low cloud and rain, and the 
 western side by north-westerly winds, fair weather, and a lower 
 temperature. The change is often very abrupt and is accom- 
 panied by squalls. It is clear that a V-shaped depression 
 affords an example of warmer air, that from the south, being 
 forced to rise over colder and denser air, that from the north- 
 west. The southerly winds are usually of high relative 
 humidity so that the dynamical cooling caused by their forced 
 elevation is amply sufficient to produce the low cloud and the 
 rain which so characterise one side of the system. 
 
 90. Wedges of High Pressure. Between two depressions 
 there occurs, generally, what is known as a " wedge " of high 
 pressure. The wedge is usually shaped like an inverted V, and 
 it moves in an easterly direction with the two depressions which 
 it separates. As would be expected from what has been said 
 about depressions, the eastern, or front, side of a wedge is a 
 region of fair weather with light northerly or north-westerly 
 winds, and the western, or rear, side, a region of southerly 
 winds and increasing cloud as the second depression approaches. 
 In the centre of the wedge winds are very light and variable. 
 If fine weather quickly follow a depression, it is frequently the 
 sign of a wedge ; in which case, bad weather will follow again 
 comparatively quickly. 
 
 91. Straight Isobars. Quite often, the isobars on a synoptic 
 chart run almost straight over a large area. Such straight 
 
72 
 
 isobars are, usually, the outermost margins of a large depression 
 whose centre is located a considerable distance away. The 
 winds approximately follow the lie of the isobars and their 
 direction is determined by the position of the centre of low 
 pressure. Almost any type of weather may accompany straight 
 isobars, largely dependent upon the direction of the parallel 
 isobars, their relation to the depression centre, and the season of 
 the year. For example, in winter, in the British Isles, straight 
 north-south isobars bring northerly winds with squalls, and 
 sometimes snow or sleet ; but in summer, cloudy and thundery 
 conditions often prevail with an entirely similar isobaric 
 arrangement. 
 
 The commonest type of straight isobars over the British 
 Isles is the "south-westerly," which is the one in which the 
 isobars run practically parallel from south-west to north-east, 
 with the centre of the big depression, of which the isobars form 
 a part, somewhere in the Icelandic region. Straight isobars of 
 this type are very favourable to the formation of secondary 
 " bulges." The weather under these south-westerly conditions 
 varies considerably, from fair or fine to overcast with showers, 
 and, sometimes, steady rainfall is even experienced. 
 
 CHAPTER XY. 
 
 Special Phenomena. 
 
 92. Line Squalls. A line squall is a heavy squall of wind 
 which is accompanied by the passage of a long arch of low, 
 black cloud, which often stretches in approximately a straight 
 line for several miles, and from which heavy rain or hail falls 
 for a short time. Thunder, too, is often heard. The squall is 
 also accompanied by a veer of wind from some southerly or 
 south-westerly point to a westerly or north-westerly one accom- 
 panied by a sudden drop in temperature and a sudden rise of 
 the barometer. The actual squall of wind lasts for only a few 
 minutes, but is extremely violent. 
 
 The phenomena accompanying a line squall are quite 
 characteristic and cannot be mistaken. 
 
 The explanation of such a squall is to be found in the 
 sudden undercutting of a warm southerly air current of high 
 relative humidity by a colder, and hence, denser westerly or 
 north-westerly one. The warmer, southerly current is forced 
 to rise abruptly over the colder one. The ascent is accompanied 
 by the dynamical cooling which is an inevitable result of such 
 elevation. As the warmer current coming from the south is 
 generally of high relative humidity, very little ascent that is. 
 very little cooling is sufficient to produce low cloud and rain 
 As the undercutting takes place along a long length of line, the 
 
To face p. 73. 
 
 s* N C 4' 
 
 ^3^4 111 
 IM* IN 
 
 111 l>ll. 
 
 |!v il ; 
 4^ llfi 
 
 il^SJl' 
 
 lii ii*- 
 
 R^i^li 
 
 1540G 
 
73 
 
 long, black arch of cloud is explained. The violence of the 
 uprush is sufficient to explain the heavy rain, and also the hail 
 (see section 53), which is so often a feature of line squalls. 
 The sudden decrease of temperature is due to the fact that a 
 colder air stream has replaced a warmer one on the surface, and 
 this, too, explains the sudden rise in pressure. 
 
 The isobaric distribution which ends in the development of 
 line squalls will be revealed by a reference to Fig. 16, which 
 shows the synoptic charts for 7 a.m. and 1 p.m. on October 14, 
 1912, on which day a line squall crossed Great Britain. It will 
 be noticed in the 7 t a.m. chart that northerly or north-westerly 
 winds were blowing down the western coast of Scotland, 
 whilst south-westerly winds were blowing across Southern 
 Scotland. The line of meeting of these two air streams of 
 considerably different temperature was marked by a land squall 
 and all its attendant phenomena. The wind and pressure 
 " traces " taken at Aberdeen on the day in question are also 
 shown in Fig. 16, and should be carefully studied in the light 
 of what has already been said. 
 
 Line squalls are most frequent near the trough line on the 
 southern sides of depressions, and are especially liable to 
 happen if the depressions be at all V-shaped, for then, the two 
 wind streams, the one southerly and warm, the other, north- 
 westerly or westerly and colder, are brought more abruptly into 
 conjunction. 
 
 Line squalls constitute an exceptionally serious menace both 
 to shipping and to aircraft. Happily, however, once one appears 
 it is regular in its habits and travels steadily from west to east, 
 or from south-west to north-east, with a velocity of 20 to 30 
 miles per hour ; and hence, due warning of its coming can be 
 given to localities in their track. It is interesting to notice that 
 not one line squall has yet been known to travel from an easterly 
 point to a westerly one. 
 
 93. Thunderstorms. All the features of thunderstorms point 
 to their dependence upon a convectional overturning of the 
 atmosphere. They have been grouped* into three main 
 classes : 
 
 (a) Those due mainly to heated surface air in fine summer 
 
 weather. 
 (6) Those associated with powerful upper currents from the 
 
 south-west, the surface wind being light, variable or 
 
 south easterly, 
 (c) Those associated with low temperature in the upper air, 
 
 chiefly in the south-westerly currents of depressions. 
 
 * " Professional Notes No. 8. Temperature and Humidities in the Upper 
 Air conditions favourable for thunderstorm development, and Temperatures 
 over Land or Sea." By Capt. C. K. M. Douglas. Issued by the Meteoro- 
 logical Office, Air Ministry, London, 1920. 
 
74 
 
 This classification is, probably, not thoroughly comprehensive, 
 some thunderstorms occurring which seem to find no place in it, 
 but it is very useful. 
 
 Thunderstorms of type a occur chiefly in the warmer regions 
 of the earth during the warm seasons, and they are most violent 
 or active at that time of day, three or four o'clock in the 
 afternoon, when the convectional up currents, due to the great 
 heating of the land, are likely to have reached their maximum, 
 The clouds characteristic of these summer thunderstorms all 
 point to the importance of the convection currents engaged. 
 The cumulus clouds of an ordinary summer's morning are 
 seen to develop upwards. The development is continued and, 
 soon, some of them reach towering sizes, their summits 
 attaining heights up to 20,000 feet. If the summits of these 
 largely developed clouds be watched, a curious change is seen 
 to take place, the tops seeming to become flattened and then 
 covered with wisps or filaments of cirrus-looking cloud, termed 
 " false cirrus " which cover them as with fleecy mantles. 
 Investigation has shown that the " false cirrus " is made of 
 snow. 
 
 It is not every cloud present which goes through the full 
 development outlined. "When the full development is reached 
 the whole appearance of the " thunder cloud," as it is called, is 
 that of an anvil of great size which has been covered with a 
 cloth of exceptionally delicate texture. Generally, a thunder- 
 storm soon follows the appearance of an "anvil cloud." 
 
 It will be recognised that the lower the upper air tempera- 
 tures are, the better are the conditions for the development of 
 these summer thunderstorms. 
 
 Thunderstorms of type b are due to a definite flow of cooler 
 air over warmer, instead of to direct heating of the land 
 surface. Generally a warm surface southerly or south-easterly 
 current is overlain by a cooler south-westerly one. In such 
 cases the curve representing the fall of temperature upward is 
 very distorted and atmospheric instability is set up. Especially 
 does this seem to occur in the south-east quadrants of depres- 
 sions. There the surface wind is south-easterly or southerly, 
 and in this country warm and of comparatively high relative 
 humidity. It is known, too, that the upper winds in those 
 quadrants are mainly south-westerly or westerly, and of lower 
 temperature and relative humidity. It becomes clear, therefore, 
 that here opportunities for convectional overturning exist, over 
 and above that afforded by any local heating of the ground. 
 
 Thunderstorms of this type seem frequent and it may be 
 due to their prevalence that, as a matter of statistics, thunder- 
 storms occur more frequently in the south-east quadrants of 
 depressions than in any other isobaric grouping. It is worthy 
 of notice, too, that the conditions outlined may occur during the 
 night as well as during the day, and that thus some nocturnal 
 thunderstorms may doubtless be explained. 
 
75 
 
 Thunderstorms of type c are associated with the very low 
 upper air temperatures which are known to exist in the southern 
 part of a depression. If then the surface southerly or sonth- 
 easterly currents of the south-east part of the depression be 
 fairly warm, their displacement by the colder air from above 
 may cause conditions suitable for thunderstorms. Generally 
 thunderstorms of this type are not violent, and the conditions 
 suitable for their development may occur either during the 
 night or day. 
 
 In view of what has been said in the consideration of the three 
 types of thunderstorms, it is interesting to notice that in such a 
 region as Iceland most of the thunderstorms which occur are 
 nocturnal, and, moreover, characterise the winter months more 
 than the summer ones. In such high latitudes it would appear 
 then that the effect of local heating of land surfaces is seldom 
 sufficient to produce thunderstorms ; and that the storms there 
 are mainly produced as a result of the conditions outlined in 
 the consideration of thunderstorms of types 6 and c. 
 
 94. Cloud-bursts. Cloud-bursts are only exaggerated 
 thunderstorms. They occur chiefly in hilly or mountainous 
 districts, and manifest themselves as exceptionally heavy rain- 
 falls (sometimes with hail) in very short periods of time. They 
 are probably caused by the sudden stoppage of ascending air 
 currents as the hills or mountains are crossed. As was shown 
 in section 50, an ascending air current of sufficient velocity is 
 able to support a great quantity of accumulated water or hail 
 in it. If this upward current be suddenly stopped the water, 
 which it had been supporting, falls, and if the accumulated 
 stores were great the fall is extremely heavy and rapid. 
 
 95. Tornadoes. A tornado may be defined as a very violent 
 but short-lived wind. The connotation of the name varies some- 
 what in different parts of the world. In West Africa the 
 tornado is the squall which blows out of the front of a thunder- 
 storm at about the time the rain commences. Similar squalls 
 occur in all parts of the world associated with thunderstorms. 
 
 In the United States of America the name "tornado" is 
 applied to the small but very violent whirlwinds of one or two 
 hundred yards in diameter which advance towards the east or 
 north-east at a speed of from 20 to 40 miles per hour, the very 
 strong winds destroying practically everything within the direct 
 path. The activity of one tornado lasts for about half an hour, 
 and, apart from the violent winds, the most characteristic 
 feature of these tornadoes is the funnel-shaped cloud which 
 hangs downward and is generally in a whirling condition. 
 
 Tornadoes especially favour the valley of the Mississippi and 
 occur chiefly on hot afternoons in the summer months. They 
 occur most frequently in the south-east quadrants of depressions, 
 about 500 miles from the centre of lowest pressure. 
 
76 
 
 Waterspouts are the funnel-shaped tornado clouds occurring 
 at sea. They occur more frequently in the tropics than else- 
 where. The funnel cloud descends, and, as it descends, a cloud 
 of spray or vapour forms over the sea immediately beneath the 
 point of the funnel. Finally, the funnel point touches the 
 surface and the cloud of spray OY vapour takes on the appearance 
 of a column of water. Like the tornadoes on land, waterspouts 
 last only for about half an hour. 
 
 96. Hurricanes and Typhoons. --Tropical depressions are 
 often of small diameter, from 50 to 300 miles, with very steep 
 pressure gradients, and consequently with winds circulating 
 around them of great violence. Such systems occurring in the 
 Western Pacific, near Japan and the Philippine Islands, are 
 termed " typhoons " ; if near the West Indies, " hurricanes " ; 
 and, if in the Indian Ocean, " cyclones." The use of the term 
 " cyclones " in this connection needs to be carefully differentiated 
 from the usual use of that term in Great Britain (see section 79). 
 
 Hurricanes, typhoons, or " cyclones " are accompanied by 
 great masses of cloud, from which rain falls in torrents, and 
 very strong winds which, however, are not quite so violent as 
 those experienced in tornadoes. 
 
 The systems move along fairly regular routes in a mainly 
 westerly direction. If land be encountered the storms weaken 
 and often die away. 
 
 One peculiar feature of a hurricane or typhoon is the well 
 marked nature of the centre. As the hurricane travels the wind 
 rages with great violence, but, in the centre, drops to almost a 
 dead calm. This change occurs abruptly and the clouds 
 may even break and blue sky appear. The diameter of the 
 calm area may be from 10 to 30 miles. As the hurricane 
 continues to travel, however, and the centre passes over the 
 observer away to the eastward, the winds rise again as abruptly 
 as they dropped, attaining quickly to a violence equal to that 
 which they had before but blowing now from the reverse 
 direction, as would be expected from Buys Ballot's Law. The 
 centre, on account of its peculiar characteristics, is often 
 referred to as " the eye of the storm," and has been termed, 
 poetically, " the whirlwind's heart of peace." 
 
 97. Pamperos. South America, especially the republics of 
 Argentina and Uruguay, is often the scene of fierce thunder- 
 storms, which are termed " pamperos." Pamperos are Hne 
 squalls and show all the characteristics of those disturbances 
 (see section 92). The squalls brush up great clouds of dust 
 from the dry pampas and are accompanied by drenching rains 
 and almost incessant lightning for a short period. These storms 
 are greatly feared by sailors in the estuary of the Rio de la 
 Plata. 
 
CHAPTER XVI. 
 
 Forecasting. 
 
 98. Forecasts. The first use of synoptic charts for forecasting 
 weather was made by Admiral Fitzroy in 1860, and it was he 
 who invented the special meaning of the term " forecast " to 
 avoid the doubtful connotations attaching to such terms as 
 " prognostic " and " prophecy." 
 
 As issued now, a forecast generally covers a period which is 
 not greater than 24 hours, and, usually, a statement under the 
 heading of "further outlook" is added, giving the conditions 
 likely to be experienced in the 24 hours or so following the 
 period covered by the actual forecast. 
 
 A forecast is published, as a rule, in a definite form. It 
 includes : 
 
 (a) A statement of the wind speed and direction at a height 
 of 2,000 feet, together with the changes therein likely 
 to occur during the period of the forecast. 
 
 (fr) A similar statement with regard to the wind at the 
 surface. 
 
 (c) A statement concerning the probable state of the sky 
 
 face with regard to cloud, together with statements 
 concerning the incidence of precipitation (if any) and 
 concerning the temperature likely to be experienced, 
 especially as to whether that temperature is likely to 
 be greater or less than the normal for the time of 
 year. Notes are also added, if necessary, as to the 
 probability of thunderstorms, or frosts, or any other 
 special phenomena. 
 
 (d) A statement concerning the visibility likely to be 
 
 experienced together with notes, if necessary, con- 
 cerning the probability of the occurrence or clearance 
 of mists and fogs. 
 
 The " Further Outlook " is then appended in one or two 
 short but comprehensive sentences. 
 
 99. The Practice of Forecasting. In the making of such a 
 forecast as has been described, the forecaster depends prac- 
 tically entirely upon the synoptic charts before him, together 
 with special experimental data received from certain stations 
 concerning upper air winds and temperatures. He examines 
 the current chart and determines what types of isobaric distri- 
 bution (see section 78) are present. From a further considera- 
 tion of the chart, in conjunction with the two or three previous 
 charts, he then determines the probable travel of the various 
 isobaric groupings present, and tries to visualise what the 
 synoptic charts of 6 and 12 hours hence will be like. Knowing 
 the general sequences of wind and weather connected with the 
 
78 
 
 various types of isobaric distribution, he is then able to make 
 up his mind as to how far the particular regions for which he is 
 forecasting will be affected by the various sequences in their 
 travel. 
 
 The weather and wind connected with the various pressure 
 types, together with notes upon the travel of those types, will be 
 found in Chapter XIV. 
 
 To determine the probable visibility, and the probability of 
 the occurrence of such phenomena as hail, thunderstorms, frosts, 
 and the like, the forecaster has to bear continually in mind the 
 physical principles underlying those phenomena, which prin- 
 ciples have been considered at more or less length in previous 
 sections of this book. 
 
 To determine what the " further outlook " is likely to be, he 
 has to visualise what the synoptic chart is likely to be 24 hours 
 hence, and thence to apply the same plan that has already been 
 outlined. 
 
 It will be recognised, then, that the making of a sound 
 weather forecast is by no means an easy matter, but connotes 
 that the forecaster shall have at his ready disposal a well 
 organised stock of meteorological and physical knowledge, 
 which knowledge can only be obtained by diligent study, and 
 the ability to use which can only be gained as the result of 
 much practice. The forecaster is much in the position of a 
 physician ; to be successful he must be able to diagnose rapidly 
 but soundly, must be able quickly to determine which factors 
 are essential and which non-essential, and then having gained 
 the factors by his diagnosis, his analysis must be able to 
 predict the future action of those factors, and to interpret that 
 action in the light of his particular needs. 
 
 Forecasting weather is a scientific process based upon 
 physics and soundly sifted experience ; it cannot be success- 
 fully accomplished either by reading the legends around a 
 " weather glass " or by an appeal to the rhymes and maxims of 
 weather lore, no matter how extensive may be the stock of that 
 weather lore held. 
 
 100. Forecasting by a Single Observer. It is interesting to 
 attempt to realise what a single observer, without any synoptic 
 charts or without any knowledge of what weather conditions 
 are in other localities can do to forecast coming weather. 
 
 Remembering Buys Ballot's Law, and what has been said 
 about the wind and weather sequences in a depression (see 
 section 85), and also by watching his barometer, he should be 
 able generally to recognise the approach of a depression, and 
 also to determine how he is situated with regard to the path of 
 travel of its centre. Once knowing this, he is able roughly to 
 forecast coming weather events in the next few hours, though 
 he will be very considerably handicapped, in the absence of a 
 
79 
 
 current chart, by not knowing the size of the depression with 
 which he is dealing. 
 
 He may often, too, be able to tell whether fine weather is the 
 result of a wedge or of an anti-cyclone, and a study of the 
 principles underlying the formation of fog (see sections 36-40), 
 and of frost (see section 34), may enable him quite often 
 successfully to predict whether or no those particular phe- 
 nomena are to be expected. A careful study of cloud forms, 
 especially of alto-cumulus-castellatus and of " anvil cloud," 
 may enable him, too, to be successful in forecasting some 
 thunderstorms. 
 
 But when all is said, the best that a single observer can do 
 to forecast, for even a short period, is very little, and " further 
 outlooks " are quite impossible for him. In short, synoptic 
 charts are the very foundation of successful forecasting, and, 
 without them, very little is possible. 
 
 101. Long Range Forecasts. Much work has been done in 
 the endeavour to find means to forecast the coining of weather 
 changes at a greater distance ahead than three or four days. 
 It cannot be said that, up to the present, very much success has 
 resulted. 
 
 If the problem be put in the shape of asking what the 
 weather in such a place as London, for example, will be this 
 day week, then the only answer is that the succession of 
 pressure types is, apparently, so irregular that the weather at 
 such a distance ahead cannot be forecasted beyond giving the 
 average for the time of year. There are almost certainly laws 
 underlying the apparently irregular succession of isobaric 
 groupings, but they have not yet been found, though probably 
 they will be in course of time. Three or four days ahead, in 
 such a region as the British Isles, seems the extreme distance 
 for which a forecaster can predict satisfactorily, even under the 
 most favourable conditions. 
 
 But certain results have been reached with regard to a 
 periodicity in the sequence of certain meteorological elements, 
 results which are sufficiently encouraging to lead to hopes that 
 much more may yet be done. To take an example, Bruckner, 
 of Berne, so far back as 1890, concluded, from an examination 
 of all available rainfall data, that, in Europe, there is a variation 
 of rainfall with an average period of 35 years, and hence, was 
 able roughly to predict wet periods and dry periods, each 
 approximately of 17 years' duration. Russell has suggested, 
 similarly, a period of 19 years for the variation in the rainfall 
 of Australia, a result which has been confirmed by other 
 workers. 
 
 Much, too, has been attempted in the correlation of sun- 
 spots with terrestrial weather. The mean period of frequency 
 of sun-spot numbers is, approximately, 11 years. It is known 
 that sun-spots are due to vertical disturbances in the sun's 
 atmosphere, such disturbances being of huge extent. It is 
 
80 
 
 known, too, that the spots bear a definite relation to terres- 
 trial magnetic records. Many weather phenomena such as 
 " cyclones " in the Indian Ocean (see section 96) and the 
 rainfall in Scotland have been endeavoured to be correlated 
 with the periods of the spots, but the results have not been 
 very satisfactory. 
 
 Another method of employing meteorological statistics is to 
 endeavour to prove a connection between one set of phenomena 
 occurring in one part of the wbrld and another set of different 
 phenomena occurring in another part at a later date, thus 
 placing the two in the form of cause and effect. As all weather 
 is, doubtless, the result of a general atmospheric circulation 
 over the whole globe, little objection on the logical side can be 
 urged against these attempts, and some have been very 
 successful. Sir Napier Shaw* has shown, for example, that 
 a remarkably close parallelism exists between the seasonal 
 variation in the trade wind at St. Helena and the rainfall in the 
 south of England ; and several other workers like Meinardus and 
 Hildebrandsson have gained similarly noteworthy results. 
 
 But there is a danger to be avoided and this section cannot 
 be better concluded than by a quotation which reveals the limita- 
 tions of these periods as applied to actual forecasting for a 
 given stretch of time or for a given date. The quotation! is as 
 follows : " There is at once a fascination and a curious 
 inconclusiveness about many of these attempts to identify the 
 period of variation of the meteorological elements. The period- 
 icity shows itself when records for long intervals are studied, 
 but it is apt to be illusory as a guide to the meteorological 
 character of any particular year. An example of the fascination 
 and the difficulty here referred to may be given from the 
 consideration of the seasonal variation of rainfall in our own 
 country. Nothing is more certain about the seasonal variation 
 of that element than that the rainfall tends to a maximum in 
 October and a minimum in March ; any curve of average monthly 
 values taken for a long period of years with give that result, and, 
 consequently, one is perfectly justified in assuming that the 
 assignment of a seasonal periodicity in rainfall with a maximum 
 in October represents something real, yet, if one were to hazard 
 a prediction that the coming month of October will be the 
 rainiest month in the current year, the facts might belie the 
 prediction." 
 
 * See "Nature," Vol. 73, p. 175. 
 
 f " Forecasting Weather." Sir TV. Napier Shaw, F.R.S. (Constable and 
 Co., Ltd.) 1913. Chapter XYI1. p. 357. 
 
81 
 
 CHAPTER XVII. 
 
 Weather Lore. 
 
 102. Weather Maxims. Ever since man has thought at all, 
 he has generalised about the weather with the result that a very 
 large number of weather maxims are now in existence, these 
 maxims being intended as means whereby the weather to come 
 may be forecasted.'" 
 
 A few of these maxims are true, a larger number contain 
 a partial truth, a very large number indeed, are completely 
 fallacious, and some are mutually contradictory. But this 
 does not seem to matter at all ; these maxims are part of 
 the popular current coinage of conversation, and often enough, 
 definite action is based upon them, which definite action 
 leads frequently to a failure which need not occur if the user 
 of the weather maxim but take the trouble to examine the founda- 
 tions of the belief in which he puts his trust. 
 
 Very few of the common weather sayings stand the test of 
 statistical examination, and this is hardly a matter for surprise 
 when the propensity of man to count the "hits" and to neglect 
 the " misses " be remembered. 
 
 Consider the prediction, for example, associated with St. 
 Swithin's Day (July 15th) in England : 
 
 " If St. Swithin greets, the proverb says, 
 The weather will be foul for forty days." 
 
 Even the most cursory investigation of this over a few years 
 demonstrates its unreliability. There is a further point about 
 this saying, however, which is interesting and that is, that other 
 nations close at hand have similar predictions ' connected with 
 other saints and other days. In France the day of augury is 
 that of St. Medard (June 8th), in Belgium that of St. Godelieve 
 (July 27th), and in Germany that of the Seven Sleepers 
 (June 27th). It is clear, therefore, that giving to each of these 
 predictions equal value, in quite a circumscribed portion of 
 Western Europe it would be foul weather continuously for a very 
 long period, a doctrine to which even the most inveterate 
 grumbler concerning the weather could hardly truthfully 
 subscribe. 
 
 Sayings concerned with the weather likely during a given 
 month are often equally unreliable. For example, the second of 
 the months of the year is frequently referred to as " February 
 Fill-dyke," this being, presumably, a statement of belief in it 
 as a time of much rain ; but, as a matter of statistics spread 
 
 * The interested student is referred to " Weather Lore, A Collection 
 of Proverbs, Sayings, and Rules concerning the Weather." By R. Inwards. 
 3rd edition, 1898. 
 
 J5406 F 
 
82 
 
 over a long period, February in the British Isles is one of the 
 driest months in the year. An examination, too, of actual 
 months of March over long periods reveals the fact that no 
 reliance whatever can be placed on the couplet 
 
 " March, black ram, 
 Conies in like a lion and goes out like a lamb." 
 
 Many sayings, too, are in existence that endeavour to show a 
 connection between the weather in one period of the year and 
 that in another, such as that which declares that if the end of 
 October and the beginning of November be wet, then the 
 January and February following will be frosty and cold, except 
 after a very dry summer, but statistical investigation has shown 
 clearly that such sayings have very little in them. 
 
 There is a great number of sayings existing which endeavour 
 to relate the behaviour of wild and domestic animals to weather 
 and its changes. Some of these sayings depend simply on the 
 response of the animals in question to the weather actually 
 existing at the moment, and, as such and limited to such, some 
 reliance may be placed upon them, especially is this so in the 
 case of considerable changes in humidity as both men and other 
 animals are affected by the change from dry to wet air, and 
 vice versa. But this is an entirely different matter from 
 attributing to animals a foresight, in the case of coming 
 weather, denied to man. Such a foresight has never been 
 proved. To take an example, investigation has shown that no 
 reliance at all can be placed upon the supposed instinctive 
 foresight of such animals as beavers and squirrels in preparation 
 for severe winters. 
 
 Another very popular belief is that the phases of the moon 
 are able to exert an influence upon terrestrial weather, but this, 
 too, has never been demonstrated as having an objective reality 
 to the satisfaction of meteorologists. 
 
 In some of the sailor's weather maxims, however, more truth 
 Can be found, for these can often be related to the types of 
 pressure distribution appearing on synoptic charts. For 
 example, the saying 
 
 " First rise after low 
 Foretells a stronger blow," 
 
 is quite true of the squalls and general increase in wind velocity 
 which are often experienced as the trough line (see section 80) 
 of a depression passes. Similarly, the couplet 
 " Long foretold, long last ; 
 Short notice, soon past," 
 
 is often true, inasmuch as the larger depressions with their 
 larger areas of bad weather travel more slowly, as a rule, than 
 do the smaller depressions ; and. therefore, their warning signs 
 of high cloud and backing winds are in evidence for a longer 
 time. 
 
83 / 
 
 In the land saying 
 
 " Rain before seven, 
 Fine before eleven," 
 
 there is sometimes truth as it is usually connected with the 
 passage of a small secondary depression. Similarly, the fact 
 that depressions often follow one another quite closely lends 
 point to the saying that 
 
 " A Nor- wester is not long in debt to a Sou-wester." 
 
 But it cannot be too strongly emphasised that even those 
 sayings that can be connected with synoptic charts are simply 
 general statements to which many exceptions occur. Halos, for 
 example, around the sun or moon are generally regarded as signs 
 of coming bad weather, and as they are a concomitant of cirro- 
 stratus and cirro-nebula clouds, which clouds are a feature of 
 the fronts of depressions (see sections 81 and 85), the belief has 
 a good deal of truth in it ; but high clouds with halos sometimes 
 occur in the rear of depressions when they cannot be considered 
 as a sign of bad weather but of quite the reverse. 
 
 Many sayings relate to sky colorations in their relation to 
 weather. There is, for example, the very well known one 
 
 " A red sky at night 
 Is the shepherd's delight ; 
 A red sky at morning 
 Is the shepherd's warning," 
 
 which often gives reliable prognostications but, on the other 
 hand, quite often proves " a broken reed." 
 
 Rainbows deserve some reference in this connection. By 
 reason of the physics of their formation, the centres of rainbows 
 are always exactly opposite the sun. In regions therefore 
 where areas of rain generally travel from westward to eastward, 
 as in the British Isles, rainbows give some indications of 
 coming or departing rain ; for rainbows in the east, and hence 
 in the afternoon, denote clearing weather, the rain forming the 
 bow having passed the observer, but if seen in the west, and 
 therefore in the morning, rain is approaching. 
 
 In short, no weather maxim should be accepted, no matter 
 how popular or how often quoted it may be, unless it has been 
 subjected to statistical investigation, including, if possible, a 
 comparison with many synoptic charts, over a long period and 
 has not been found wanting. Very few maxims can survive 
 such a searching test, as has been already said, and must be 
 regarded as unworthy of modern acceptance. It is a complete 
 fallacy to believe, as so many people do, that just because a thing 
 is often said that there is, therefore, something of truth in it. 
 
 103. The " Weather Glass "Iu the period between the 
 invention of the barometer (1643) and the introduction of the 
 synoptic chart (about 1860), the weather maxims handed down 
 from antiquity were reinforced by personal observations of the 
 barometric pressure. It is no matter for surprise, therefore, that 
 
 F 2 
 
\ 84 
 
 the period in question was characterised by most persistent 
 attempts to interpret barometer readings in terms of coming 
 weather. The " weather glass " was one of the results : and, as 
 used now, usually consists of a mercury or aneroid barometer, 
 the height of which is indicated by a hand moving over a dial 
 inscribed with legends which run as follows : Very Dry, Set 
 Fair, Fair, Change, Rain, Much Rain and Stormy. 
 
 These legends are obviously not very satisfactory because 
 they are written against particular readings of the barometer 
 irrespective of locality or the height of the instrument above 
 sea level. Moreover, if synoptic charts impress one fact upon 
 the student of them more than another, it is that it is the 
 relative distribution of pressure which is the all-important 
 matter, and not the actual barometer reading at any one place. 
 Nevertheless, the legends and their positions are based upon 
 experience and, hence, are not without value ; but they alone 
 are not .sufficient to give sound forecasts in the absence of 
 synoptic charts. 
 
 An extension of the legends on the " weather glass" is to be 
 found in the comprehensive list of instructions* for the use of the 
 barometer to foretell weather, issued by Admiral Fitzroy in 
 about the year 1860. Once again these rules are based upon 
 wide experience, and, consequently, give often enough sound 
 results. The coming of synoptic charts has, however, thrown 
 an entirely new light upon weather and its changes, and 
 Fitzroy's rules can now only be recommended for use in those 
 circumstances when it is impossible for the inquirer either to 
 draw and interpret a synoptic chart for himself, or to get, by 
 wireless telegraphy or otherwise, a forecast based upon such 
 charts. 
 
 PART HI. THE UPPER AIR. 
 
 CHAPTER XVIII. 
 
 The Variation of Wind with Altitude. 
 
 104. A Note on Units. Any discussion on what may be 
 termed " aerology ", the study of the upper air, introduces 
 afresh in an acute form the question of units. A previous 
 reference bearing on this vexed question in present-day 
 meteorology has been made in section 69. 
 
 The subject of units is more acute in the study of the upper 
 air for various reasons. In the first place, the study is a 
 comparatively new one and practically all the work in it has 
 been done, and the results published, in C.G.S. units (see 
 section 69) ; heights being measured in metres or in kilometres, 
 
 * Reprinted in "The Weather of the British Coasts," issued by the 
 Meteorological Office, London. 
 
85 
 
 velocities in metres per second, -and temperatures in degrees 
 Centigrade (or, more generally, marking another diversity, in 
 degrees Absolute). This means that any writer, writing for 
 English readers and, hence, desirous of using the units familiar 
 to those readers, that is to say, putting heights in feet, velocities 
 in miles per hour, and temperatures in degrees Fahrenheit, has 
 to change the data of the various original publications into the 
 English units. The transformations obviously lead to awkward 
 numbers as the scales on the one side cannot be expressed in 
 simple round numbers of the scales on the other. 
 
 In the second place, the discussion of such a matter as 
 
 upper air temperatures on the Fahrenheit scale involves the 
 
 use of negative signs, and, as has been well said,* "a scale of 
 
 ' temperature including positive and negative signs is dan- 
 
 ' gerous for the observer, troublesome for the computer, 
 
 ' awkward for the printer, and misleading for the reader." 
 
 The use of tlje Centigrade scale involves the same trouble, 
 
 which, however, is avoided if the Absolute! scale be used. 
 
 The fact must, however, be faced that the two previous parts 
 of this book have been written using the English units, as 
 those units are at present very much more familiar to the 
 readers for whom it is mainly intended. That being the case, 
 a change of units in the present part seems unjustified as 
 involving a serious break in continuity, and as likely to lead to 
 confusion. Nevertheless, it is extremely desirable that the 
 student should spare no effort to make himself familiar with the 
 C.G.S. system, inasmuch as that is the system in which most 
 research work in the study of the upper air is being published, 
 and as it will ultimately probably come into general use in 
 meteorological science. 
 
 A compromise, then, seems the best way. To secure 
 continuity the data and tables appearing in the discussion of 
 upper air conditions which follows will be expressed in English 
 units, but wherever possible the C.G.S. units will be given as 
 well. 
 
 It should be noted that, in transformation from one system 
 to another, such matters as a certain number of kilometres will 
 be expressed in an approximate round number of feet, and a 
 certain number of degrees Absolute as a round number of 
 degrees Fahrenheit, decimal points being omitted except where 
 the accuracy denoted by them seems important. 
 
 105. Wind near the Surface. The Meteorological Office, 
 London, have issued J an interesting table (Table VI.) which 
 
 * " Observer's Handbook," Meteorological Office, London, 1919 edition. 
 
 f " The measure of temperature is so chosen that the volume of a mass 
 of gas at constant pressure, or the pressure of a mass of gas at constant 
 volume, is proportional to the temperature. It is the temperature on the 
 Centigrade scale increased by 273." ("Observer's Handbook," Meteoro- 
 logical Office, London, 1919 edition.) 
 
 J Annual Summary of the Monthly Report, 1916. 
 
86 
 
 endeavours to connect the velocity of the wind at heights up to 
 100 feet (>0 metres) above open grass land with that occurring 
 at 33 feet (10 metres). 
 
 TABLE VI. 
 
 Wind at Various Heights above Open Grass Land compared 
 with Wind at 33 feet (10 metres). 
 
 Height in 
 
 
 
 
 
 
 
 
 
 
 
 Metres - - | 0'5 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 
 
 Feet (approx.) j 1 5 i 3 
 
 6-5 
 
 10 
 
 13 
 
 16 
 
 33 
 
 50 
 
 66 
 
 82 
 
 100 
 
 Katio to wind at 
 
 
 
 
 
 
 
 
 
 
 
 
 33 feet (10 
 
 50 
 
 59 
 
 73 
 
 80 
 
 85 
 
 89 
 
 1-00 
 
 1-07 
 
 1-13 
 
 1-17 
 
 1-20 
 
 metres). 
 
 
 
 
 
 
 
 
 
 
 
 
 A note appended to the table is also interesting. The note 
 reads: "It is to be understood that the ratios shown above 
 ' are only approximate. The increase of wind with height is 
 ' more rapid in proportion when the air is not disturbed by 
 ' convection, i.e., in cold weather and at night ; it is less rapid 
 " under the opposite conditions." 
 
 The note is valuable because it gets at once to the difficulty 
 that is involved in any discussion of winds near the surface. 
 The air near the surface is in a continual state of eddying, the 
 eddies being due to irregularities in the surface of the ground 
 over which the air is moving, and also to changes in the 
 temperature of that surface, and consequently of the air in its 
 proximity. Reference has already been made to these eddies 
 in the discussion of the gustiness shown by anemometer traces 
 (see sections 7 and 9). * The effect of these eddies is to promote 
 the mixing of adjacent masses of air and, in general, the lower 
 masses gain in momentum at the expense of the higher ones. 
 If eddying be reduced through any cause, the lower layers 
 become checked by friction with the surface of the land, whilst 
 the upper layers gain by not being subjected to the loss of 
 momentum due to mixture with the lower ones, which are less 
 energetic. Taylor has developed the theory of eddy motion in 
 the atmosphere in two important papers,* to which the student 
 is referred. 
 
 In view of what has been said concerning the varying 
 turbulence of the air layers near the ground, it would appear 
 that it is a very difficult matter to find any simple relationship 
 existing between the wind in one layer near the surface and 
 another. 
 
 Attempts have been made, however, to find some simple 
 formula which shall accomplish the desired end. 
 
 * G. J, Taylor, " Phenomena connected with Turbulence in the Lower 
 Atmosphere." Proceedings, Royal Society, A., Yol. XCIY., p. 147, 1918. 
 " On Eddy Motion in the Atmosphere," Phil. Trans., A., Yol. CCXY., 
 p. 1. 
 
87 
 
 In 1880, Stevenson* gave a formula for approximate wind 
 velocity up to 51 feet. His formula was 
 
 /H + 72. 
 
 where V = the wind^ velocity at height H, which is the 
 velocity to be found, 
 
 and v = the known wind velocity at height h. 
 H and h are measured in feet. 
 
 Later, in 1885, Archibald,! as a result of observations made 
 with kites, gave a formula for wind velocities between 300 feet 
 and 2,000 feet. It will be noticed that he avoided the layer in 
 immediate proximity to the surface. His formula was 
 
 v= /Hxt 
 
 where V, v, H and h, have the same meanings as in Stevenson's 
 formula. 
 
 Sir Napier Shaw and Captain Cave, in 1910, as a result of 
 observations on small, hydrogen-filled balloons, put forward the 
 
 formula!]: 
 
 V = H + a V , 
 a 
 
 where V = wind velocity at height H, 
 
 V = wind velocity given by a well-exposed anemo- 
 
 meter, 
 and a = a numerical constant for the particular site. 
 
 The observations were carried out at Ditcham Park, in 
 Hampshire, and the value of the constant a there was 550 feet 
 (167 metres^, the height of the station above sea-level. 
 
 None of the formulae mentioned, however, have been found 
 to be entirely satisfactory when compared with actual results. 
 
 Some later attempts, too, merit mention. Chapman, for 
 example, has put forward an empirical rule that the speed of 
 the wind in the lower layers is a linear function of the logarithm 
 of the height. His formula takes the form 
 
 where V is the wind velocity required at height H, 
 and a and 6 are constants which not only vary from place 
 to place, but also for different sets of observations taken 
 at different times of the day at the same place. 
 
 * Journal, Royal Scottish Meteorological Society, 1880, p. 348. 
 
 t " Nature," Vol. XXVII., p. 243. 
 
 J First Report, Advisory Committee for Aeronautics, Reports and 
 Memoranda, No. 9. 
 
 E. H. Chapman, Professional Notes, No. 6. Meteorological Office, 
 London, 1919. 
 
88 
 Hellman,* in 1917, gave a very similar formula 
 
 where a, ?>, and c are constants, and H and V have the same 
 meaning as in Chapman's formula. 
 
 106. Extent of Surface Turbulence upwards. It is pertinent, 
 at this point, to inquire as to how far upwards the turbulence 
 attributable to eddy motion extends. In this connection the 
 results obtained from the anemometer placed on the top of the 
 Eiffel Tower in Paris are of great interest. The head of that 
 anemometer is, roughly, at a height of 1,000 feet (more exactly, 
 305 metres) above the ground and its records reveal the fact 
 that the disturbing effects of the surface certainly extend to 
 that height. It would be expected, too, that the effect of 
 turbulence should be felt at greater heights in summer than in 
 winter, as it is in summer that the temperature cause forming 
 eddies is given its fullest opportunities. The term " surface 
 layers," then meaning the layers affected by surface turbulence, 
 is usually taken to mean the air up to 6,000 feet in winter and 
 up to 9,000 feet in summer. 
 
 It must be recognised, however, that even above the level 
 where surface turbulence ceases to affect matters, the atmosphere 
 still remains complex, and that the behaviour of the wind there 
 cannot be expressed in simple statements. 
 
 107. Methods of Measuring Upper Winds. The measure- 
 ment of winds in the upper air, by means other than by that of 
 determination of cloud velocities, is of comparatively recent 
 growth. The measurements are usually made either by following 
 the motions of small hydrogen-filled balloons,! or by observing 
 the smoke drift from shells bursting at known heights.^ Kites, 
 too, have been used, but are hardly to be compared with the other 
 two methods. 
 
 It is clear, however, that both the small balloon and the 
 shell-burst methods are only possible in those conditions when 
 low cloud is more or less absent. A method has been devised 
 whereby the wind at a given height even above cloud sheets can 
 be calculated from the bursting of bombs at that height ; but 
 not very much work has been done on it as yet. It will be 
 recognised, therefore, that the means available for determining 
 the wind at heights are limited, and are available only in clear- 
 sky (that is, clear of low cloud) weather. This limitation needs 
 to be borne in mind in what follows. 
 
 108. Wind Velocity between the Surface and 8,000 feet. 
 This layer includes the surface layers and gives, also, a little 
 information about what is just beyond the height where the 
 surface turbulence is, at times, eliminated (see section J06). 
 
 * Preuss, Akad. der Wiss., Berlin, 10, pp. 174-197, 1917. 
 
 f For details of the methods employed, see either "The Structure of 
 the Atmosphere in Clear Weather " by C. J. P. Cave (Cambridge University 
 Press), 1912 ; or " The Computer's Handbook," Section II., Subsection I., 
 Metereological Office, London. 
 
 % " The Computer's Handbook," Section II., Subsection I., Meteorological 
 Office, London. 
 
89 
 
 It is best treated by reference to a special set of observations, 
 and the set chosen is that performed by Dobson,* at Upavon, 
 using the small balloon method with one theodolite. The 
 observation ground was 600 feet (183 metres) above sea-level, 
 and was characterised by an exceedingly open exposure. 
 
 Dobson arranged his 97 ascents with regard to the wind 
 direction at the surface. With north-easterly surface winds, 
 the gradient wind velocity (see section 77) as calculated from 
 the current synoptic charts was reached at a height of about 
 3,000 feet (915 metres), and then the velocity began to 
 diminish : first, an increase up to 3,000 feet, and then a falling 
 off. With south-easterly surface winds the velocity increased 
 to reach the gradient wind speed at just under 1,000 feet 
 (approximately 300 metres), and then showed little deviation 
 from that value with increasing height. With south-westerly 
 surface winds, the velocity increased to reach the gradient 
 wind speed at about 1,500 feet (approximately 500 metres), 
 and thenceforward continued to increase slightly. With north- 
 westerly surface winds, the gradient wind was reached at just 
 under 1,000 feet (approximately 300 metres), and, thereafter, 
 the wind showed a fairly regular, considerable increase up to 
 the limit of Dobson's observations, namely, 8,000 feet (approxi- 
 mately 2,500 metres). 
 
 The general result, then, to be obtained from this work is 
 that mean wind velocity increases from the surface up to about 
 1,000 or 1,200 feet (roughly 300 to 400 metres), and that, 
 thenceforward, the behaviour is more eccentric, depending to 
 some extent upon the surface direction, and, as he showed in 
 another analysis of the same observations, also upon the 
 strength of the wind. The curves which he drew to illustrate 
 his conclusions show very clearly that the increase of wind 
 speed between the surface and the 1,000 or 1,200 feet level is 
 not linear. This statement should be compared with the 
 various formulas given in section 105. 
 
 It is interesting to notice that similar work done at coast 
 stations reveals that the wind behaviour between the surface 
 and 8,000 feet (roughly 2,500 metres) is even more cornpli; 
 cated there than that shown by Dobson's results at Upavon, an 
 inland station. This would be expected as a consequence of 
 the juxta-position of land and sea ; the land on the one side 
 with its atmosphere filled with eddies and general turbulence, 
 and the sea on the other with its atmosphere, to a great extent, 
 free from turbulence. 
 
 109. Wind Direction between the Surface and 8,00 feet. 
 So far as general changes of direction are concerned in this 
 layer, the most marked feature is the "veer" (that is, change 
 of wind in a clockwise direction) from the surface direction. 
 This veer is particularly marked in the first 2,000 feet. The 
 
 *G. M. B. Dobson, "Pilot Balloon Ascents at the Central Flying 
 School, Upavon, during the year 1913." Quarterly Journal, Royal 
 Meteorological Society, Vol. XL., p. 123. 1914. 
 
90 
 
 surface wind deviates from the lie of the isobar at the surface, 
 as has been previously stated (see section 75), and the veer 
 brings it more or less into parallelism with that isobaric lie at 
 the height of about 1,500 to 2,000 feet (approximately 500 to 
 650 metres). The veer is then carried on slightly beyond the 
 isobaric lie at heights between 2,000 feet (650 metres approxi- 
 mately) and 8,000 feet (2,500 metres approximately). The one 
 curious exception appears to be the case of north-easterly winds 
 at the surface. These winds veer as do the winds in the other 
 three quadrants, but do not quite reach the lie of the surface 
 isobars even at heights well above 2,000 feet (650 metres). 
 
 110. Some Abnormalities. In general, as has been said, the 
 wind shows an increase in velocity from the surface up to, say, 
 2,000 feet. Not infrequently, however, it is found to fall off 
 from the surface upwards, or from some point near the surface. 
 This is exceedingly likely to happen when there is no very 
 definite pressure gradient on the surface so that katabatic or 
 anabatic winds are likely to prevail there (see section 17). Such 
 winds have no relation to the pressure gradient, and in such 
 cases the wind very near the surface may have a velocity which 
 is much greater than that prevailing, say, at 500 feet and above. 
 Quite often, too, the wind at quite a low height in such cases 
 is found to be entirely reversed in direction from what it is 
 actually on the surface ; this, again, is to be expected from the 
 lack of relationship between katabatic and anabatic winds and 
 the pressure distribution. 
 
 It is interesting to notice, too, that when the wind at the 
 surface is easterly, there is often a tendency for the wind 
 velocity to decrease with height above it. Such easterly 
 currents, too, are often of comparatively little thickness, so that 
 it is quite common to find great changes of direction at no 
 great height above them. 
 
 Moreover, it needs to be emphasised that to any general 
 statement made about wind behaviour at heights above the 
 surface numerous exceptions exist. 
 
 111. Wind between 8,000 and 25,000 feet. The first point 
 to be stressed under this heading is that very little help need be 
 anticipated from the pressure distribution at the surface as a 
 means of predicting what the wind direction or force at 8,000 
 feet (approximately 2,500 metres) or over is likely to be. 
 
 It needs to be remembered, too, in what follows, that owing 
 to the fact that actual experimental observations, by following 
 small balloons from the ground or by watching shell-bursts 
 similarly, are only possible in the absence of much low cloud, 
 that the results obtained for this particular layer are those for 
 practically clear sky weather. 
 
 Captain Cave* has divided his observations of winds up to 
 25,000 feet into five groups : 
 
 (a) Those which show " solid currents " ; that is, those on 
 which, after the gradient velocity has been reached, 
 
91 
 
 there is little change either in direction or velocity 
 up to approximately 25,000 feet (7,500 metres 
 approx.). 
 
 (&) Those which show a continued increase in velocity after 
 the gradient velocity has been reached, which con- 
 tinued increase is unaccompanied by much change 
 in direction. 
 
 (c) Those which show a decrease of velocity after the 
 
 gradient velocity has been reached. 
 
 (d) Those which show great changes, or even reversals, in 
 
 direction in the layer under consideration. 
 (0) Those which show an upper wind (either in the north- 
 west or south-west quadrant) crossing the lower wind 
 and probably blowing out from a distinct low 
 pressure centre. 
 
 The following comments upon the various groups are also 
 largely taken from Captain Cave's book. 
 
 So far as the British Isles are concerned, it appears that 
 group 6 is the commonest form of wind distribution. Most of 
 the increase in velocity however occurs in the lower part of the 
 range. 
 
 Group c, comprising those winds which show a velocity 
 falling off with height, seems almost entirely restricted to 
 easterly winds and is not -manifested frequently. 
 
 The big directional changes, which are the characteristic 
 feature of group d, seem nearly always to be preceded by 
 the dropping of the wind to very low velocity at the particular 
 heights where the changes are to occur. They are due to 
 the superposition of two currents of air of great difference in 
 direction upon one another under conditions which are not 
 made evident by the current synoptic charts of the surface. 
 
 Groups d and e are closely connected with the production of 
 thunderstorms ; and a sounding of the upper air which reveals 
 either a great change of direction in the upper layers (group d\ 
 or an upper wind in the north-west, or south-west quadrant 
 crossing a lower wind (group e), may be taken as indicating 
 a great risk of a thunderstorm occurring in some part of the 
 neighbourhood of the place of sounding. 
 
 Much further invest! gational work is needed, however, and 
 also the perfecting of some experimental means of measuring 
 wind direction and velocity above low cloud, before really 
 definite statements concerning the wind behaviour between 
 8,000 and 25,000 feet can be made. 
 
 112. Winds above 25,000 feet. Winds above 25,000 feet 
 (roughly 7,500 metres) are mainly in what is termed " the 
 stratosphere," and will be treated in the chapter which deals 
 with that part of the atmosphere. 
 
 *"The Structure of the Atmosphere in Clear Weather," by C. ,T. P. 
 Cave. Cambridge University Press. 1912. 
 
92 
 CHAPTER XIX. 
 
 The Troposphere and Stratosphere. 
 
 113. Temperature Measurements in the Upper Air. Measure- 
 ments of temperatures in the upper air are obtained chiefly by 
 the use of either balloons or kites. * 
 
 The data obtained reveals when plotted a most remarkable 
 result. This result will become evident upon reference to 
 Fig. 17, which shows the temperatures obtained in 45 soundings 
 done for the Meteorological Office, London, in 1907-1908. 
 
 CURVES SHOWING CHANGE OF TEMPERATURE WITH HEIGHT ABOVE SEA-LEVEL 
 DETAINED FROM BALLON- 50NDE ASCENTS 1907-8. 
 
 20- 
 
 16- 
 
 [OiTCHAM flW 2JWB. PVKIDN HIU 
 ofcnjJofTCHAM RWK Oftes. PTRTON HIU 
 
 HO -60 -40 -Z ZO 
 
 iTMPCRATURE ABSOLUTE 
 
 The separate curves represent the relation between temperature and 
 height in miles or kilometres in the atmosphere. The numbers marking 
 the separate curves indicate the date of ascent at the various stations as 
 shown in the tabular columns. The difference of height at which the 
 isothermal layer is reached, and the difference of its temperature for 
 different days or for different localities, is also shown on the diagram by the 
 courses of the lines. 
 
 FIG. 17. 
 
 * For details of the methods employed, the reader is referred to the 
 Meteorological Office, London, Publication No. 202 : " The Free Atmosphere 
 in the Region of the British Isles," which contains a Report by W. H. Dines, 
 F.R.S., " On Apparatus and Methods in Use at Pyrton Hill." ' 
 
93 
 
 The diagram shows that all the soundings manifest a decrease 
 of temperature with elevation up to a height of somewhere near 
 to 30,000 feet (approximately 10 kilometres), but that after that 
 the temperature ceases to fall. 
 
 Considering anyone ascent, it is found that the change from 
 the region of falling temperature to that of steady, or hardly 
 changing, temperature is generally fairly abrupt. The place of 
 change is, indeed, often marked by an inversion of temperature, 
 that is, by an increase of temperature with height, which inver- 
 sion, however, does not persist through any great thickness of 
 atmosphere, and soon the steady or hardly changing state is 
 reached. 
 
 A further examination of Fig. 17 will show that the height 
 at which the change occurs varies. 
 
 It is clear, then, that the atmosphere may be regarded as 
 divided into two definite parts so far as temperature is 
 concerned : 
 
 (a) a lower portion in which there is a fairly regular and 
 considerable lapse, or fall, of temperature with height ; 
 and 
 
 (6) an outer shell in which there is no material change of 
 temperature with height. 
 
 The lower portion, that in which the fall of temperature 
 with height does occur, is called the " troposphere/' and the 
 outer portion, that in which no material change of temperature 
 occurs with height, is called the " stratosphere." Both these 
 terms are suggested by Teisserenc de Sort. The level of 
 change from the troposphere to the stratosphere is termed the 
 " tropopause." 
 
 114. The Height of the Tropopause. As a general rule, the 
 boundary between the troposphere and the stratosphere is quite 
 definite, but, in some cases, the temperature gradient ceases 
 gradually, and, in others, it stops for some distance and then 
 starts again. Also, inversions of temperature in the neighbour- 
 hood of the boundary are, as has already been mentioned, quite 
 frequent. 
 
 Some definition of what, in these varying cases, is to be 
 taken as the level of the tropopause is clearly called for ; and 
 the following are the instructions issued on the point by the 
 Meteorological Office, London. 
 
 Type 1. When the stratosphere commence with an inversion, 
 the level of the tropopause is the height of the first 
 point of zero temperature gradient. 
 
 Type 2. When the stratosphere begins with an abrupt 
 transition to a temperature gradient below 2 Abso- 
 lute per kilometre (that is, 3 '6 Fahrenheit per 
 3,280*8 feet) without inversion, the level of the 
 tropopause is the height of the abrupt transition. 
 Type 3. When there is no such abrupt change of tempera- 
 ture gradient, the level of the tropopaus is to be 
 
94 
 
 taken at the point where the mean fall of tempera- 
 ture for the kilometre next above is 2 Absolute 
 or less (that is, for the 3,280*8 feet next above 
 is 3 ' 6 Fahrenheit or less), provided that it does not 
 exceed 2 Absolute (3 '6 Fahrenheit) for any subse- 
 quent kilometre. 
 The height of the tropopause varies at the same place at 
 
 different times. The mean height of the tropopause at various 
 
 places is given in Table VII. 
 
 TABLE VII.* 
 Mean Height of the Tropopause. 
 
 Place. 
 
 Period of 
 Observation. 
 
 Number of 
 Observa- 
 tions. 
 
 Mean Height of 
 Tropopause. 
 
 Kilometres. 
 
 Feet (to 
 nearest 100). 
 
 Scotland - 
 
 1908-1914 
 
 29 
 
 9-8 
 
 32,200 
 
 Ireland ... 
 
 1908-1914 
 
 27 
 
 10-1 
 
 33,100 
 
 Manchester 
 
 1908-1914 
 
 73 
 
 10-3 
 
 33,800 
 
 England, S.E. - 
 
 1908-1914 
 
 167 
 
 10-7 
 
 35,100 
 
 British Isles 
 
 1908-1910 
 
 150 
 
 11-1 
 
 36,400 
 
 J> 5 
 
 1912 
 
 52 
 
 10-0 
 
 32,800 
 
 Berlin 
 
 1 
 
 
 
 
 Strassburg 
 
 [ 1904-1909 
 
 212 
 
 10-5 
 
 34,400 
 
 Yienna 
 
 J 
 
 
 
 
 Petrograd 
 
 1904-1909 
 
 41 
 
 9-6 
 
 31,500 
 
 Paris 
 
 1904-1909 
 
 90 
 
 10-5 
 
 34,400 
 
 Italy - - . 
 
 1904-1909 
 
 46 
 
 11-0 
 
 36,100 
 
 Paris 
 
 i 
 
 
 
 
 Hamburg - 
 
 1 1904-1909 
 
 158 
 
 10-6 
 
 34,800 
 
 Brussels - 
 
 J 
 
 
 
 
 A glance at Table VII. suggests that the height of the 
 tropopause is dependent to some extent upon latitude ; and this 
 is the case, it being lowest near the poles. The boundary 
 level between the stratosphere and the troposphere seems, in 
 fact, to slope downward from the equator towards the poles. 
 This suggestion has received strong support from some ascents 
 carried out on Lake Victoria in the equatorial regions of Africa 
 in which the stratosphere was not encountered until a height of 
 55,800 feet (approximately 17 kilometres) was reached, which is 
 about 16,000 feet (approximately 5 kilometres) above the usual 
 maximum height in Europe. 
 
 The level of the tropopause over Enrope, too, varies with 
 cyclonic and anti-cyclonic conditions on the surface, being 
 lower with cyclonic conditions than with anti- cyclonic. 
 
 This suggests that there might be a definite connection 
 between the level of the tropopause and the barometric pressure 
 
 * Geophysical Memoirs, No. 13. "The Characteristics of the Free 
 Atmosphere." By "W. H. Dines, F.R.S. Meteorological Office, London, 
 1919. The table has been re-arranged, part omitted, and heights expressed 
 in feet as well as in kilometres, ; 
 
95 
 
 at the ground level. This is, in fact, found to be the case, the 
 mean level being lowest with the lowest pressures, and rising 
 with increasing pressures. 
 
 115. The Stratosphere. Care needs to be exercised in 
 thinking about the distribution of temperature in the strato- 
 sphere, a care which is rendered all the more necessary because 
 of the rather unfortunate name of " isothermal layer," which 
 was formerly given to that outer portion of the stratosphere. 
 The stratosphere is practically isothermal in the vertical 
 direction, but it is not isothermal in the horizontal one. That 
 is to say, if an observer with a thermometer, starting at the 
 base of the stratosphere, could ascend straight upwards, he 
 would find practically no change in the readings afforded by 
 his thermometer at various stages of his ascent, but if he 
 moved horizontally through the stratosphere, either at the level 
 of the tropopause or along some level higher up, he would 
 find definite and appreciable changes recorded. It is clear, 
 therefore, that the name " isothermal layer " for the stratosphere 
 is very unsatisfactory as liable to lead to wrong ideas ; and it 
 has now been superseded, the term " stratosphere " having 
 come into general use. 
 
 116. Temperature in the Troposphere. As would be expected 
 from what has been already said about the variation of inso- 
 lation received under different conditions by the earth's surface, 
 and by the further differentiation effected by the two components 
 of that surface, land and water (see section 14), the temperature 
 of the air in the 2,000 or 3,000 feet (650 or 900 metres approxi- 
 mately) just above the earth's surface is subject to considerable 
 fluctuations. It has been stated, too, that inversions of tempe- 
 rature near the ground (see section 23) are quite common. 
 
 Despite these fluctuations, however, it is valuable to examine 
 the mean temperature at various heights. Table VIII. is that 
 given by Mr. \V. H. Dines, F.R.S.* ; it shows the mean monthly 
 temperatures for England. Mr. Dines has " smoothed " it to get 
 rid of artificial irregularities. He gives his table in kilometres 
 and degrees Absolute. In Table IX! his data has been trans- 
 ferred to read in feet and degrees Fahrenheit to meet the 
 situations concerning units set out in section 104. 
 
 117. The Seasonal Variation of Temperature in the Tropo- 
 sphere. It is interesting to inquire as to how far upward the 
 seasonal variation of temperature, which is so marked a feature 
 of the ground level, extends. An examination of the data given 
 in either Table VIII. or IX. for England, and also of similar 
 tables which have been prepared for other localities, reveals the 
 fact that the seasonal variation extends upwards for about 
 30,000 feet (roughly 10 kilometres), after which there is little 
 change. 
 
 * Geophysical Memoirs, No. 13. " The Characteristics of the Free 
 Atmosphere." W. H. Dines, F.R.S. Meteorological Office, London 
 (M.O. 220c), 1919. 
 
96 
 
 ,3 
 
 JS 
 
 s 
 
 XL, 
 
 0> 
 
 t/J 
 
 2 
 Er, 
 
 
 f 
 
 [> 
 
 H 
 
 EH 
 
 Oi(MCOdO5CDOlOSiOOCO<M 
 
 T? CO d r-l O Ci 
 
97 
 
 
 * -*-> 
 
 O G> 
 
 I .8 
 
 I .a 
 
 f 
 
 XI 
 
 K 
 
 c3 ^ 
 
 m ^ 
 
 H 
 
 o -P 
 
 ^1 
 
 81 
 
 S 
 
 B 
 
 (X 
 
 O 15406 
 
 T 1 C3 CO CO 
 
 i IGOOOiOCOi^-CDCOO 
 
 i i i i M ?7 i 
 
 ocooocooocoioooT-icqi 10 
 
 CO O CO CO u, CO Cj. M 
 
 OO(MCOCOI>-iO^T I rH <M CO (M CO 
 CO CO CO CO 10 ^ CO GJ rH 
 
 OlCOCO 
 
 i CO ?Q rH rH i I (M CO 
 I I I I I I I I I I 
 
 r-l<N<MCO 
 
 O O OfO OOOOOOOOOOO 
 
 O5l>^i lOOiOtMOIr-^rHOOCOCO 
 
98 
 
 118. The Daily Temperature Range in the Troposphere. A 
 similar inquiiy may be put with regard to the diurnal variation 
 of temperature, which is again so marked a meteorological feature 
 of the ground level. Whilst it is certain that this diurnal 
 variation decreases very rapidly with height and becomes very 
 small, being, indeed, less than 2 Fahrenheit at 4,000 feet (that 
 is, less than 1*2 Centigrade at about 1| kilometres), it seems 
 impossible to say at what height, if any, it ceases to exist at all. 
 
 It depends to some extent upon locality. Lieut -Col. Gold,* 
 for example, working up results obtained at Berlin gives a 
 daily variation there at 1 kilometre (approximately 3,280 
 feet) of 0'85 Centrigade (1'53 Fahrenheit) with the maxi- 
 mum occurring at 6 p.m., and a daily variation at 2 kilometres 
 (approximately 6,560 feet) of 0'6i Centigrade (I'lfi Fahren- 
 heit) with the maximum occurring at noon ; whilst for Petrograd 
 he gives the daily variation at 1 kilometre as 0'72 Centigrade 
 (1*30 Fahrenheit), the maximum occurring at about 2.30 p.m. 
 
 119. Yearly Mean Temperatures for Different Pleiyhts in 
 Various Areas. A point of great interest is seen when the 
 mean temperatures for different heights in certain selected 
 areas are examined. Table X, showing some of these mean 
 temperatures, is extracted from one given by Mr. \V. H. Dines, 
 F.R.S.j 
 
 The remarkable feature revealed by the table is the extra- 
 ordinarily low temperature prevailing in the stratosphere over 
 the equator. 
 
 120. Temperatures over Depressions and Anti-Cyclones. 
 Equally interesting results are revealed by the observations 
 that have been taken in the upper air over depressions and 
 anti-cyclones on the surface. Once again, we are indebted 
 to Mr. W. H. Dines for the statistics that illustrate the matter. 
 His figures appear in Table XI. which has been extracted with 
 some re-arrangement and addition from one given in " The 
 Computer's Handbook " (section II.), issued by the Meteorological 
 Office, London. The depression figures are those connected 
 with low pressures on the surface, the mean of the pressures 
 being 984 millibars (29 '06 inches), and the anti-cvclonic figures 
 those connected with high pressures on the surface, the mean 
 of those pressures being 1,031 millibars (30 '45 inches). It 
 will be noticed, too, that the mean pressures, as well as the 
 mean temperatures, are given for the various heights. 
 
 * Geophysical Memoirs, No. 5. " The Intel-national Kite and Balloon 
 Ascents." E. Gold, M.A. Meteorological Office, London, Publication 
 210e, 1913. 
 
 f Geophysical Memoirs, No. 13. " The Characteristics of the Free 
 Atmosphere." W. H. Dines, F.R.S , Meteorological Office, London. 
 Publication 220c, 1919. 
 
09 
 
 ! 
 a 1 
 
 ClOlfMG^IQOCO-^OS^OOCO^t^xOlr^OCvlCOClrHO 
 rHrHrHrHOOaiI>-CO^jCO<M rH CO -* lO CO l> 00 
 
 rH rH i 1 rH rH rH 
 
 1 1 1 I 1 1 1 1 1 1 1 I I 
 
 n 
 
 i i 
 
 oo i 
 "i 
 
 1 
 
 *8 
 
 i 
 
 li I 
 4*4 
 
 ill 
 
 a-* S 
 
 
 1 
 
 1 1 1 1 |>OCOOCOI^t>COCOC5rHCMCOCNI 
 
 |cO<^>COCOiO^CO(M rH (N "CO 
 
 1 
 
 OOGOOOOCOOOO<MC5CO<MOOCOI>-O 
 
 I 1 
 
 
 COCOCOCOCOCO00^1^t^ 
 O CO CO CO CO CO CO CO CO CO ^i CO (M rH rH (M CO CO T? 
 
 1 1 1 1 1 1 i [ ! ! i ! 1 1 i 
 
 2^ 
 
 
 
 d rH rH 
 
 G 2 
 
100 
 
 TABLE XI. 
 
 Average Values of Pressure and Temperature at Various 
 Heights over High-Pressure and Low-Pressure Areas. 
 
 Height, 
 
 Depressions. Anti-Cyclones. 
 
 Kilo- 
 metres. 
 
 Feet to 
 nearest 100. 
 
 Pressure 
 (Milli- 
 bars). 
 
 Tempera- 
 ture 
 ( Abso- 
 lutej. 
 
 Tempera- 
 ture 
 ( Fahren- 
 heit). 
 
 Pressure 
 (Milli- 
 bars). 
 
 Tempera- 
 ture 
 ( Abso- 
 lute). 
 
 Tempera- 
 ture 
 ( Fahren- 
 heit). 
 
 
 
 
 I 
 
 
 
 
 15 
 
 49,200 
 
 116 
 
 
 
 
 
 123 
 
 
 
 ' 
 
 14 
 
 45.900 
 
 135 
 
 224 
 
 56 
 
 146 
 
 215 
 
 72 
 
 13 
 
 42,700 
 
 157 
 
 226 
 
 -53 
 
 171 
 
 215 
 
 72 
 
 12 
 
 39,400 
 
 183 
 
 225 
 
 54 
 
 201 
 
 217 
 
 -69 
 
 11 
 
 36,100 
 
 212 
 
 225 
 
 -54 
 
 235 
 
 221 
 
 62 
 
 10 
 
 32,800 
 
 247 
 
 225 
 
 54 
 
 273 
 
 226 
 
 -53 
 
 9 
 
 29.500 
 
 288 
 
 226 
 
 53 
 
 317 
 
 233 
 
 -40 
 
 8 
 
 26,200 
 
 335 
 
 227 
 
 -51 
 
 366 
 
 2-10 
 
 -27 
 
 7 
 
 23,000 
 
 388 
 
 232 
 
 -42 
 
 422 
 
 247 
 
 15 
 
 6 
 
 19,700 
 
 449 
 
 240 
 
 27 
 
 483 
 
 254 
 
 - 2 
 
 5 
 
 16,400 
 
 516 
 
 248 
 
 13 
 
 552 
 
 261 
 
 10 
 
 4 
 
 13,100 
 
 591 
 
 255 
 
 
 
 628 
 
 267 
 
 21 
 
 3 
 
 9,800 
 
 H75 
 
 263 
 
 14 
 
 713 
 
 272 
 
 30 
 
 2 
 
 6,600 
 
 767 
 
 269 
 
 25 
 
 807 
 
 277 
 
 39 
 
 1 
 
 3,300 
 
 870 
 
 ' 275 
 
 36 
 
 913 
 
 279 
 
 43 
 
 
 
 
 
 984 
 
 279 
 
 43 
 
 1,031 
 
 282 
 
 48 
 
 An examination of Table XI. shows the following points : 
 
 (a) That up to 10 kilometres (roughly 33,000 feet) 
 temperatures over cyclonic areas are lower than 
 those over anti-cyclonic ones ; but that, above that 
 level, that is, roughly from the base of the strato- 
 sphere upwards, temperatures over the low 
 pressures are higher than those over the high. 
 
 (I) That even at the greatest heights examined the low 
 pressure evident at the surface in depressions 
 continues to manifest itself. 
 
 This latter point (6) seems to suggest that depressions on 
 the surface must be deemed to extend to great heights, a matter 
 of interest in any discussion as to the origin of cyclones and 
 anti-cyclones. It is now thought, indeed, that pressure changes 
 on the surface are mainly due to changes occurring in the 
 stratosphere. 
 
 The old idea that the cores of depressions are coin posed of 
 relatively warm moisture-laden air must also be discarded. 
 Table XL shows clearly enough that the lower layers over 
 depressions have temperatures which are relatively low com- 
 pared with the surroundings. Comparison, too, of the 
 
101 
 
 temperatures over depressions reveal ths^fect thnc tii;ey/4ire 
 actually lower than the normals for the levels considered. 
 
 121. Temperature Gradients in the Troposphere. Normally, 
 the temperature falls off with height until the stratosphere is 
 reached. The rate of change per kilometre step of vertical 
 height is termed the "temperature gradient." Although 
 normally, as has been said, the temperature falls with height, 
 the reverse is often noticed, so that the temperature gradient 
 may be either positive or negative. By international agree- 
 ment it is deemed positive when the temperature is falling on 
 the particular step of height examined, and negative when 
 the temperature is rising. An inversion of temperature (see 
 section 23), then, is just a change from a positive temperature 
 gradient to a negative one. Such inversions in the 'various 
 layers of the troposphere are quite common, happening almost 
 invariably in and just above layers of fog, and frequently in 
 and just above other forms of clouds. 
 
 The term " gradient " seems to have become somewhat over- 
 worked in the subject and it is better to restrict it to changes 
 on the horizontal. Another term, "lapse" (Latin, lapsus, a 
 slip), is coming now into use in the discussion of temperature 
 and pressure changes in the vertical. Using the newer 
 nomenclature, "vertical temperature gradient" becomes "lapse 
 rate." 
 
 Measurements of this lapse rate have been fairly numerous. 
 
 Berson,* for example, dealing with data obtained from 
 manned balloons towards the close of the last century, drew 
 attention to the marked constancy of the lapse rate, about 
 5 Centigrade per kilometre (9 C ' Fahrenheit per 3,300 feet, 
 approximately), up to a height of 4 kilometres (roughly 
 13,000 feet), and to the sudden and considerable increase in 
 its value in the next and succeeding layers. He attributed the 
 change to the fact that the level of 4 kilometres marks roughly 
 the upper limit of the lower clouds, and that near this height 
 inversions of temperature are more frequent than at neighbouring 
 heights, inasmuch as temperature inversions are characteristic 
 of the top of such cloud layers. 
 
 Values deduced by E. G old from later manned balloon ascents 
 on the Continent (1901-1907) do not show the peculiarity com- 
 mented upon by Berson quite so clearly, but, nevertheless, 
 it is still distinctly noticeable. Gold's valuesf are given in 
 Table XII. 
 
 * " Wissenschaftliche Luftfehrten," Braunschweig, 1899, 3 vols. 
 f Geophysical Memoirs, No. 5. " The International Kite and Balloon 
 Ascents," E. Gold, M.A. Meteorological Office, London, 1913. 
 
102 
 
 TABLE XII. 
 Upward Temperature Gradient. 
 
 Height: 
 
 
 
 | 
 
 
 
 
 
 Kilometres 
 
 0-1 
 
 1-2 
 
 2-3 i 3-4 
 
 4-5 
 
 5-6 
 
 6-7 
 
 7-8 
 
 Feet (in hun- 
 
 0-33 
 
 33-66 
 
 66-98 
 
 98-131 
 
 131- 
 
 164- | 197- 
 
 230- 
 
 dreds). 
 
 
 
 
 
 164 
 
 197 i 230 
 
 262 
 
 
 
 
 i 
 
 Gradient : 
 
 
 
 
 
 
 
 
 
 Degrees Abso- 
 
 4-3 
 
 5-1 
 
 5-1 
 
 5-8 
 
 6-2 
 
 6-9 
 
 7-5 
 
 6-2 
 
 lute. 
 
 
 
 
 
 
 
 
 
 Degrees Fahr. 
 
 7-7 
 
 9-2 
 
 9-ii 
 
 10-4 
 
 11-2 
 
 12-4 
 
 13-5 
 
 11-2 
 
 Number of Cases 
 
 50 
 
 50 
 
 44 
 
 40 
 
 34 
 
 22 
 
 10 
 
 3 
 
 Similar work in determining temperature gradients or lapse 
 rates in the upper air done in various places reveals the fact that 
 the gradient ceases to exist at different heights in different 
 latitudes, the higher the latitude, that is, the nearer the Pole, 
 the lower being the height. This would be expected from what 
 has already been said (see section 114) about the lowering of 
 the level of the tropopause with higher and higher latitudes. 
 
 122. Winds in the Stratosphere. Any statement concerned 
 with winds at such a great height as is connoted by the strato- 
 sphere must be tempered by the fact that comparatively few 
 experimental soundings have been made reaching such heights, 
 and then only in clear sky condition, inasmuch as the small 
 balloon method, practically the only one possible, cannot be 
 used to high altitudes in the presence of low cloud. The results 
 obtained, then, can hardly be used, at present, to make sound 
 generalisations. What generally seems to happen, however, is 
 that there is very little change of direction shown, either in the 
 transition from the troposphere to the stratosphere, or in the 
 actual stratosphere itself ; but that velocity falls off fairly 
 rapidly. The falling off in velocity is in agreement with the 
 equality of pressure which is found in the stratosphere. It 
 seems hardly possible to say much more at present about the 
 winds in the stratosphere ; better experimental methods for 
 sounding the atmosphere at such great heights being needed, 
 and especially some method applicable in all types of sky 
 conditions. 
 
103 
 CHAPTER XX. 
 
 Pressure, Density, and Humidity in the Upper Air. 
 
 123. Variation of Pressure with Height. The variation of 
 pressure with height is closely connected with the temperature 
 of the air. The formulae in general use is : 
 
 \ fy) = C (kg Po tog Pi) 
 
 where h^ and h are the heights at two points vertically over 
 each other, p { and p the pressures at those points, 6 the 
 absolute temperature of the column of air between those points, 
 and is a numerical constant. If. heights are measured in 
 metres, C is 67 '4 ; if in feet, 221 ' 1. 
 
 Speaking quite strictly, the formula only holds if temperature 
 conditions are uniform, but no great error is introduced if 6 is 
 taken as the mean temperature of the air between the two 
 points. 
 
 But it is just the determination of what is to be taken as 6, 
 the mean temperature, that presents difficulty in practice. As 
 has been already shown, the temperature in the first few 
 thousand feet is subject to great fluctuations, and no lapse rate 
 could be assumed which would be adequate to meet all possible 
 cases. 
 
 Obviously, the best plan would be to make actual observa- 
 tions of the temperature, say, at every 500 feet, on every 
 occasion that the formula was to be used. The mean of these 
 many readings would give a fairly accurate 6. But this is 
 quite clearly out of the question on most occasions, and, even 
 if it were possible and if the temperatures were taken by actual 
 observation of a thermometer on an aeroplane, it must be 
 remembered that the accuracy of such observations is somewhat 
 seriously in doubt ; first, because there is a vicious circle in 
 that aeroplane heights are measured by altimeters (see section 
 124) which depend entirely upon measurements of pressure and 
 have to be calibrated on a temperature assumption, and, second, 
 because the thermometers carried are, to some extent, affected 
 by the aeroplane, and especially by the necessarily hot engine. 
 
 In view of this, the one course remaining is to adopt such 
 a standard lapse rate of temperature which shall afford the best 
 possible approximation to general conditions. This rate is 
 usually taken as 6 centigrade per kilometre (that is, 3 '3 
 Fahrenheit per 1,000 feet). Hence, if one temperature actually 
 existent be known, the value of can be readily calculated.* 
 
 * Recently, Toussaint's formula for the change of temperature with 
 height has received much recognition. The formula is 
 
 T = T -6-5fe, 
 
 where h is measured in kilometres and T is the temperature at the surface 
 and T the temperature at height h, both T and T being measured in degrees 
 Centigrade. 
 
104 
 
 When cases of temperature inversion at the surface are sus- 
 pected, the temperature actually occurring at or near that 
 surface would obviously be a very bad starting point. 
 
 124. The Altimeter. In view of the almost universal use of 
 the altimeter in modern aircraft as an indicator of height, some 
 discussion of it seems pertinent. 
 
 It is an ordinary aneroid barometer (see section 70) graduated 
 to show heights instead of actual pressures, and the mechanism 
 is so devised that the height scale may be uniform. The fact 
 that it is really an instrument for measuring upper air pressures 
 even though it expresses them as heights justifies the inclusion 
 of the consideration of it in this particular chapter. 
 
 It is clear that the scale of the altimeter must be graduated 
 with regard to some definite temperature. That temperature 
 is taken in England as 50 Fahrenheit (283 Absolute), this 
 particular standard being due to Airy. Airy's rule for correct- 
 ing the readings given by his height scale reads as follows : 
 When the temperature differs from 50 Fahrenheit, the recorded 
 
 height is to be ^creased by the -nnnrth part for every 
 
 f\ \\C\\7P 
 
 below ^ Fahrenheit. Speaking more strictly, the fraction 
 
 should read TuW instead of Tinnr- When the Absolute scale is 
 used, the rule becomes as follows : When the temperature 
 differs from 283 Absolute, the recorded height is to be 
 
 creased by the -gg-g-rd part for every degree Ve 283 
 
 Absolute. 
 
 When no actual temperature observations are made on the 
 aeroplane itself, so that the temperature at any level is not 
 experimentally known for the purpose of correcting the height 
 recorded by the altimeter, then a lapse rate of 1 ' 5 Absolute 
 (2*7 Fahrenheit) per 1,000 feet should be assumed. 
 
 The temperature correction is not, however, the only one 
 that must be made if the height scale is to show even an 
 approximately correct reading. 
 
 In the formula previously quoted (section 123) : 
 7ij h = C0 (log p log Pi), 
 
 if h be sea level, then it may be written as zero and p Q becomes 
 the pressure at sea level. 
 
 Further, remembering that the formula applies to points 
 vertically beneath one another, p , at any moment of observation, 
 at any moment of reading the altimeter, is implied to be the 
 pressure at sea level immediately underneath the observer. 
 But the pressure at sea level differs from place to place and 
 from time to time. The observer has set his altimeter scale at 
 zero against the sea level pressure of his starting point. 
 Obviously, then, if the observer retain the same zero on his 
 
105 
 
 instrument throughout a long flight, his height readings will 
 be much in error. 
 
 Meteorology can do much to help the observer in this matter. 
 Let it be assumed that he is preparing to make a flight from 
 London to Carlisle, and that he contemplates starting at 10 a.m. 
 By the time of start he could be supplied with a synoptic chart 
 showing the distribution of pressure over England at, say, 
 7 a.m. that same morning. Fortunately, horizontal changes of 
 pressure are fairly persistent, and such a synoptic chart would 
 be of service to him for some hours after it was drawn. From 
 it, he would be able to see at a glance the pressures that 
 prevailed at 7 a.m. along his route and, by looking at the 
 barometric characteristics and tendencies (see section 73) also 
 recorded on the chart, would be able to see, especially if provided 
 with a forecast of coming pressure changes by a competent 
 meteorologist, what the pressure at sea level would be likely to 
 be at any known point of his route. During the flight, then, 
 he could alter the zero of his altimeter to agree with the sea 
 level pressures over which he is flying, and hence his instrument 
 will give him a much closer approximation to his true height. 
 The method outlined may not be perfect but it is a far superior 
 one to that which assumes that the sea level pressure at London 
 is the same as that at Carlisle, and that it will not alter 
 appreciably with time. 
 
 An even better method, when it is available, is that, on 
 passing over or near to a meteorological station, he should 
 obtain the sea level pressure prevailing at that station directly 
 from it by means of wireless telegraphy and then correct his 
 altimeter accordingly. 
 
 The altimeter is, however, an unsatisfactory instrument, 
 mechanically, for measuring pressures, and hence heights, 
 because of the " lag " which characterises it. The "lag" 
 becomes evident in that the mechanism will not record the 
 proper pressure at once. Especially is this the case if it has 
 been kept at a low pressure, that is, a great height, for an hour 
 or so ; on its return, it indicates less than the proper pressure, 
 that is, more than the proper height, and only gradually 
 recovers. Other mechanical defects due to the effect of the 
 acceleration of the aeroplane upon it also arise in practice. 
 
 125. Mean Pressures in the Upper Air. There is no method 
 of measuring directly the pressure at various heights in the 
 upper air with accuracy. Mr. W. H. Dines, F.R.S., uses the 
 following to arrive at the pressures required. A small 
 hydrogen-filled balloon has a meteorograph attached. A meteoro- 
 graph* is a small light instrument which gives an automatic 
 
 * For a description of the meteorograph, see " The Computer's Hand- 
 book." M.O. 223, Section II., Subsection II. Meteorological Office, 
 London. 
 
106 
 
 record of pressure and temperature. The balloon, carrying the 
 meteorograph, is released and its course is followed from the 
 ground by means of two theodolites at the ends of a long base 
 line. This means that the height of the balloon, at any given 
 time, can be calculated trigonometrically without much likeli- 
 hood of serious error. The surface pressure is read from a 
 standard mercurial barometer. Let it be supposed that this 
 surface pressure is x millibars. When the meteorograph is 
 recovered (after it has fallen owing to the bursting of the 
 balloon), its continuous record of temperature will enable the 
 mean temperature between the recorded pressures of x and, say, 
 y millibars to be determined. Hence the height at which the 
 pressure was y millibars can be computed. A similar operation 
 can be worked through. using the pressure of the y millibars at 
 the computed height as starting point ; and, so continuing, the 
 pressures at various heights can be obtained. 
 
 Table XIII. is extracted from the larger table given by Mr. 
 Dines.** It shows the mean pressures at each kilometre height 
 up to 20 kilometres (roughly 66,000 feet) at various places, and 
 reveals that there is no appreciable variations in mean pressure 
 with longitude, but that there is with latitude, the pressure being 
 lower in higher latitudes than in lower though the Canadian 
 values seem somewhat exceptional, perhaps owing to paucity of 
 observations. 
 
 126. Seasonal Variation of Pressure at Various Heights. 
 Another interesting store of statistical data is given by Mr. 
 Dines in the same publication as that from which Table XII. 
 was extracted. Table XIV. is a copy of that given by him (with 
 the exception that a column giving heights in feet is included) 
 and shows the mean pressures at various heights over England 
 for the various months of the year. The lower pressures in the 
 winter at the higher levels should be noted. They are produced 
 chiefly by the seasonal changes of temperature at those heights. 
 Mr. Dines shows the range in the last column and states that, if 
 uniform temperature above 15 kilometres (approximately 49,000 
 feet) had been assumed, the range at 20 kilometres (approxi- 
 mately 66,000 feet) would have been 5 milibars. 
 
 127. Densities in the Upper Air The density of a sample 
 of air is the weight, or, better, the mass, of a measured volume 
 of it. It depends upon a variety of factors ; the composition, the 
 pressure, the temperature. It increases with pressure, decreases 
 in reversed proportion to the absolute temperature, and also 
 decreases with increasing content of water vapour. 
 
 * Geophysical Memoirs, No. 13. " The Characteristics of the Free 
 Atmosphere." W. H. Dines, F.R.S. Meteorological Office, London, 1919. 
 
107 
 
 "* <M 
 
 CO CO CO 
 
 1 
 
 w 
 
 
 C5 -H O O 
 
 GO O GO -fi 
 
 <M CO ^ O 
 
 <M <M CO CO 
 
 o r*. co I-H 
 
 "* "* iO CO 
 
 CO CO t^ QO 
 
 'T-^ 
 
 OiCQCOOiOOCOCO? I 
 i l(M(MCOCO'^l"^iOCO 
 
 CO'^l"^iOCOCOI>- 
 
 O <M QO CO 
 OCOI>OOOrHCOCOQO 
 
 100 
 
 
 M| 
 
 s 
 
 i IrHrHrHrH 
 
108 
 
 c r^ 
 
 35 ' 
 
 
 r-Tr-T tH rH Cfl <N 00 CO ** P CO t 00 
 
 II 
 
 OOiOCOCC-CO'-tiCOOOOl^COOOj-lQO 
 
 1 
 
 CO Ol (M CO O r-l t^ Tf) rH O O5 O5 
 
 O 
 
 -I 
 
 'I 
 
 8 
 
 IT 
 
 a, 
 
 O 
 
 O 
 
 I 
 
 fi 
 
 0, 
 
 i IOO5O 
 
 r- ii |xOCO>O(?l^OCOOii ( 
 
 O5 OO rH Oi 
 
 S'S rH in '3 S C0 -^1 * >:00 
 
 ii 
 
 tJD 
 
 1 
 
 aT CD o" os co o os cc co os cc 
 
 o os cc 
 
 CQrHt-l 
 
309 
 
 The actual , formula, taking account of these various factors, 
 which allows of the density being calculated is as follows : 
 
 where d is the density to be found, and d Q is the density of 
 perfectly dry air under the conditions p = 1,000 millibars 
 (29 '53 inches) and T =290 Absolute. These conditions give 
 d Q = l,"20l grammes per cubic meter, p is the barornetic 
 pressure in millibars of the sample of air under consideration, 
 and e is the pressure of water vapour in the sample. 
 
 Neglecting water vapour altogether, the formula takes 
 the simpler form 
 
 and the error introduced is quite small. 
 
 TABLE XV. 
 Mean Densities for Different Heights, South-East England. 
 
 Height. 
 
 Temperature. 
 
 Pressure. 
 
 Density. 
 
 Kilometres. 
 
 Feet, 
 nearest 100. 
 
 Degrees 
 Absolute. 
 
 Millibars. 
 
 Grammes per 
 Cubic Metre, 
 
 20 
 
 65,600 
 
 219 
 
 55 
 
 i 87 
 
 19 
 
 62,300 
 
 219 
 
 64 
 
 102 
 
 IS 
 
 59,100 
 
 219 
 
 75 
 
 119 
 
 17 
 
 55,800 
 
 219 
 
 88 
 
 139 
 
 16 
 
 52,500 
 
 219 
 
 102 
 
 162 
 
 15 
 
 49,200 
 
 219 
 
 120 
 
 191 
 
 14 
 
 45,900 
 
 219 
 
 140 
 
 223 
 
 13 
 
 42,700 
 
 219 
 
 164 
 
 261 
 
 12 
 
 39,400 
 
 219 
 
 192 
 
 305 
 
 11 
 
 36,100 
 
 220 
 
 224 
 
 355 
 
 10 
 
 32,800 
 
 222 
 
 261 
 
 409 
 
 9 
 
 29,500 
 
 228 
 
 303 
 
 463 
 
 8 
 
 26,200 
 
 234 
 
 352 
 
 524- 
 
 7 
 
 23,000 
 
 241 
 
 407 
 
 589 
 
 6 
 
 19,700 
 
 248 
 
 469 
 
 658 
 
 5 
 
 16,400 
 
 255 
 
 538 
 
 735 
 
 4 
 
 13,100 
 
 262 
 
 615 
 
 819 
 
 3 
 
 9,800 
 
 268 
 
 699 
 
 909 
 
 2 
 
 6,600 
 
 273 
 
 795 
 
 1,014 
 
 1 
 
 3,300 
 
 278 
 
 900 
 
 1,128 
 
 
 
 
 
 282 
 
 1,014 
 
 1,253 
 
 Table XV., again extracted from a larger one given by 
 Mr. W. H. Dines,* gives some idea of the value obtained. The 
 
 * Geophysical Memoirs, No. 13. " The Characteristics of the Free 
 Atmosphere." W. H. Dines, F.B.S. Meteorological Office, London, 1919. 
 
110 
 
 values given take no account of water vapour and this makes 
 them a little too high as a reference to the first of the two 
 formulas just given will make clear. The error, as has been 
 already said, is, however, quite small even in the lower layers, 
 where it is greatest : in the higher layers, it is almost negligible. 
 The variation of gravity with height has also been neglected, 
 but, once again, the error introduced by so doing is very small. 
 
 It should be noticed that the density increases with eleva- 
 tion. This is due to the fact that the diminution of pressure and 
 the decrease of density due to that diminution has a greater effect 
 than the decrease of temperature and the increase of density 
 due to that cause. 
 
 These questions of pressure temperature, and density of 
 various levels in the atmosphere have a very great importance in 
 the problems of buoyancy arising in connection with balloons or 
 airships. 
 
 128. Humidity in the Upper Air. As would be expected, 
 relative humidity shows, in general, an increase with height up 
 to the levels of the lower cloud layers, that is, up to the height 
 of 1 to 2 kilometres (3,300 to 6,600 feet approximately), but 
 after that it falls off fairly rapidly and at great heights seems 
 always to be very low. An inversion of temperature in the 
 upper atmosphere is nearly always characterised by a low 
 relative humidity, this again being in accord with what would 
 be expected. 
 
 129. The Height of the Atmosphere. Very few soundings of 
 the upper air to a greater height than 20 kilometres (roughly 
 12J miles) have been made, but it is quite certain that the 
 atmosphere extends much above that height. Formerly, the 
 height of the atmosphere was given as about 80 kilometres 
 (roughly 50 miles), that being the height at which the air 
 would have no appreciable pressure. This figure, however, 
 has had to be considerably amended as a result of the observa- 
 tions made on meteors. Meteors, or "shooting stars" are 
 small solid bodies, entering the earth's atmosphere from outer 
 space Their velocity is so great that they generate enough 
 heat, either by actual friction or by the compression of the air 
 in front of them, to render themselves luminous. If various 
 observers note the apparent position of the same meteor with 
 regard to the stars, data is afforded whereby the angular altitude 
 of each observer's line of sight may be determined. The inter- 
 section of these different lines of sight enables the height of the 
 luminous streak due to the incandescent meteor to be computed. 
 From the results so obtained, it is thought that the atmosphere 
 extends with sufficient density to produce the phenomena of 
 
 * For a treatment of this subject the student is referred to the article 
 of " Buoyancy " in " The Meteorological Glossary," issued by the 
 Meteorological Office, London. 
 
Ill 
 
 luminosity in meteors to a height of about 300 kilometres 
 (roughly 188 miles). How far beyond this level traces oi 
 the atmosphere extend is, at the moment, purely a matter of 
 speculation. 
 
 It is clear that much, very much, still remains to be done in 
 the investigation of the upper air. Point has been added quite 
 recently to this remark by the publication of an important paper 
 by Professor S. Chapman and Mr. E. A. Milne.* Formerly, it 
 was thought that the highest layers of the atmosphere were 
 almost entirely composed of hydrogen. Chapman and Milne 
 doubt the presence of hydrogen in the atmosphere and show 
 that if hydrogen be neglected, the higher layers, say above 130 
 kilometres (roughly 80 miles), are composed mainly of helium, 
 and that the very highest layers are composed practically 
 entirely of that gas, a speculation of great interest in regard to 
 theories concerning magnetic storms, the aurora borealis, and 
 the like. 
 
 As was stated in section 1, however, that part of the atmo- 
 sphere with which the meteorologist is concerned in his daily 
 work does not differ appreciably in chemical composition from 
 the actual surface layers, and, in any case, chemical composition 
 is not, at present, of great importance to him. Nevertheless, it 
 is of interest and may become of more vital importance to him 
 as knowledge advances. 
 
 The change of view so recently brought about does, however, 
 emphasise the fact that meteorology, like all live sciences, is 
 dynamic and not static, and that its bounding edge is continually 
 advancing. 
 
 * The Composition, lonisation, and Viscosity of the Atmosphere at great 
 Heights." S. Chapman, F.R.S., and E. A. Milne. Quarterly Journal Roval 
 Meteorological Society, Yol. XLYI. 1920 
 
112 
 APPENDIX. 
 
 Further Reading. 
 
 For information concerning the conduct of a meteorological station, and 
 for details of the various instruments employed, together with tables neces- 
 sary for the proper working up of the readings afforded by the instruments, 
 the student is referred to " The Observer's Handbook," issued by the 
 Meteorological Office, London. ~ * ? 
 
 For questions more particularly relating to physical geography, " The 
 Realm of Nature," by Dr. H. R. Mill (Murray), still remains one of the 
 best written in English, containing a fund of information in a comparatively 
 small compass. 
 
 Of text-books : on general meteorology, B. G. K. Lempf ert's " Meteoro- 
 logy " (Methuen & Co.), A. E. M. Geddes' " Meteorology " (Blackie), and 
 W. J Humphreys' " Physics of the Air " (Philadelphia, Franklin Institute), 
 may be thoroughly recommended. 
 
 A. W. Clayden's "Cloud Studies" (John Murray, London), and 
 G-. A. Clarke's " Clouds " (Constable & Co.) are also strongly to be recom- 
 mended on their particular subject. 
 
 For definitions of terms employed and for short but authoritative 
 articles upon various aspects of the subject, arranged alphabetically, " The 
 Meteorological Glossary," issued by the Meteorological Office, London, 
 should be in the hands of all earnest students of the subject. It is, at once, 
 very comprehensive, very clear, and very cheap, and its information is 
 thoroughly to be trusted. 
 
 For fuller discussions of the problems arising in connection with fore- 
 casting weather from synoptic charts, Sir Napier Shaw's " Forecasting 
 Weather" (Constable & Co., London) will repay the most careful study. 
 It is extremely valuable in that it forces attention to the intimate relation- 
 ship existing between physics and meteorology. From the historical point 
 of view, " Weather," by the Hon. R. Abercromby, the writer and worker 
 who did so much to crystallise knowledge concerning forecasting from 
 synoptic charts, is full of interest and still remains full of instruction. 
 
 For certain broad aspects of meteorology in relation to man, Dr. H. N. 
 Dickson's "Climate and Weather" (Home University Series, Williams and 
 Norgate, London), is very valuable. 
 
 For fuller reading on the conditions prevailing in the upper air, the 
 student is referred to the "Manual of Meteorology" (Part IV.), by Sir 
 Napier Shaw, F.R.S., which has, as its sub-title, " The Relation of the Wind 
 to tht3 Distribution of Barometric Pressure," and which is published by the 
 Cambridge University Press ; and to C. J. P. Cave's " The Structure of the 
 Atmosphere in Clear Weather," published by the same press. Reference 
 should also be made to " The Characteristics of the Free Atmosphere," by 
 W. H. Dines, F.R.S., which is issued as Geophysical Memoirs, No. 13, by 
 the Meteorological Office, London, and to "The International Kite and 
 Balloon Ascents " (Geophysical Memoirs, No. 5), by E. Gold, F.R.S., 
 issued by the same office. Both these last-mentioned are standard works 
 upon the subject of which they treat. 
 
 Meteorology is, however, a distinctively live science, and new and 
 important work is being done in it daily in various countries and published 
 in various forms and languages. So far as the United Kingdom is con- 
 cerned, most new work is to be found in the current publications of the 
 Royal Society of the Meteorological Office, London (especially in the two 
 series known respectively as "' Geophysical Memoirs " and as " Professional 
 Notes "), and of the Royal Meteorological Society (Quarterly Journal). 
 
 If the student wishes to maintain himself abreast of new knowledge and 
 of new view points in the subject, a periodical inspection of these publica- 
 tions is essential. 
 
113 
 
 INDEX. 
 
 (Numbers refer to Sections,) 
 
 Abercromby, Hon. R. - - 80 
 
 Absorption - - 14 
 
 , Effect of clouds on - 14 
 
 Airy - - 124 
 
 Altimeter - - 123, 124 
 
 Anemometer, Dines* Tube - - 6 
 
 Aneroid barometer - 70, 103 
 
 Anti-cyclones - - , - 86 
 
 , Temperature over - 120 
 
 Archibald - 105 
 
 Assmann - - 36 
 
 Atmosphere - - 1 
 
 , Adiabatic changes in - 21 
 
 , Height of - 129 
 
 , Helium in - - 129 
 
 , Pressure of - - - - - 66 
 
 Bar - - 69 
 
 Barograph - - 71 
 
 Barometer, Aneriod - 70, 103 
 
 , Diurnal variation of - 72 
 
 readings, Corrections to be applied to - 63 
 
 , Units used in - 69 
 
 , Simple - - 67 
 
 , Standard mercurial - 67 
 
 Barometic characteristic 73 
 
 tendency - - 73 
 
 units - 69 
 
 Beaufort, Admiral Sir F. Notations for weather - - 59 
 
 Wind scale - 5 
 
 Bersoii - 121 
 
 Bora - - 17 
 
 Bruckner, E. - -101 
 
 Buys Ballot's Law - 75 
 
 Cave, E, J. P. - 105, 107, 111 
 
 Chapman, E. H. . 105 
 
 Chapman, S. , . . 129 
 
 Cloud Atlas, International - 45, 46 
 
 Cloud bursts - . 94 
 
 Cloud particles, Size of - 32, 49 
 
 Clouds Alto-cumulus - - 46 
 
 Alto-stratus - 4(? 
 
 Anvil - 46, 93 
 
 Cirro-cumulus - * - - 46 
 
 Cirro-nebula - - 4G 
 
 Cirro- stratus - . 46 
 
 Cirrus 44, 46 
 
 O 15406 H 
 
114 
 
 Clouds, Cumulo-nimbus - 46 
 
 , Cumulus 44, 46 
 
 , False cirrus - 46, 93 
 
 , Formation of - - 49 
 
 , Fi-acto- cumulus - 46 
 
 , Heap - - 47 
 
 , in sheets - 47 
 
 , Lenticular - - 46 
 
 ,, , Marnmato-cumulus - - 46 
 
 , Nimbus - 46 
 
 , Sequence of, in depressions - - 81 
 
 , Strato-cumulus - - - 4t> 
 
 , Stratus 44, 46 
 
 , Study of - 48 
 
 Cols - - 87 
 
 Cooling due to expansion - 20 
 
 Compression, Heating due to - 19 
 
 Condensation - 23, 32 
 
 at temperatures below the freezing point 52, 55 
 
 by dynamical cooling - - 49 
 
 , The necessity of a nucleus for - 36 
 
 Cyclone, Yarying use of term - - 79 
 
 De Bort, Leon Teisserenc 45, 113 
 
 Density - 127 
 
 in the upper air - 127 
 
 , mean, for different heights - - 127 
 
 Depression - - 79 
 
 , Distribution of cloud in - 81 
 
 , To determine the direction of travel of a - - 83 
 
 , Nomenclature of a - 80 
 
 ,; , Rate of advance of a - 84 
 
 , Secondary - - 88 
 
 , Sequence of wind and weather in a - 85 
 
 , Y-shaped - 89 
 
 , Weather in a .- 81 
 
 Depressions, Temperature over - 120 
 
 , Travel of - 82 
 
 , Tropical - - 96 
 
 Dew - - - 33 
 
 Dew point - - 29 
 
 Dines, W. H. - - 6, 113, 114, 11G. 119, 120, 125, 126, 127 
 
 Diurnal, variation of pressure - - 72 
 
 Dobson, G. M. B. - 108 
 
 Doldrums - 11 
 
 Douglas, C.K.M. - - 93 
 
 Eddies 9, 105 
 
 Ekholm,N. - - 83 
 
 Evaporation - - 27 
 
 Expansion, Cooling- due to - - i2<> 
 
 Fiducial temperature - - - 69 
 
 Fitzroy, Admiral - -S, 103 
 
 Fog at sea, Conditions for - - 39 
 
 Fog on land - - 40 
 
115 
 
 Fogs - - 38 
 
 Forecasting by a single observer - 100 
 
 Procedure of - 99 
 
 Forecasts 98 
 
 Long range - - 101 
 
 Foucault - 10 
 
 Frost, Conditions for - 34 
 
 Gla/ed - - 58 
 
 Further Outlook - - 98, 99 
 
 Glazed frost - - 58 
 
 Gold,E. - 78, 118, 121 
 
 Gradient, Adiabatic temperature - 22 
 
 Temperature, in troposphere - 121 
 
 Vertical temperature - - 22 
 
 Wind - - 77 
 
 Guilbert, C., Method of Forecasting - - 83 
 
 Gusts - - - - -.,:-.- - - 7, 8 
 
 Halos - -46, 47, 85 
 
 Hail - - 53 
 
 Hail, Soft - 54 
 
 Haze - - 42 
 
 Heating due to compression - - 19 
 
 Hellmann - 105 
 
 Hildebrandsson, H. H. - - 45 
 
 Howard, Luke - - 44 
 Humidity, Absolute 
 
 Relative - - 30 
 
 Measurements of - - 31 
 
 in Upper Air - - 128 
 
 Hurricanes - - 96 
 
 International Cloud Atlas 45, 46 
 Insolation 
 
 Inversions of temperature 23, 121 
 
 Inwards, R. - - 102 
 Isallobars 
 
 Isobars 74 
 
 , Straight - 91 
 
 Land and sea breezes - - 15 
 
 Land and water, Unequal heating of - 
 
 Lapse rate - Ipl 
 
 Latent heat of vaporisation 
 
 Line squall - . - - 92 
 
 Maxims, Weather 102 
 
 Meteorograph - - 125 
 
 O 15406 I 
 
116 
 
 Meteorology, Definition of - 1 
 
 , Importance of means of rapid communication in - 65 
 
 , Yalue of 2 
 
 , , to airmen - -4 
 
 , Relation of, to Physics , - 1 
 
 , Units in - 69, 104 
 
 Meteors - 129 
 
 Millibar - - - 69 
 
 Milne, E. A. - - 129 
 
 Mist - - 41 
 
 Molecular Theory of Matter - - 27 
 
 Monsoons ........ 8 
 
 Nuclei for condensation - - t ,, - - - - 36 
 
 Pamperos - - 97 
 
 Periodicity in meteorological elements - 101 
 
 Pressure, Atmospheric - 66 
 
 , Conversion table for - 69 
 
 , Mean, in the upper air - 125 
 
 , Measurement of 67 
 
 over Earth - - 10 
 
 , Seasonal variation of, with height - - 126 
 Pressure, Yapour 
 
 , Variation of, with height - - 123 
 
 Radiation - - 14 
 
 , Effect of clouds on - 14 
 
 Rainbows - 102 
 
 Rain, formation of - 51 
 
 Riggenbach, A. - 45 
 " Roaring Forties " 
 
 Rotation, Effect of earth's - - 10 
 
 Russell, H. C. - - - - 101 
 
 Scud - - 4(j 
 
 Sea and land breezes - - 15 
 
 Secondary depressions - 
 
 Shaw, Sir Napier - 82, 101, 105 
 
 Simpson, G. C. 5, 26, 33, 55 
 
 Sleet - -57 
 
 Smoke aggregations, Lessons from 
 
 "Smudging" - 
 
 Snow - 
 
 Squall, Line - 
 
 Squalls - 
 
 Stratosphere - . 113, 115 
 
 , Winds in 
 
 Standard temperature of barometer - 
 Stevenson, T., Screen - 
 Straight isobars 
 Supersaturation 
 
 Synoptic Charts, Meaning of - ... - 65 
 
 , Making of - - 73 
 
117 
 
 Taylor, G. I. - - 105 
 
 Temperature, Annual range of - 25 
 
 Daily range of, in troposphere - - 118 
 
 Distribution of, over globe - - 24 
 
 Fiducial, of barometer - - 69 
 
 gradients in troposphere - 121 
 
 in troposphere - - 116 
 
 inversions 23, 121 
 
 , measurements in upper air - 113 
 
 over depressions and anti- cyclones - - 120 
 
 , Seasonal variation of, in troposphere - 117 
 
 , Standard, of barometer - 69 
 
 , Upward, gradient - 22, 113 
 
 , Yearly mean, for different heights - - 119 
 
 Thermometer, Wet and dry bulb - 31 
 
 Thunderstorms 93, 111 
 
 Torricelli - - - 67 
 
 Tornadoes 79, 95 
 
 Trade winds - - 10 
 
 Travel of depressions - -. - 82 
 
 Tropopause, - - 113 
 
 , Height of- - 114 
 
 Troposphere - - 113 
 
 , Temperature gradients in - 121 
 
 , Temperature in - - 116 
 
 , The daily mean range in - 118 
 
 , Seasonal variation of temperature in - - 117 
 
 Typhoons - 79, 96 
 
 Upper air, Density in - - 127 
 
 , Humidity in - 128 
 
 ,, , Mean Pressure in - - 125 
 
 Upper winds Methods of measuring - - 107 
 
 Vapour pressure - 28 
 
 Visibility - 43 
 
 V-shaped depressions - .... 89 
 
 Water drops, fall of, through the air - - 50 
 
 , Terminal velocities of - - 50 
 
 Waterspouts - - 95 
 
 Weather, Beaufort Notation for - 59 
 
 changes, The control of - 60 
 
 glass - 103 
 
 , International Symbols for - - 59 
 
 lore and sayings - - 102 
 
 in a depression 81, 85 
 
 of the Frigid Zones - - 64 
 
 of the Temperate Zones - - 63 
 
 of the Torrid Zones - - 62 
 
 Wedges of high pressure - 90 
 
 Wegand - 36 
 
 Wegener, A. - ..... 33, 3G 
 
118 
 
 Wind above 25,000 feet 
 , Backing of 
 
 between 8,000 and 25,000 feet - 
 direction between surface and 8,000 feet 
 , Gradient 
 in the stratosphere 
 . Lull of, at night 
 - *^ surface - 
 
 UNIVERSITY OF CALIFORNIA LIBRARY 
 
 - 112 
 
 - 83 
 
 - Ill 
 
 - 109 
 
 77 
 
 - 122 
 
 - 18 
 
 - 105 
 
 - 85 
 
 LD2l-ioo W -i2,>46(A201 2s i 6 )4120 
 
 Printed under the authority of His MAJESTY'S STATIOSTEBY OFFIC*; 
 
 By Byre and Spottiswoode, Ltd., East Harding Street, E.C. 4, 
 
 Printers to the King's most Excellent Majesty. 
 
YD 10849- 
 YC 10850 
 
 519641 
 
 
 UNIVERSITY OF CALIFORNIA LIBRARY