THE LIBRARY 
 
 OF 
 
 THE UNIVERSITY 
 OF CALIFORNIA 
 
 PRESENTED BY 
 
 PROF. CHARLES A. KOFOID AND 
 MRS. PRUDENCE W. KOFOID 
 
UNIVERSITY EXTENSION SERIES. 
 
 EDITED EY J. E. SYMES, M.A., 
 
 Principal of University College, Nottingham. 
 
 METEOROLOGY 
 
nt&er*ttg Extension Series. 
 
 UNDER the above title MESSRS. METHUEN are publishing a series of books on 
 Historical, Literary, and Scientific subjects. The University Extension Move- 
 ment has shown the possibility of treating History, Literature,_ and Science in a 
 way that is at once popular and scholarly. The writers in this series aim at a 
 similar combination. Each volume will be complete in itself, and the subjects 
 will be treated by competent writers in a broad and philosophic spirit. 
 
 Ube ^following Dolumes are 
 THE INDUSTRIAL HISTORY OF ENGLAND. H. DE B. GIBBINS, M.A. 
 
 Second Edition Revised. With Maps and Plans. [25. 6d. 
 
 A HISTORY OF POLITICAL ECONOMY IN ENGLAND: from Adam 
 
 Smith to Arnold Toynbee. L. L. PRICE, M.A. [as. 6d. 
 
 PROBLEMS OF POVERTY. An Inquiry into the Industrial Conditions of 
 
 the Poor. J. A. HOBSON, M.A. [as. 6d. 
 
 11 The object of this -volume is to collect, arrange, and examine some of the 
 
 leading facts^nd forces in modern industrial life which have a direct j 
 
 bearing upon Poverty, and to set it in the light they afford some of the 
 
 suggested palliatives and remedies" Extract from Author's Preface. 
 VICTORIAN POETS.. AMY SHARP, Newham College, Cambridge. [25. 6d. 
 Chapters on Browning, Tennyson, Swinburne, Mrs. Browning, A. If. Clough, 
 
 Matthew A mold, Rossetti, Win. Morris, and others. 
 
 PSYCHOLOGY. F. S. GRANGER, M.A. [25. 6d. 
 
 THE EVOLUTION OF PLANT LIFE : Lower Forms. An Introduction to 
 
 the Study of Cryptogamic Botany. G. MASSEE. [25. 6d. 
 
 " The aim of the present book is to briefly indicate in a broad sense the most 
 
 prominent features, structural and physiological, that characterise plant 
 
 life as manifested at the present day." Extract from Author's Preface. 
 THE FRENCH REVOLUTION. J. E. SYMES, M.A., Principal of University 
 
 College, Nottingham. With Map of France. [25. 6d. 
 
 AIR AND WATER. Professor V. B. LEWES, M.A. Illustrated. [25. 6d. 
 
 ENGLISH SOCIAL REFORMERS. H. DE B. GIBBINS, M.A. [25. 6d. 
 
 CONTENTS: I. Langland and Ball. II. More, III. Wesley and Wilberforce. 
 
 IV. The Factory Reformers. V. Kingsley. VI. Carlyle and Ruskin. 
 ENGLISH TRADE AND FINANCE IN THE SEVENTEENTH CENTURY. 
 
 W. A. S. HEWINS, B.A. [ 2S . 6d. 
 
 AGRICULTURAL BOTANY. M. C. POTTER, M.A. Illustrated. [ 3 s. 6d. 
 
 [In the Press. 
 KIMMINS. THE CHEMISTRY OF LIFE AND HEALTH. By C. W. 
 
 KIMMINS, Downing College, Cambridge. Illustrated. Crown 8vo. [23. 6d. 
 THE MECHANICS OF DAILY LIFE. V. P. SELLS, M.A. Illustrated. [ 2S . 6d. 
 ELECTRICAL SCIENCE. G. J. BURCH, B.A. [ 3 s. 
 
 Ube ^following Dolumes ave in preparation; 
 
 ENGLISH PAINTERS. D. S. MACCOLL, M.A., Fellow of University College 
 
 London, U.E. Lecturer in Art and Literature. 
 
 THE PRE-SHAKESPERIAN DRAMA. H. M. FITZGIBBON, M.A 
 THE EARTH. An Introduction to Physiography. E W SMALL M A 
 AGRICULTURAL CHEMISTRY. M. J. R. DUNSTAN, M.A. 
 HYGIENE. J. LORRAINE SMITH, M.A., M.D. 
 PLANTS AND ANIMALS. P. CHALMERS MITCHELL, M.A 
 
 METHUEN & CO., 18 BURY STREET, W.C. 
 
METEOROLOGY 
 
 Elements of aHeatfjer an* Climate 
 
 BY 
 
 H. N. DICKSON 
 
 F.R.S.E., F.R.MF.T.SOC. 
 
 JSetfjuen & Co, 
 
 18 BURY STREET, LONDON, W.C. 
 1893 
 
QC 
 
 3)55 
 
 CONTENTS 
 
 CHAP. PAGE 
 
 I. FUNDAMENTAL FACTS .... 1 
 
 II. FUNDAMENTAL PRINCIPLES ... 23 
 
 III. CYCLONES ...... 44 
 
 IV. ANTI-CYCLONES : OTHER FORMS OF PRESSURE 
 
 AREAS .... . . 73 
 
 V. WEATHER FORECASTING .... 94 
 
 VI. METEOROLOGICAL INSTRUMENTS AND OBSERVA- 
 TIONS ...... 103 
 
 VII. THE ELEMENTS OF CLIMATE . . .136 
 
 VIII. APPLICATION OF METEOROLOGY TO AGRICUL- 
 TURE . . . . . .156 
 
 INDEX . 1 86 
 
PREFACE 
 
 IN the earlier stages of the study of meteorological science, 
 the student has in most cases the advantage of familiarity 
 with the general nature of the phenomena he has to con- 
 sider. I have endeavoured in this book to utilise this 
 common knowledge, and to show the methods of eliminating 
 its errors, and of arranging it in scientific order. An 
 attempt is thus made to lay down a certain amount of 
 " permanent way," specially adapted to practical purposes, 
 but at the same time leading towards the more theoretical 
 grounds of modern research. 
 
 For the benefit of those who wish to pursue any branch 
 of the subject farther, numerous references are given in 
 foot-notes. Since this book was written, the results of 
 recent work in Germany have been presented in elementary 
 form to English readers by Professor Frank Waldo in his 
 " Modern Meteorology/' to which the non-mathematical 
 reader, who is interested in the theory of the subject, is 
 referred. Like other departments of physical science, 
 
v iii PREFACE. 
 
 meteorology does, however, steadily tend to become in- 
 debted to the higher mathematics ; and its development can 
 be best studied, failing the original papers, from Professor 
 Abbe's collection of translations, recently published by the 
 Smithsonian Institution. 
 
 My best thanks are due to Dr. Hugh Robert Mill, 
 Librarian to the Royal Geographical Society, and to Mr. 
 R. S. Cole, R. N. E. College, Devonport, for friendly help 
 in revising the proofs. 
 
 H. N. D. 
 
 OXFORD, September, 1893. 
 
METEOROLOGY 
 
 CHAPTER I. 
 
 FUNDAMENTAL FACTS. 
 
 i. Weather. The word weather is ordinarily used to de- 
 note (or, rather, connote) the general condition of the atmo- 
 sphere at any particular time, and especially of the thin 
 layer of it near the surface of the earth which more immedi- 
 ately concerns us. Thus the word is commonly coupled 
 with such adjectives as "good" and "bad," which are 
 strengthened or not according to the effects produced upon 
 us or our interests individually. But weather may be good 
 or bad in a great variety of ways. A very rainy day is not 
 "good," although it be dead calm ; neither is a rainless day 
 on which there is great wind ; and worst of all, perhaps, is 
 one with neither rain nor wind, but a still, dank mugginess, 
 or a raw, penetrating mist. 
 
 2. Elements of Weather. Such variations suggest that 
 weather is a general result produced by the combined action 
 of several different elements, each consisting of a special set 
 of phenomena in the physics of the atmosphere, such as 
 those depending on its motion, warmth, transparency, or the 
 like. We are accordingly led by a process of induction to 
 analyse weather into various constituents, the study of each 
 of which forms a separate branch of meteorology. 
 
 3. Meteorology. Meteorology is the name given to the 
 science which deals with the physical conditions and 
 
 A 
 
2 METEOROLOGY. 
 
 changes of the atmosphere generally, whether directly con- 
 cerned with weather or not ; and its ultimate object is to 
 account for atmospheric phenomena by considering the 
 earth as an astronomical body revolving round the sun as a 
 source of heat and rotating on its own axis, and then de- 
 ductively to predict them from such considerations. It 
 will be our business to inquire first what are the beliefs 
 about the weather suggested by everyday experience, how 
 far those beliefs are untrustworthy and why, and in what 
 direction more reliable information is to be sought. We 
 shall in this way be led to rediscover for ourselves some of 
 the leading principles of modern scientific meteorology, and 
 to make use of their practical applications. 
 
 4. Sources of Information. To begin, then, with the re- 
 sults of everyday experience of the weather, we must first 
 ascertain from what sources our facts are to be derived, and 
 then examine the nature of the evidence. There can be no 
 doubt that the best experience is the most dearly bought, 
 and we may therefore look for it amongst those who have 
 most to gain or lose from accurate observations of the 
 weather, such as, for example, farmers, sailors, or fishermen. 
 Again, it will be admitted that those beliefs are most 
 worthy of examination which gain the concurrence of the 
 largest number of competent judges. Hence we select those 
 most commonly accepted as accurate, and in this there is, 
 fortunately, little difficulty, for we find everywhere a current 
 weather folk-lore, formulated in the shape of sayings and 
 proverbs, which may fairly be taken to represent general 
 experience, and discussed accordingly. 
 
 5. Prognostics. The proposition laid down in a weather 
 " saying " or proverb is usually to the effect that if one 
 event happens, a second is to be regarded as probable ; the 
 probability of the second event succeeding the first being 
 deemed sufficiently great to justify an observer of the latter 
 forecasting or predicting the other. The chance that if the 
 
FUNDAMENTAL FACTS. 3 
 
 first event happens the second will follow it is expressed 
 in the usual way, as the ratio of the number of times the 
 second event follows the first to the total number of times 
 the first occurs. Take, for example, the Scotch proverb : 
 
 " A wet May and a winnie, 
 Makes a fou stacky and a finnie." 
 
 This means that out of a large number of cases in which the 
 month of May has been wet and windy, the majority have 
 been followed by an abundant harvest, and that therefore 
 such a harvest is likely to follow any particular wet and 
 windy May. The value of this saying as a prognostic is simply 
 the ratio of the two numbers ; thus, if out of 50 wet and 
 windy Mays all are recorded as having been followed by 
 good harvests, the prediction for any one season would al- 
 most amount to a certainty ; if 33 good harvests were re- 
 corded, the odds would be about 2 to i in favour of a 
 successful prediction ; while if we had 25 good years, the 
 chances would be even and the prognostication worthless. 
 Or, generally, if we denote "certainty" by i, the chance of 
 any event happening again which has already been known to 
 happen a times out of n is a and the chance against its 
 
 happening is - . n 
 
 It may be said that as yet all weather forecasting depends 
 on the principle just explained. The aim of the meteor- 
 ologist is, of course, to obtain a knowledge of the chain of 
 causes and effects connecting the first event with the second, 
 and in that way to be able to demonstrate that, given the 
 first event, the second must happen, whether it ever hap- 
 pened before or not. The changes of the weather, although 
 not " uncertain," are nevertheless extremely complex, and a 
 deductive method of forecasting is still in the distant future. 
 In the meantime, we adhere to the old method of "pro- 
 
4 METEOROLOGY. 
 
 babilities," trusting to a better selection of " first events " 
 for more successful prediction of the second. 
 
 6. Causes of Error. This leads us to inquire what are 
 the causes of error which make certain weather prognostics 
 so much less accurate than others. The first is obviously 
 to be looked for amongst errors of observation. Observa- 
 tional errors are of two kinds, personal errors depending on 
 individual peculiarities, and errors common to all observers. 
 With the first class we are not at present concerned, since we 
 are dealing with results obtained by many observers, whose 
 personal errors may be assumed to neutralise each other. 
 But such an assumption cannot be made about the second 
 class of errors, since the probability is not that they will 
 counteract each other, but that they will be all in one 
 direction. By way of illustration, suppose an iron plate 
 placed on a table and allowed to remain until it is known 
 to be neither hotter nor colder than the air. If two people 
 try to estimate the temperature of the plate by touching it, 
 they are more likely than not to form different impressions. 
 Again, let the same people try to estimate the temperature 
 of the plate by touch after it has been immersed in boiling 
 water for a time ; in this case the impressions will also pro- 
 bably differ, and we observe that the individual who thought 
 the plate coldest before now thinks it hottest. But now 
 suppose we place a wooden slab alongside the iron plate 
 and set anybody to determine by touch which is the warmer 
 of the two. If they are at a temperature near that of the 
 air the iron will invariably feel the colder, and if near that 
 of boiling water invariably the warmer. In the former case 
 the difference was entirely a personal matter, depending on 
 whose hands were warmer to begin with or had the thicker 
 skin, while in the latter it was a genuine case of deception 
 by the senses; both iron and wood were at the same 
 temperature, but the former being a good conductor ab- 
 stracted heat rapidly from the hand, and gave the impression. 
 
FUNDAMENTAL FACTS. 5 
 
 of being colder than the wood which retained the heat from 
 the hand in a thin layer near its surface. 
 
 In order to get rid of the latter class of errors it is usual 
 to bring the evidence of another sense to bear on the facts 
 under investigation, by employing a special device or instru- 
 ment of some kind ; thus in the case of the iron and wood 
 slabs we make the sense of sight available by means of a 
 thermometer. We may, therefore, define the function of an 
 instrument as being to confirm or contradict the evidence 
 of one sense by that of another, or more generally to afford 
 direct evidence of phenomena which appeal only indirectly 
 to the unaided sense. 
 
 So far errors in observing the facts ; the next cause of 
 failure lies in drawing wrong conclusions, and of these the 
 most fruitful source is the want of accurate records. 
 Amongst the practical observers from whom prognostics are 
 derived, the best are perhaps the least likely to record their 
 observations accurately, and after the lapse of time occur- 
 rences are retained in the memory not because of their 
 frequency or actual importance, but because of their relative 
 importance to the observer or the industry in which he is 
 interested. Such errors are evidently likely to diminish 
 the more frequently the occurrence takes place \ hence we 
 find a very large percentage of proverbs relating to storms 
 quite trustworthy, because gales occur in different parts of 
 these islands some 10 to 20 times annually, while proverbs 
 relating to seasons are seldom so reliable, only one observa- 
 tion a year being obtainable. In the latter case people are 
 apt to draw premature conclusions from a small number of 
 observations, and then to become unconsciously biased, 
 finding confirmatory evidence where there is none, until the 
 supposition becomes a fixed article of belief. One of the 
 best examples of errors arising in this way is to be found in 
 the "equinoctial gales." Careful records kept for a long series 
 of years have shown that the probability of gales occurring 
 
6 METEOROLOGY. 
 
 is not specially increased at or about the time of the equin- 
 oxes, i.e. that the " equinoctial gales " as such do not exist. 
 The belief that changes of the moon occurring at particular 
 hours are followed by special types of weather is so strong 
 and so widespread that it demands special notice. The 
 following table, due in its present form to the older Herschel, 
 may still be found in many books of tide tables, almanacs, 
 and the like. 
 
 If New Moon, Second Quarter, 
 Full Moon, or Last Quarter, 
 happen. 
 
 Weather likely to follow during the Quarter. 
 
 In Summer. 
 
 In Winter. 
 
 Between 
 Midnight and 2 Morning 
 2 Morning 4 Morning 
 4 Morning 6 Morning 
 6 Morning 8 Morning 
 
 8 Morning 10 Morning 
 
 10 Morning 12 Noon 
 Noon 2 Afternoon 
 2 Afternoon 4 Afternoon 
 4 Afternoon 6 Afternoon 
 
 6 Afternoon ,, 10 Afternoon 
 10 Afternoon ,, 12 Midnight 
 
 Fair 
 
 Hard frost unless S. or W. 
 Snow and stormy 
 Snow and stormy 
 Stormy 
 ( Cold, Rain, if wind W. 
 1 Snow, if E. 
 Cold, with high wind 
 Snow or Rain 
 Fair and Mild 
 Fair 
 Fair, frosty if N. or N.E. 
 Rain or snow if S. or S.\V. 
 Fair and frosty 
 
 Cold, with Showers. 
 Rain 
 
 Wind and Rain.. . 
 Changeable 
 
 Frequent Showers. 
 Very Rainy 
 Changeable 
 
 f Fair,if windN.W. 
 ( Rainy,ifS.orS.\V. 
 Fair 
 
 
 We remark that any errors in the first column of this table 
 are likely to produce consequences in astronomical and tidal 
 matters which would at once call attention to them, and the 
 published data of the moon's age may accordingly be ac- 
 cepted as correct. The indefmiteness of the predictions in 
 the second and third columns leaves a large margin for suc- 
 cess, but as the margin is equally large for failure, we may 
 assume that observations of whether the weather is " Fair," 
 " Very Rainy," or " Stormy," are correct, i.e. that the pro- 
 bable errors of observation are equal in both directions. 
 Now, since no reservations are made with reference to dis- 
 turbing influences other than that of the moon, the relation 
 of the weather to the lunar phases must be direct ; and, since 
 
FUNDAMENTAL FACTS. 7 
 
 the latter can be calculated for any time with certainty, we 
 ought to be able to predict the weather for any indefinite 
 time in advance, which we cannot do. Again, since no re- 
 servation is made as to the place for which the table holds 
 good, and since the change of the moon's phase is not a 
 local event, the sequence of weather-changes must be every- 
 where the same, which it is not. Hence, it appears that 
 the conclusions upon which the table are based are a priori 
 absurd, and, therefore, even a successful prediction in any 
 particular case cannot be more than a coincidence, as may 
 be easily shown by comparison of predicted and observed 
 weather. 
 
 7. Analysis of Weather. Following up the suggestion of 
 2, we proceed to resolve the weather into five constituent 
 elements, limiting ourselves for the present to those which 
 appeal directly to the senses, i.e. which can be recognised 
 without any instrumental aid. 
 
 (a) Wind. 
 
 [ Temperature. 
 ' ' | Dryness and Dampness. 
 
 (c) Cloud. 
 
 (d) Rainfall. 
 (f) "Meteors." 
 
 Since we are concerned at present only with the broader 
 changes of weather, the detailed examination of each element 
 may be deferred until we are in possession of better 
 methods of observation. Certain leading facts about each, 
 however, are evident. 
 
 Wind is motion of the atmosphere, and must therefore, 
 like all other motions, have velocity, i.e. direction and 
 speed. But since the atmosphere has mass, it must, when in 
 motion, be capable of exerting force, and as practical experi- 
 ence deals most frequently with effects produced by the 
 
8 METEOROLOGY. 
 
 wind, it is usual to speak of its force rather than of its 
 speed. 
 
 The idea of Temperature, or warmth and coldness, must 
 be taken along with that of dryness or dampness, because 
 without the help of instruments it is impossible to keep 
 from confusing the two ; moist air at a comparatively high 
 temperature feels colder than dry air which may really be far 
 below freezing-point, and the moderate heat of the tropics 
 seems often more intense than the excessive temperatures of 
 a desert. , We may in the meantime use words or phrases 
 describing merely the joint physiological effects, such as 
 " damp heat," " mugginess," " closeness," " dry cold," " fresh- 
 ness," and the like. 
 
 Clouds may be denned as portions of the atmosphere which 
 from natural causes have become temporarily visible. We 
 shall find afterwards that a varying proportion of the at- 
 mosphere consists of water vapour or steam, and that under 
 certain circumstances part of the vapour becomes condensed 
 in the form of cloud. The classification of clouds, which 
 usually depends on their form and height above the ground, 
 is a matter of great difficulty. A first rough division into 
 three classes may be made : clouds lying on the ground 
 called mist or fog according as they are wet or dry, heavy 
 rolled clouds at moderate elevations, and light fleecy clouds, 
 at great heights. 
 
 Rain is taken as a generic name for moisture precipitated 
 from the clouds by further condensation, and includes snow, 
 hail, sleet, and " drizzle." 
 
 Under the term " meteors " we include special phenomena 
 connected with the weather which do not directly come 
 under any of the foregoing heads, such as dew, hoar-frost, 
 pale or watery sun, rings or halos round the sun and moon, 
 rainbows, thunder and lightning, aurora, etc. 
 
 8. Example of Successive Changes. Selecting now a 
 change of weather at a particular place, we may observe the 
 
FUNDAMENTAL FACTS. 9 
 
 variations in the different elements, and see how far they 
 agree with previous experience as set forth in weather lore, 
 or how far we could have successfully predicted each stage of 
 the change from those preceding it. 
 
 At Holyhead, on the morning of 2Qth June, 1890, 
 after a few days of bright sunshiny weather with light 
 north-westerly winds, a solar halo was observed, and the sun 
 became pale and watery ; towards afternoon the wind 
 backed to S.W., and at six o'clock it blew a fresh breeze 
 from S.S.W., heavy clouds banking up from the westward. 
 The evening was damp and muggy, rain fell during the 
 night, and the morning of the 3oth broke gloomy and misty, 
 with the wind from S., and moderately rough sea. In the 
 course of the day the sky cleared and the wind veered 
 again to S.W., decreasing in force until night, when it 
 again freshened, and on the morning of July ist was 
 blowing in strong squalls from N. and N.W. with hard de- 
 tached clouds, threatening showers at intervals, but bright, 
 dry, invigorating atmosphere. 
 
 We note first the changes of the direction of the wind ; 
 from N.W. it gradually backed "against the sun " to S., and 
 after blowing steadily from that point for a time, veered 
 again to N.W. and N., becoming squally and unsteady. 
 This reminds us of the proverb 
 
 " When the wind veers against the sun, 
 Trust it not, for back it will run." 
 
 But, almost before the wind began to back, the sun was 
 encircled with a halo, and became pale and watery. Since 
 these appearances apply equally to the moon, it being 
 merely a question of which luminary happens to be visible, 
 we note indifferently, 
 
 " If the sun goes pale to bed 
 'Twill rain to-morrow, it is said." 
 
10 METEOROLOGY. 
 
 " The moon with a circle brings water in her beak." 
 " The sun is in his house, it will rain soon." 
 
 After the halo the sky became "dirty," and the dampness 
 increased. In many localities such conditions cause clouds 
 or " caps " to form round the tops of hills, hence such local 
 proverbs as 
 
 "When the clouds are on the hill 
 They'll come down by the mill." 
 
 " When Halldown has a hat, 
 Let Kenton beware of a skat." 
 
 " When Ruberslaw puts on his cowl, 
 
 The Dunion on his hood, 
 
 Then a' the wives of Teviotside 
 
 Ken there will be a flood." 
 
 Next, heavy soft clouds, banking up till the sky becomes 
 overcast, are recorded, easily recognised by their lowering 
 appearance, due to the dampness of the air, which also 
 causes them to reflect the glare of large fires or the lights of 
 cities : thus in Cumbrae and Bute rain is expected if the 
 glare of the Ayrshire ironworks is seen. 
 
 The combined physiological effects of dampness and high 
 temperature, described as " mugginess " or "closeness," is 
 felt by all living organisms, for, as the reverend and ingenious 
 Mr. Pointer remarks, " The very body of all animals and 
 vegetables is (as it were) a contexture of hygrometers, baro- 
 meters, and thermometers." Thus it is a sign of rainy 
 weather when peacocks cry much, when fowls pick up their 
 feathers with their bills, when birds that usually perch upon 
 trees fly to their nests, when swallows fly low and skim the 
 surface of water, or when sea-birds come in to the land. 
 
FUNDAMENTAL FACTS. II 
 
 " If the cock goes crowing to bed 
 He'll certainly rise with a watery head." 
 
 Concerning the goose " She is no witch, or astrologer, to 
 divine by the starres, but yet hath a shrewd guesse of rainie 
 weather, being as good as an almanack to some that believe 
 in her." 
 
 Again, "A bee was never caught in a shower." Wasps, 
 gnats, and midges, are said to be more than usually vicious 
 before rain. Rain may be expected " when sheep leap 
 mightily and frisk about, and push at one another with 
 their heads." 
 
 " It is time to stack your hay and corn 
 When the old donkey blows his horn." 
 
 Fishes rise more than usual, and dolphins and porpoises 
 are specially active before rain. Many plants close their 
 flowers and leaves on the approach of rain, as chickweed, 
 trefoil , pimpernel, convolvulus, etc. 
 
 A muggy, close atmosphere affects the spirits and tempers 
 of the human species, and aches, old wounds, rheumatism, 
 and corns become painful. The damp also makes an im- 
 pression on inanimate objects. Stones sweat or turn black, 
 and wood swells so that doors and windows creak or be- 
 come difficult to shut ; candle flames flare or snap ; ditches 
 or drains smell offensively ; soot loosens and falls down the 
 chimney ; cordage becomes taut. 
 
 Recurring now to our observations of 2Qth June, 1890, 
 we find all these prognostics justified. The muggy, damp 
 evening was followed by a night of rain, which began with 
 mist and drizzle. This state of things continued till 
 8 a.m. on June 30th ; but in the forenoon came a change. 
 The steady rain ceased with a " clearing shower," and 
 shortly after the damp muggy feeling passed away ; patches 
 
12 METEOROLOGY. 
 
 of blue sky became visible, and the wind began to veer to 
 westward and became squally ; at 2 p.m. " Wind S.W. and 
 sky cloudless, air dry and cool," is recorded. At 6 p.m. the 
 wind was westerly, and during the night it veered to north, 
 and blew in strong squalls, which continued during the 
 whole of July ist. These conditions are well described in 
 the sayings : 
 
 " When as much blue sky is seen as will make a Dutch- 
 man a jacket, the weather may be expected to clear 
 up." 
 
 " When, after a shower, hard clouds open up overhead, 
 leaving broken, ragged edges pointing upwards, wind will 
 follow." 
 
 " If it begin to rain from the south, with a high wind 
 for two or three hours, and the wind falls but the rain con- 
 tinues, it is like to rain twelve hours or more ; and does 
 usually rain till a strong north wind clears the air." 
 
 "A warm wind has a cold tail." And finally, 
 
 " Do business with a man when the \vind is in the north- 
 west." 
 
 9. Weather in Different Places : Synoptic Charts. Now 
 the fact that we have been able to describe every stage in 
 the weather-changes of the two days we have examined by 
 means of different proverbs and sayings, many of which are 
 to be found in the writings of Virgil, Pliny, or Aratus, and 
 most of which are current at the present day in numberless 
 languages and dialects, shows that the sequence observed at 
 Holyhead at that time has been a common and frequent ex- 
 perience over large parts of the globe for indefinite time 
 past. It seems, therefore, that at least one class of weather- 
 changes follows certain broad general outlines which are in- 
 dependent of local conditions, and the inquiry naturally 
 suggests itself How are these successive changes arranged 
 with regard to place ? Hitherto we have taken a fixed point 
 and discussed the variations as time progressed, the next step 
 
FUNDAMENTAL FACTS. 13 
 
 is obviously to take a specified instant of time and observe 
 how the weather differs from one place to another. 
 
 The most convenient method of doing this is to mark the 
 positions of a number of stations on a sketch map, and to 
 fill in at each station notes of the weather observed at the 
 time to which the map refers. This introduces us to what has 
 proved the most efficient instrument of modern weather re- 
 search, the synoptic chart, which enables us to see at a glance 
 the general conditions of weather at any given instant over a 
 large tract of country. In order to simplify and condense 
 as far as possible the notes placed on the chart, it is con- 
 venient to have a notation or uniform system of abbrevia- 
 tions, the one usually adopted being that given below. 
 
 10. Notation of Charts. Wind is represented by an 
 arrow flying with it, thus f means S., ^- W., \ N., -^ E., 
 and so on. The force of the wind is indicated by the 
 number of barbs or feathers on the arrow, thus : 
 
 j light breeze ; f fresh breeze ; $ strong wind ; gale ; 
 signifies calm. 
 
 Temperature and moisture being usually given as numeri- 
 cal results of instrumental observations are omitted for the 
 present. The remaining elements ( 7) are described by 
 letters and symbols as follows : 
 
 b = . . . . blue sky. 
 
 c = .... detached opening clouds. 
 
 d = . . . drizzling rain. 
 
 f or *S - foo- 
 
 1 \JI vvvv 1U O* 
 
 g= .... dark, gloomy weather. 
 
 h or ^ = . . hail. 
 
 1 or < = . . lightning (without thunder). 
 
 m = . . . . mist. 
 
 o = .... overcast sky. 
 
 p = .... passing showers. 
 
 q= .... squally. 
 
14 METEOROLOGY. 
 
 r or = . . . rain, 
 
 sn or * = . . snow. 
 
 T = thunder, 
 
 u = .... ugly, threatening. 
 
 v= .... visibility, i.e. distant objects 
 appearing unusually clear. 
 
 w or O - dew. 
 
 The above notation, devised by Admiral Beaufort, has 
 long been in universal use in this country. The following 
 symbols have been added more recently, and are officially 
 recognised by the various European meteorological institu- 
 tions. 
 
 K . - Thunderstorm. / . . . Strong Wind. 
 
 A ... Soft Hail. . . . Solar Corona, 
 
 i i ... Hoar Frost. O Solar Halo. 
 
 V ... Silver Thaw. CD Lunar Corona. 
 
 . . . Glazed Fro-t. (1) . . . Lunar Halo. 
 
 4*- ... Snow Drift. ^-\ . . . Rainbo\v. 
 
 >. . . Ice Crystals. \^ . . . Aurora. 
 
 co ... Dust Haze. 
 
 Amount or intensity may be indicated by the exponents 
 o and 2, e.g. : 
 
 thin fo ' ; ^ 2 dense fog. 
 
 1 1. Application to Previous Example. On the morning of 
 June 29th, 1890, we find from observations at twenty-five 
 stations scattered over the British Isles, southerly winds in 
 the west and north of Ireland and Scotland, westerly in the 
 south of Ireland, and north-westerly over England, nowhere 
 stronger than a light breeze. The weather everywhere was 
 bright with passing clouds (b c in the notation), except in 
 the west and north of Ireland, where it was overcast or 
 showery (o p m). As we take this merely as a starting 
 
FUNDAMENTAL FACTS. 1 5 
 
 point, a chart is unnecessary. Now note the observations at 
 6 p.m. on the same day. The general direction of the arrows 
 at the different stations is from south-west in the southern 
 counties, south in the midlands, and a little easterly in the 
 north : this is more easily seen if we draw large heavy 
 arrows following the general trend at the stations amongst 
 which it passes, as in Fig. i. Looking next at the weather, 
 we find three sets of symbols, all the stations reporting 
 " b r," or "<?," or "r," and a study of these show that it is 
 possible to divide the country into three quite definite belts, 
 one where the sky is more or less covered by detached 
 clouds, one where it is overcast, and a third where rain is 
 actually falling (see the broken lines, Fig. i.), which are 
 divided by lines running nearly north and south, the cloudy 
 area lying to the east and the rainy to the west, with over- 
 cast sky between. 
 
 Fig. 2 shows the state of things at 8 o'clock next morn- 
 ing. The large arrows are now as it were doubled up, 
 showing that in the southern districts the wind has shifted 
 more to the westward, and in the northern more to the east, 
 while between the two little change has taken place. The 
 distribution of weather symbols is now very different : the 
 " cloudy " area is only represented by one station, Yarmouth, 
 having apparently passed eastwards into the North Sea, and 
 its former position is chiefly occupied by the " overcast " 
 area, which now covers the whole of Scotland and the north- 
 east of England. The rainy area has also travelled during 
 the night from the extreme west of Ireland, and has assumed 
 a triangular shape, covering the north-east of Ireland, Wales, 
 and the southern, south-eastern, eastern, midland, and north- 
 western counties of England. In the south-west of England 
 and Ireland the rain has been replaced by squally, cloudy 
 weather with passing showers (c q p\ 
 
 At 6 p.m. on June 3oth (Fig. 3), the heavy arrows have 
 become still mere bent, forming a nearly closed curve. 
 
16 
 
 METEOROLOGY. 
 
 Fig. i. 
 
 Fig. 3. 
 
FUNDAMENTAL FACTS. I/ 
 
 The wind in the southern districts is therefore west, in the 
 northern east, while between them lies an eastern district 
 over England, in which it is still southerly, and a western 
 over Ireland, where it has changed rapidly round to north. 
 The " overcast " area has now also disappeared, and only 
 the apex of the rain area is left, covering the north of 
 England and south of Scotland ; elsewhere the weather has 
 become showery and squally. Fig. 4 shows that at eight 
 next morning (July ist) winds were easterly and north- 
 easterly in the north, westerly and north-westerly in the 
 south, and northerly between, while the rain area was limited 
 to the north-east of England and south-east of Scotland. A 
 narrow overcast belt intervened between it and the squally, 
 showery area which now extended over all the southern and 
 western parts of the kingdom. 
 
 12. Winds Arranged in a Rotating System. Recurring 
 now to the heavy wind arrows of Figs. 2, 3, and 4, if we 
 draw perpendiculars at various points of these curves we shall 
 find that they all intersect within a small central area around 
 which the winds everywhere circulate in a kind of vortex, 
 the direction in which it turns being against the sun or the 
 hands of a watch. Hence it appears that the motion is 
 analogous to that of a wheel revolving on its axis there 
 being no rotation at the axis, and the rate of motion, or the 
 force of the wind, depending on the speed at which it moves 
 round the centre. 
 
 13. Motion of Centre. Having found the central area 
 of the system for Figs. 2, 3, and 4 separately, compare its 
 position at different times. At 8 a.m. on June 3oth it 
 lay west and north from Erris Head, on the coast of 
 County Mayo ; at 6 p.m. off the mouth of the Mersey, and 
 at 8 a.m. on July ist over Lincolnshire. Joining these 
 points we get a line showing the path of the central area, 
 and therefore of the whole system for 24 hours, approxi- 
 mately a distance of 400 miles in an east-south-easterly 
 
 n 
 
IS METEOROLOGY. 
 
 direction, an average of nearly 17 miles an hour, which, be 
 it observed, has nothing to do with the force of the wind. 
 
 14. Direction of Wind in Different parts of the System. 
 No\v, given the rotating system and the path of its centre, 
 it is evident that the direction of the wind at any station can 
 be found by considering simply its position with regard to 
 that centre. Let A B (Fig. 5) represent the path of the 
 system, O the centre at a given instant. Draw C D through 
 O at right angles to A B. Then we have obviously three 
 cases, according as the point of observation, P, lies on the 
 path A B, or to the right, or to the left of it. Suppose a 
 northerly or westerly wind blowing at P, as the line through 
 the centre C D approaches, the wind will always back to some 
 point of south; on the line A B it will remain steadily S.S.W. 
 
 until the centre actually arrives, 
 then after a short period of calms 
 and variable winds it will sud- 
 denly shift to N.N.E., exactly the 
 opposite direction, and again re- 
 main steady ; to the left of A B 
 it will blow from a south-easterly 
 direction, shifting to east, and 
 after the line C D has passed, to 
 
 N.E. ; and to the right of A B, it will shift from S.W. to W. 
 
 and then to N.W. 
 
 The student must satisfy himself that this explanation 
 
 really satisfies the observations charted on Figs. 2, 3, and 4 ; 
 
 and should then study the changes of wind which would 
 
 occur in the three cases if the direction of A B were N.E., 
 
 E., S.E., etc. 
 
 15. Weather in Different Parts of the System. Having 
 ascertained the general system of wind circulation and its 
 movements as a whole, let us now look how the belts of 
 weather are placed with reference to it. In Fig. i the central 
 area of the system is evidently still far to the west, although 
 
FUNDAMENTAL FACTS. 19 
 
 the general change of wind shows that the British Isles are 
 well within the outer circle. Hence it appears that the 
 "cloudy," "overcast," and " rainy " belts are arranged in 
 order in advance of the central area, a fact confirmed by 
 Figs. 2 and 3, which show that these belts advance along 
 with the wind system, and keep invariably in front of a 
 position corresponding to the line C D in Fig. 5. We note 
 further that the belt of squalls and showers lies behind the 
 line C D for the most part, only advancing in front of it at 
 the extreme outskirts of the system, and never approaching 
 the line of advance. 
 
 Now observe specially that these belts are not placed 
 symmetrically with respect to the path of the wind system. 
 We know that the path at 8 a.m. on June 3oth was E.S.E., 
 and at that time the rain area lies chiefly in the "right 
 front " quadrant (DOB, Fig. 5), while the overcast area is 
 mostly in the "left front" quadrant. At 6 p.m. again, the rain 
 area has moved considerably towards the " left front " quad- 
 rant, and the overcast area has disappeared ; while at the 
 same time the squally area has advanced relatively on the 
 right. At eight next morning the change has progressed 
 still further, the overcast area has reappeared, and it now 
 lies along with the rain area, in the " left front " quadrant, 
 the squally area having meantime spread into the " right 
 front." But this involves one of two things, either all stations 
 in the right front, after being warned by halos, clouds, and 
 the like, will experience squalls and passing showers, and all 
 those in the left rear after receiving the promised rain will 
 return into damp, muggy weather ; or, the whole system 
 must have changed front by altering the direction of its 
 path. The observations of July 2nd show that the latter 
 event actually occurred, and at 8 a.m. the central area lay 
 over the Cattegat, thus restoring the original arrangement of 
 all the weather belts with regard to the different quadrants 
 of the wind circulation. 
 
20 METEOROLOGY. 
 
 We have therefore been able, with the help of synoptic 
 charts, to develop from a particular case a system capable 
 of accounting for the observed facts ; and since these facts 
 have been shown to agree with common experience, we may 
 for the present accept it as a typical weather system, and 
 summarise its principal characteristics as follows : 
 
 (1) A central calm area, around which the wind circu- 
 lates in a direction opposite to that in which the hands of a 
 watch move, the whole system progressing in a definite 
 direction. 
 
 (2) A belt of more or less cloudy weather, in which 
 halos, coronae, etc., are observed (considerably in advance 
 of the central area). 
 
 (3) A belt still in advance of the centre, but nearer to it 
 than (2), in which the sky is overcast, and the atmosphere 
 close, warm, and muggy. 
 
 (4) An area, generally triangular or oval, of steady con- 
 tinuous rain, the whole of which lies in advance of the 
 centre, and the greater part to the right of its path. 
 
 (5) An area of bright, dry atmosphere, with squally winds 
 and occasional showers, lying entirely in rear of the central 
 area. 
 
 In the case discussed we have seen that the path of the 
 system while approaching these islands was nearly E.S.E. 
 Compare the saying 
 
 " A rainbow at night is the shepherd's delight ; 
 A rainbow in the morning is the shepherd's warning." 
 
 Now a rainbow can only be observed when the sun is 
 shining, and when rain is falling in the quarter of the sky 
 opposite to the sun ; hence a rainbow in the morning requires 
 thit there should be clear weather to the east and foul to the 
 west, and one in the evening clear to the west and foul to 
 east ; from which, we may argue that in the former the rain 
 
FUNDAMENTAL FACTS. 21 
 
 is approaching, and, in the latter, passing away, i.e. that an 
 easterly course is more usual than a westerly for the system 
 we have described ; a supposition further confirmed by 
 
 " Sun set in a clear, 
 Easterly wind near ; 
 Sun set in a bank, 
 Westerly will not lack." 
 Hence we add 
 
 (6) The path of the system as a whole is usually to- 
 wards some point of east, at least in approaching these 
 islands. 
 
 From (i) we may deduce at once the following practical 
 rules : 
 
 (1) Stand with your back to the wind, and the central area 
 lies away to your left. 
 
 (2) If the wind blows steadily for a long time without 
 shifting a point, expect a sudden calm and then squalls 
 from exactly the opposite direction. 
 
 (3) If the wind shifts steadily but slowly, you are on 
 the outskirts of the system ; if quickly, you are near the 
 path of the centre. ' 
 
 These rules should be observed with a due regard to the 
 other peculiarities of the system. 
 
 16. Intensity of System not accounted for. Having stated 
 a few empirical results and suggested a few practical 
 applications, the question naturally arises, why should 
 this arbitrary and apparently rather uncertain relation 
 amongst the elements into which we have analysed the 
 weather exist ? And again, how is it that we have been able 
 to develop a definite system, by employing a considerable 
 proportion of the known weather sayings and prognostics, 
 out of a particular case which ended in nothing more serious 
 than a moderate night's rain and a fresh sailing breeze ? 
 How are we to distinguish such a system from one bringing 
 
22 METEOROLOGY. 
 
 a flood or a violent hurricane ? Yet again, we have seen that 
 in changing its course, the system we have considered 
 brought about a failure of many of what seemed the most 
 reliable signs, no rain or wind ensuing in some quarters, 
 while others received a double allowance. How are we to 
 provide against such failures ? 
 
 This somewhat formidable list of questions suggests that 
 there is yet another element underlying those already ap- 
 parent j one which does not appeal directly to the unaided 
 sense, and must therefore be sought for with the aid of 
 instruments. 
 
 But where is the other element to be looked for ? On 
 reviewing the whole evidence, it appears that the most stable 
 and definite part of our weather system is that of the wind 
 circulation. Hence we inquire, why does this rotational 
 motion arise, and how is it set up ? Which reminds us that 
 we do not yet know why the wind blows at all. 
 
CHAPTER II. 
 
 FUNDAMENTAL PRINCIPLES. 
 
 17. Motion of the Atmosphere. We have defined wind as 
 motion of the atmosphere, and may restate the question, 
 "Why does the wind blow?" in the form, "What are the 
 forces which set the atmosphere in motion, and how are they 
 applied ? " We shall in this chapter remind the reader of a 
 few physical facts and principles required for our special 
 purposes, referring him to treatises on mechanics and heat 
 for evidence and formal proofs. 
 
 1 8. Physical Constitution of the Atmosphere. The atmo- 
 sphere consists of two quite distinct parts, or rather the 
 earth is enveloped in two separate atmospheres one of air 
 and one of steam or aqueous vapour. 
 
 19. Air. Air is a true gas, and consists of a mechanical 
 mixture of about 21 parts of oxygen and 79 of nitrogen in 
 100 by volume, with a small quantity of carbonic acid. 
 Since its composition is practically constant throughout the 
 atmosphere, it need not be further inquired into. Now the 
 distinguishing property of all gases is their power of inde- 
 finite expansion. If we introduce a small quantity of air 
 into an empty vessel, however large, it will expand until the 
 vessel is completely filled, and having filled, it will press 
 upon the containing walls, trying, as it were, to expand still 
 further. 
 
 The amount of this expansive force ox pressure has been 
 found for a given quantity of air to depend on the volume 
 and the temperature, and the relations of these three quan- 
 
 2 3 
 
24 METEOROLOGY. 
 
 titles to each other are expressed in two laws, known by the 
 names of Boyle and Charles. 
 
 Boyle's Law. The volume of a given mass of gas, kept at 
 a constant temperature, is inversely as the pressure. 
 
 Charles' Law. The volume of a given mass of gas, kept 
 at a constant pressure, increases by a definite fraction of 
 its amount for a given rise of temperature. 
 
 The density of a gas varies directly as the pressure and in- 
 versely as the volume. Regnault found the density of air 
 at temperature o C. and pressure 760 mm., to be i'293 
 grammes per litre. 
 
 20. The Barometer. Since pressure is a force, it is most 
 conveniently measured in terms of a weight. The barometer 
 is simply a " weighing machine " adapted to this purpose. 
 A glass tube, closed at one end, is filled with mercury and 
 inverted in a vessel also containing mercury. If the 
 instrument is now placed in the gas whose 
 pressure is to be measured, mercury flows out 
 of the tube until a certain point B (Fig. 6) is 
 reached. As the part of the tube B C 
 was originally full of mercury, and nothing 
 has obtained access from without, the space 
 must now be a vacuum, and no pressure can 
 be exerted upon the mercury surface at B. 
 Hence the column A B must just counter- 
 balance the external pressure on the mercury 
 in the vessel A. 
 
 Fig. 6. Now if we define a fluid as a body which 
 
 offers no resistance to change of shape (thereby including 
 both liquids and gases), the following are necessary condi- 
 tions of its remaining at rest : 
 
 (i) The pressure exerted by any portion of the fluid is per- 
 pendicular to the surface of that portion. It follows that we 
 may define the pressure at any point of a fluid as the average 
 pressure per unit of area on a plane surface containing the 
 
FUNDAMENTAL PRINCIPLES. 2$ 
 
 point, and perpendicular to the pressure, when the area is 
 supposed infinitely small. 
 
 (2) In a fluid not acted upon by external forces, the pres- 
 sure is the same at all points and in all directions. 
 
 (3) When the fluid is acted upon by an external force such 
 as gravity 
 
 (a) The pressures in the same horizontal plane are equal. 
 
 (b) The pressures at any tivo points differ by the weight of 
 a cylinder of the fluid of unit sectional area, whose length is 
 equal to the perpendicular distance between them. 
 
 Applying these laws to the mercury at rest in the baro- 
 meter, we learn that the pressure at the horizontal surface 
 A is the same inside and outside the tube, and that the 
 difference of pressure at the levels A and B in the tube is 
 equal to the weight of a cylinder of mercury of unit sectional 
 area, whose length is the vertical distance A B. But since 
 there is no pressure in the upper part of the tube B C, this 
 difference is the pressure we desire to measure. We shall 
 in general express pressures in terms of the height of a 
 column of mercury balanced by them, but in some inquiries 
 it is often convenient to use the average pressure of the at- 
 mosphere at the earth's surface as a unit. This unit, called 
 an atmosphere, is represented by a mercurial column 760 mm. 
 in height at temperature o C. Now the mass of i cub. cent, 
 of mercury is 13*596 grammes, and therefore that of a column 
 of unit sectional area 76. cm. long is 1033.3 grammes; 
 hence the pressure of one atmosphere is a force of io33'3 
 grammes weight per square cm. = 14-7 Ibs per sq. in., 
 nearly. 
 
 21. Specific Heat of Air. The specific heat of a substance 
 is the number of units of heat required to raise the tem- 
 perature of unit mass of it i C. In warming a quantity 
 of air we may either keep the volume constant and allow 
 the pressure to increase, or allow the volume to increase so 
 that the pressure remains constant. In the latter case, it is 
 
26 METEOROLOGY. 
 
 evident that more heat will be required, because the air has 
 not only to be warmed i, but has to be expanded a given 
 fraction of its volume. Hence we have two specific heats 
 for air; that at constant pressure (Cp) is 0*2375, and at 
 constant volume (Cv) 0-1684, whence g = 1-41. 
 
 This consideration involves the idea of a relation between 
 heat and mechanical work. Joule found that in order to 
 raise the temperature of unit mass of water i C. it was neces- 
 sary to expend 424 g units of work on it, and conversely, in 
 order to get a unit of work out of it -^- units of heat 
 must be abstracted. Suppose we expend iv units of work 
 upon a quantity of air kept at constant pressure, this is 
 equivalent to communicating ( 4 ~ units of heat to it, 
 which would raise its temperature ( ~ x gJj~ )so*oio w 
 degrees. On the other hand, if thel eat in a quantity of air 
 does w units of work in expanding, it will be cooled o'oio w 
 degrees. 
 
 A particular case of the dynamical heating and cooling of 
 the air just explained, which has special interest in theoretical 
 meteorology, is that of an ascending or descending current of 
 air, supposed to move with such rapidity that no heat is ex- 
 changed with the atmosphere immediately surrounding. 
 
 Let a kilogramme of air ascend through a vertical distance 
 of h metres, then the work done against gravity is g h units 
 equivalent to ~ g h units of heat, which must be obtained 
 from the air itself. If / and t- are the initial and final 
 temperatures, by what precedes 
 
 ^ =Cpx(/-/ 1 ) 
 or, /-/! = 0*00993 h. 
 
 In the case of a descending current the work is, of course, 
 done by gravity. Hence we may say that, provided no heat 
 is lost or gained on the way, dry air is cooled or warmed i 
 C. for every 100 metres of ascent or descent. 
 
 22. Aqueous Vapour. If water be introduced drop by 
 
FUNDAMENTAL PRINCIPLES. 2/ 
 
 drop into a perfectly exhausted receiver, it will be found that 
 the drops are evaporated as they fall, until a certain pressure 
 of vapour is reached, after which they retain the liquid con- 
 dition. The space inside the receiver is then said to con- 
 tain saturated vapour, and experiment has shown that the 
 pressure of saturation depends only on the temperature. If 
 the experiment be performed at the same temperature in a 
 receiver containing dry air at any pressure, exactly the same 
 amount of water will be taken up, only the process will 
 take longer, because the vapour has to make its way through 
 the air by diffusion. 
 
 The maximum or saturation pressures of vapour at 
 various temperatures were fully determined by Regnaulr, 
 and were not found to follow a simple law. Empirical 
 formulae may be devised which fairly represent the connec- 
 tion between temperature and the pressure of saturated 
 vapour, but it is better, on the whole, to employ a table. 
 
 23. Non-saturated Vapour. When the amount of vapour 
 present in a given volume of air is not sufficient to reach the 
 pressure of saturation corresponding to its temperature, the 
 air is then partially dry. Non saturated vapour behaves in 
 much the same way as a true gas, and closely follows to 
 the laws of Boyle and Charles. 
 
 24. Saturated Vapour : The Dew Point. The tempera- 
 ture at which any given quantity of vapour becomes saturated 
 is called the dew-point, and if the temperature fall below 
 this point, part of the vapour is condensed into the liquid 
 form, until the pressure is reduced to the saturation-pressure 
 corresponding to the new temperature. In like manner, if 
 the temperature is again raised, water is evaporated and the 
 state of saturation continued as long as the supply lasts, 
 after which the vapour becomes dry, there being no longer 
 enough to reach saturation pressure. Hence a method of 
 determining the amount of vapour present at any time is to 
 cool a body having a bright surface in it until condensation 
 
28 METEOROLOGY. 
 
 begins, which is shown by the formation of dew on the 
 bright surface. The temperature at the surface is then that 
 corresponding to saturation for the pressure of vapour at the 
 time. 
 
 25. Latent Heat. In order to effect the change from 
 the liquid to the vaporous form, a certain quantity of heat 
 is required, as is obvious if we allow damp clothes to dry 
 on the body the heat required for the work of drying 
 being abstracted so rapidly from the body that most people 
 suffer a severe chill and on the other hand condensation 
 liberates an equivalent amount of heat. The heat required 
 to change the state of a substance is distinguished from that 
 which goes to change its temperature by being called latent 
 heat, and the quantity we are discussing is called the latent 
 heat of evaporation. 
 
 26. Supersaturated Vapour; Condensation Nuclei. We 
 have thus far dealt with stable conditions, but Aitken has shown 
 that in perfectly pure air an unstable state may be realised, 
 in which a pressure of vapour greater than that correspond- 
 ing to the temperature of saturation is maintained. In fact, 
 condensation will not in general begin unless some nucleus 
 is present to which the particles of water can attach them- 
 selves. 
 
 In the atmosphere the required nuclei are supplied by 
 the impalpable dust which we know to be present even at 
 great elevations, and it is not probable that supersaturated 
 vapour, or vapour in the unstable state, occurs to any extent 
 in nature, unless, as suggested by Von Bezcld, in the case 
 of thunderstorms or the phenomena known as cloud-bursts, 
 where sudden falls of very heavy rain are experienced. 
 These may be due to sudden condensation of supersaturated 
 vapour in a dust-free region of the atmosphere, and the 
 rise of pressure accompanying such precipitation would 
 seem to give at least plausibility to the supposition ; but in 
 the majority of cases the tendency is probably the other 
 
FUNDAMENTAL PRINCIPLES. 29 
 
 way, for it may happen that the temperature of a number of 
 dust-particles becomes. lower than that of the air in which 
 it is floating, and vapour is condensed upon them although 
 the air itself is not quite saturated. Much also depends on 
 the amount of dust present, and on the size and constitution 
 of the particles. 
 
 27. Dry Fog. As water condenses on the particles of 
 dust, their size is soon so far increased that they become 
 visible. When the quantity of vapour condensed is small 
 and the number of dust-particles large, or their size con- 
 siderable, it is obvious that the allowance of vapour particles 
 for each particle of dust must be small, and we have the 
 phenomenon known as dry fog. The dry fogs peculiar to 
 hrge towns are due not so much to deficient vapour as to 
 the excessive size and quantity of the particles, which, as we 
 know, are often sufficiently visible as smoke without the 
 help of condensation. On the other hand, sea-fogs pro- 
 bably occur when there is considerable dryness, because the 
 dust consists largely of salt particles derived from spray or 
 surf, which greedily absorb moisture. 
 
 28. Mist. When the quantity of condensed vapour 
 becomes larger or the allowance of dust smaller, each dust 
 particle receives more vapour, and a cloud of wet mist is 
 formed, in which the particles are coarser and heavier than 
 in the fog. After a certain point the dust particles are 
 unable to retain all the moisture condensed upon them, 
 and an unsteady condition sets in, giving rise to inequalities 
 in the distribution of the moisture, which ends in the 
 formation of rain-drops. The details of this process are 
 not, however, fully understood. 
 
 29. Radiation. It is now abundantly evident that the 
 changes occurring in the various meteorological elements are 
 ultimately dependent on changes of temperature. If the 
 atmosphere were freed from temperature disturbances it 
 would shortly become and would remain entirely stagnant. 
 
30 METEOROLOGY. 
 
 Now we know that the amount of heat derived from the 
 interior of the earth, although measurable in quantity, is 
 insignificant in comparison with the radiant energy received 
 from the sun, and it therefore becomes important to know 
 what is the behaviour of the atmosphere with regard to 
 radiation. 
 
 30. The Sun's Rays. The sun sends forth rays of widely 
 diversified wave lengths; and of each particular wave 
 length in different quantities. It is convenient for most pur- 
 poses to classify the rays according to their effects on bodies 
 upon which they impinge ; thus a certain range of wave 
 lengths affects the retina of the eye. and gives rise to sen- 
 sations of light and colour the shorter waves to blue and 
 violet, and the longer to orange and red, with intermediate 
 greens and yellows ; another series is remarkable for produc- 
 ing chemical changes in certain substances these include 
 the " photographic " rays, and are in general of short wave 
 length, blue and violet, with others (ultra-violet) too short to 
 be visible ; and a third series is sensible to us as radiant 
 heat these chiefly of longer wave lengths, yellow and red, 
 with " dark-heat " rays (ultra-red) too long lo affect the 
 retina. Of all rays reaching the earth's surface, those have 
 the maximum intensity which give the eye the impression of 
 yellow ; yellow rays are both the brightest and the hottest. 
 
 31. Effect of the Atmosphere: Absorption. If there were 
 no atmosphere the amount of heat and light received at any 
 part of the earth's surface would depend solely on the 
 altitude of the sun, and the length of time it remained above 
 the horizon, and the same would hold good for all the 
 celestial bodies. The sun, moon, and planets would appear 
 as luminous discs and the stars as points of light without 
 any twinkling, while the rest of the sky would be perfectly 
 dark. All places exposed to the sun's direct rays would be 
 illuminated with dazzling brilliancy, and the heat would be 
 intolerable, while in the shade there would be almost total 
 
FUNDAMENTAL PRINCIPLES. 3! 
 
 darkness, and great cold. On the other hand, while the 
 sun was below the horizon, the earth would freely part with 
 its heat by radiation into space, and the temperature would 
 fall far below the freezing-point every night. 
 
 The atmosphere then acts as a kind of covering to the 
 earth, just as clothes do to the body, tempering the fierce- 
 ness of the sun's rays by day, and preventing excessive loss 
 of heat by night. When the mingled waves of heat, light, 
 and chemical action from the sun reach the atmosphere, a 
 certain proportion of each is absorbed, reflected, and scattered 
 in all directions and the remainder transmitted directly. 
 The relation between the amounts absorbed and transmitted 
 depends upon what we call transparency in the case of light 
 rays, or diathermancy in the case of heat rays. According 
 to S. P. Langley 1 at least 40 per cent, of the radiant energy 
 received from the sun is absorbed by the atmosphere ; and 
 this fraction diffused in all directions illuminates the sky, 
 makes objects in shadow visible, and extending into regions 
 beyond the sun's direct rays, gives rise to the phenomena of 
 twilight. 
 
 But the proportions absorbed and transmitted are very 
 different for different wave lengths ; the atmosphere is much 
 more transparent to red rays than to blue and violet. On a 
 very clear day, when a comparatively large proportion of all 
 rays are transmitted, the atmosphere is scarcely able to 
 arrest any but blue rays, and the sky therefore has a very 
 intense blue colour ; while in hazy weather not only blue 
 but green and yellow rays are stopped, the sky has a sickly 
 yellowish colour, and at the same time the sun appears red, 
 the only light rays able to penetrate being of that colour. 
 It is easy to see that when the sun is in the zenith the rays 
 from it traverse the shortest possible (i.e. the vertical) path 
 through the atmosphere, and that, as it sinks towards the 
 horizon that path becomes longer and longer, so that there 
 il. Mag. 1884. II. p. 289. 
 
32 METEOROLOGY. 
 
 must be more and more absorption. Hence the sun's disc 
 most frequently appears red just after sunrise or just before 
 sunset. 
 
 Little is as yet known with certainty about the nature of 
 this process of absorption. It is believed that pure dry air 
 is almost quite transparent to all the sun's rays, and it seems 
 likely that dry unsaturated vapour has little absorptive 
 power. The investigations of Crova go far to show that the 
 absorbing element is in the first place the dust. Aitken 
 found that dust particles exposed to the action of heat rays 
 experienced very slight changes of temperature provided 
 they were small enough so that a cloud of dust was not 
 warmed by sunshine unless the particles of which it was 
 composed were comparatively large. Again, Crova has 
 been able to photograph invisible clouds clouds which 
 emitted rays too short in wave-length to be sensible to the eye, 
 but which nevertheless produced chemical effects. These 
 facts suggest that the excessively minute particles of dust 
 are able to reflect and scatter the chemical and violet rays, but 
 that it is only when their size is increased by the condensa- 
 tion of moisture upon them that they are able to arrest the 
 yellow and red and the heat rays. When the particles be- 
 come large enough to be visible as dry fog, we have seen 
 that only red and heat rays are transmitted and when they 
 are sufficiently large to form cloud, radiation is almost en- 
 tirely stopped. It must be remarked, however, that the pro- 
 portion of rays transmitted does not increase regularly with 
 the wave-length, certain selected rays of special wave-lengths 
 being almost entirely stopped ; these give rise to dark 
 " telluric " bands at different parts of the solar spectrum. 
 
 Crova has found that in summer the intensity of solar 
 radiation increases steadily for some hours after sunrise, 
 then suddenly becomes very irregular, changing rapidly from 
 minute to minute, decreasing on the whole till some time 
 after mid-day, when it again increases, and, after reaching its 
 
FUNDAMENTAL PRINCIPLES. 33 
 
 maximum, diminishes till sunset. He suggests that the un- 
 steadiness begins when the sun's warmth becomes sufficient 
 to cause a considerable increase in the evaporation of water 
 from the soil, which condenses on the dust particles in the 
 air, increasing their size and causing a haze which checks 
 the intensity at mid-day. As the heating process continues, 
 the water is evaporated from the dust particles, again reduc- 
 ing their size and allowing freer transmission. In winter, 
 when the quantity of vapour is small, the curve of solar in- 
 tensity shows only a single maximum at noon. 
 
 We thus see that if the air is dry and clear, compara- 
 tively little of the sun's heat is stopped on its way through 
 the atmosphere, and most of the changes of temperature are 
 brought about by heat communicated from the earth's 
 surface in a manner we shall presently discuss. But a 
 dampness insufficient to form visible haze may nevertheless 
 increase the size of the dust particles sufficiently to cause 
 considerable absorption and radiation of heat in the atmo- 
 sphere, and consequent variations of temperature. We may 
 thus account for the fact that the variation of temperature in 
 the air over the ocean is some four times as great as in the 
 surface water of the ocean itself. 1 
 
 Hence also we find that temperature changes as we ascend 
 in the atmosphere at a rate depending chiefly on the heat 
 received from that surface. The usual condition is such 
 that temperature falls as the elevation increases at an average 
 rate of i C. per 163 metres. This gradient is, however, 
 liable to very great variation \ it may be steeper or less steep, 
 or it may be reversed, as when a layer of cold air lies next 
 the ground. In any case the distribution of temperature 
 becomes more uniform at great heights. Observe that we 
 are not here dealing with ascending or descending currents. 
 
 32. Transmission of Rays. The transmitted rays may be 
 
 Challenger Reports. "Atmospheric Circulation," p. 7- 
 
 C 
 
34 METEOROLOGY. 
 
 considered separately according as they fall upon the sea or 
 upon land. 
 
 In the former case the rays are partly reflected, partly 
 absorbed near the surface, and partly transmitted through 
 the water. The observations made during the cruise of 
 H.M.S. Challenger show that some rays penetrate to a depth 
 of 500 feet below the surface. On account of this trans- 
 mission, and also because of the high specific heat of water, 
 the changes of temperature over the ocean are small, and 
 take place slowly. The " daily range " of air temperature 
 the difference between the hottest part of the day and 
 the coldest part of the night and the " annual range " are 
 small, the nights and winters are warm, and the days and 
 summers cool compared with those on land. 
 
 The heat received on a land surface is also in part 
 reflected, but the remainder is wholly arrested by the thin 
 exposed layer, which is quite opaque, and on account of its 
 low specific heat becomes much heated : the rise of tempera- 
 ture depends of course largely on the nature of the surface 
 if the substance composing the surface stratum is a good 
 conductor, the heat is quickly transferred to a lower stratum 
 and the warming of the surface checked ; if it is a bad 
 conductor, the heat is retained at the surface, which may 
 then attain a very high temperature. Dense heavy clay soils 
 are much better conductors than loose sandy ones, because 
 the latter imprison large quantities of air, which is a bad 
 conductor ; hence extremes of heat and cold penetrate to a 
 much greater depth in clay than in sand. Vegetation also 
 acts as a protective covering to the soil, and its temperature 
 never becomes very high, partly because the heat is used up 
 in evaporating moisture from the plants, and partly because 
 they present a large surface to the air, which when heated 
 moves away and is replaced by colder air. 1 The changes 
 of temperature are therefore more rapid and of greater range 
 1 Buchan : Handy Book of Meteorology > p, 83. 
 
FUNDAMENTAL PRINCIPLES. 35 
 
 over the land than over the sea least over clay soils or re- 
 gions covered with vegetation, and excessive over arid sandy 
 deserts and bare rocks. 
 
 33. Terrestrial Radiation. Since, on the whole, the 
 temperature at the earth's surface from year to year is the 
 same, it is evident that an amount of heat exactly equivalent 
 to that received must be lost by radiation into space ; and 
 since the heat radiated back into space has to pass through 
 the atmosphere., in the same way as that received, it is evi- 
 dently subject to .the same conditions with regard to trans- 
 parency, When the atmosphere is clear and dry, nocturnal 
 radiation goes on freely, and the surface of the earth is 
 rapidly cooled just as by day it is rapidly heated ; and the 
 same conditions of haze or cloud which cut off the sun's rays 
 by day prevent loss of heat by night. Again, surfaces 
 which absorb heat freely radiate freely, and hence the nights 
 are relatively warmer over the sea than over the land, and 
 over clay soils than over sandy. 
 
 34. Air Temperature : Dew. We must now look a little 
 more closely into the manner in which the changes of tem- 
 perature at the earth's surface are communicated to the 
 atmosphere. Air, especially if it be dry, is a very bad con- 
 ductor of heat, and therefore if the atmosphere remained 
 motionless the thin layer immediately in contact with the earth 
 would quickly take the temperature of its surface. This is 
 very much what happens on a calm, clear night when the soil 
 is cooled by radiation; the stratum of air lying next the surface 
 is cooled by contact, sometimes to very low temperatures, 
 and as it becomes colder it becomes heavier and has no 
 tendency to move away if the surface be level. When the 
 temperature falls below a certain point the vapour rising 
 from warm under-layers through the pores of the soil or 
 amongst the blades of grass is condensed on this thin sur- 
 face layer, either as water, when it forms dew, or as ice when 
 it is called hoar-frost. Hence we see why dew and hoar- 
 
36 METEOROLOGY. 
 
 frost are associated not with cloudy, damp weather, but 
 with clear skies and little wind, when there is free radiation. 
 Even after prolonged dry weather it is found that the ground 
 exudes moisture to form dew on the tips of grass blades and 
 the under sides of stones ; but we must be careful to distin- 
 guish between the pearly deposit which is the genuine dew, 
 and the " sparkling dew-drop" hanging on the leaves, which 
 is not dew at all, but water expired by the leaves themselves. 1 
 
 During the day, when the earth's surface is warmed by 
 the sun's rays, the action is somewhat different. Part of 
 the heat is propagated from the surface downwards by con- 
 duction as a wave, at a rate depending, as we have seen, on 
 the conductivity of the soil, the daily variation of tempera- 
 ture being sensible to a depth of about four feet. The heat 
 not drained off by conduction warms the thin stratum of air 
 lying just in contact with the surface but this air being 
 now warmer than the layer above it, is lighter, and, rising, is 
 replaced by more cold air, which in its turn is warmed, rises, 
 and is replaced. Hence, " convection currents " are set 
 up, and the atmosphere is warmed to a considerable ele- 
 vation. 
 
 In what precedes, we have for the sake of clearness sup- 
 posed the air to be perfectly clear and dry or rather have 
 neglected the heat absorbed in transmission. Since a body 
 radiates heat just in proportion as it absorbs it, it is evident 
 that a damp, dusty layer of air behaves much in the same 
 way as a water surface, so that we have the two effects as it 
 were superposed ; the temperature of the damp air is 
 changed by the heat absorbed or radiated by it, and also by 
 that from the earth's surface, which has already been trans- 
 mitted through it. 
 
 35. Diurnal Variation : Temperature. This prepares us 
 for a daily variation of the physical conditions of the atmo- 
 sphere in the first place, of the temperature, and through 
 Aitken : Trans., A'.S.I?., vol. xxxiii., p. 29. 
 
FUNDAMENTAL PRINCIPLES. 3? 
 
 the temperature of the pressure, humidity, and other ele- 
 ments indirectly, depending on the intensity of the sun's 
 rays, and on the length of the day. It is usual to discuss 
 these variations with reference to their daily mean the 
 theoretical value of which is the average of all the values 
 for every instant during the day although for practical pur- 
 poses the average of twenty-four hourly values is sufficiently 
 accurate. 
 
 The daily curve of temperature, although varying 
 considerably in different parts of the world, is in 
 general of a simple form. Over the open sea, in the 
 North Atlantic, the Challenger^ observations showed that 
 the minimum occurs about 5 a.m. and the maximum 
 about 2 p.m., and this may be taken as fairly repre- 
 sentative. Near the equator, where the days and nights 
 are nearly equal throughout the year, there is little difference 
 in the daily curve at different seasons ; thus Batavia gives 
 minimum 5.50 a.m., maximum 1.20 p.m. In higher lati- 
 tudes, where the days vary greatly in length, the hours at 
 which these phases occur differ considerably. During 
 summer the maximum usually occurs between 2 and 4 p.m., 
 unless, as in some places, sea-winds tend to increase cloudi- 
 ness towards the afternoon, when the maximum occurs 
 earlier; and during winter the maximum is almost always 
 about an hour earlier than in summer. The minimum dur- 
 ing summer occurs just at dawn, and, therefore, earlier in 
 high latitudes than in low. The first rays of the sun 
 striking the upper strata of the atmosphere are reflected to 
 the ground and absorbed by it, so that the air becomes 
 slightly warmed before sunrise. When the ground is 
 covered wit.h snow, there is very little absorption of the re- 
 flected rays, and temperature does not begin to rise till the 
 sun is above the horizon. 
 
 Sir David Brewster showed that in general if the tempera- 
 1 Challenger Reports, " Atmospheric Circulation," p. 7. 
 
38 METEOROLOGY. 
 
 ture is above the average for the day at any given hour, it 
 will be nearly as much below it twelve hours later. We 
 may therefore obtain a close approximation to the true mean 
 temperature for the day by taking the average of observa- 
 tions at two hours of the same name, as 3 a.m. and 3 p.m., 
 or 10 a.m. and 10 p.m. The best results are obtained at 
 hours when the temperature is not much different from the 
 actual average ; 9 a. m. and 9 p.m. are those usually adopted 
 in this country. 
 
 Another method of determining the mean temperature is 
 from observations of self-registering thermometers, which re- 
 cord the highest and lowest extremes for the day ; the mean 
 of these is usually about half a degree higher than the 
 average of twenty-four hourly observations. The reading of 
 the maximum thermometer minus that of the minimum 
 gives us another extremely important element of climate, the 
 daily range of temperature. This depends chiefly upon 
 proximity to the sea ; in mid-Atlantic it amounts to about 
 i*8 C., increasing to 2'4 near land; while in dry, continental 
 climates it may average 15 or more. 
 
 36. Diurnal Changes of Pressure. The diurnal oscilla- 
 tions of atmospheric pressure show in general two maxima 
 and two minima, the former at 10 a.m. and 10 p.m., and 
 the latter at .4 a.m. and 4 p.m. These four phases are most 
 sharply defined and of greatest amplitude in low latitudes 
 and in damp climates. At stations near the sea. especially on 
 the west coasts of continents and islands, it is found further, 
 that in summer the morning maximum and the afternoon 
 minimum are delayed, and the former becomes unusually 
 large, while the latter is greatly diminished. At continental 
 stations, again, the morning maximum occurs early in the 
 day, and is much diminished, while the afternoon minimum, 
 although still delayed, is increased in amplitude. 
 
 The occurrence of this double wave over the ocean, the 
 surface temperature of which remains nearly constant 
 
FUNDAMENTAL PRINCIPLES. 39 
 
 during the day, proves that it is not caused by variations of 
 temperature of the earth's surface, but of the atmosphere 
 itself; it depends on the varying amounts of heat absorbed 
 and transmitted according to the size and quantity of the 
 atmospheric dust particles. 
 
 37. Seasonal Variations. The seasonal variations of the 
 meteorological elements are in many respects analogous to 
 the diurnal phenomena we have just described. The earth 
 gains heat in the daytime of summer, loses it during the 
 night of winter, and passes through the transition stages in 
 the morning of spring, and the evening of autumn. Just as 
 the daily changes of temperature are least near the equator 
 and greatest in higher latitudes, least over the sea and 
 greatest over continents, so in the tropics the seasons are 
 scarcely marked; in the temperate and frigid zones they 
 oscillate through a wide range ; over the sea winters are mild 
 and summers cool, and over the continents intense cold 
 alternates with burning heat. From the nature of the case 
 we do not find the double diurnal oscillation of pressure 
 reproduced in the seasons : the conditions are for the most 
 part determined by the temperature. In summer the land 
 is warmer than the sea, and pressure is accordingly lower 
 over the former ; while in winter the sea is warmer than 
 the land, and the distribution of pressure reversed. 
 
 38. Corrections for Daily and Monthly Range. We are 
 not at present concerned to pursue this subject further. 
 Our object so far has been to ascertain the normal con- 
 dition of things, to separate out the processes which are 
 always going on from hour to hour and day to day, for 
 diurnal and seasonal variations are always taking place 
 under all circumstances and conditions. We thus come to 
 regard not only the infinite small variations which are 
 constantly taking place from minute to minute throughout 
 the atmosphere, but all storms, moving systems such as 
 that described in Chapter I, and spells of fine weather and 
 
40 METEOROLOGY. 
 
 the like, as disturbances > occurring over and above the 
 normal march of events. Hence in discussing, for example, 
 a disturbance of temperature or pressure, we cannot com- 
 pare observations made at different hours of the day, or on 
 different dates, until they have been corrected for daily or 
 monthly range as the case may be, which is usually ac- 
 complished by reducing all to the mean for the day or 
 month. It may, of course, happen that the disturbance is 
 so great that the diurnal variation is a negligible quantity in 
 comparison with it ; in temperate regions the diurnal varia- 
 tion of pressure is small and comparatively unimportant for 
 many purposes, but in the tropics it is of greater amplitude 
 than most disturbances, and cannot be overlooked. 
 
 39. Position of Bodies on the Earth 's Surface: Absolute 
 Motion. Any body on the surface of the earth is subjected 
 to the action of two principal forces, that of gravitation or 
 the earth's attraction and the so-called centrifugal force due to 
 the earth's rotation on its axis. 
 
 Now, Newton showed that in the case of a sphere, the 
 attraction of gravity acts as if the whole mass were collected 
 at the centre, and this result holds good, to a close ap- 
 proximation, for the earth ; and again, we know that a 
 body revolving in a circle tends to increase its distance 
 from the centre. But the circles described by bodies at 
 rest on the earth's surface are circles of latitude, and there- 
 fore the centrifugal force will act in the direction of a line 
 drawn from the earth's axis of rotation perpendicular to it, 
 that is, it will act vertically upwards at the equator, horizon- 
 tally near the poles, and in some intermediate direction at 
 all intermediate latitudes. Hence, for all intermediate 
 latitudes we may resolve the centrifugal force into two parts 
 one vertically upwards, and another horizontal the latter 
 always tending towards the equator. The vertical com- 
 ponent is obviously counteracted by gravity but how 
 about the second ? What prevents a free body su.ch as the 
 
FUNDAMENTAL PRINCIPLES. 
 
 water in the sea from moving towards the equator ? Unless 
 we introduce some new force altogether, the answer must 
 clearly be that gravity does not act vertically and that 
 there must be a horizontal component sufficient, and just 
 sufficient^ to counteract the centrifugal force. But it was 
 stated that gravity acted towards the centre of the earth, 
 hence a vertical line downwards from the earth's surface 
 cannot pass through the centre unless it be drawn from the 
 equator or the poles, which is the same as saying that the 
 earth is not a sphere, but a spheroid. 
 
 Fig. 7 will make this clearer. Let O be the centre of the 
 earth, O F a radius 
 of the equator, and 
 N the pole. Then 
 at a point P in lati- 
 tude P O F, A P 
 B is the horizon- 
 tal plane and C 
 P D the vertical. 
 On account of 
 the earth's rotation 
 about N O, P 
 moves in a circle 
 perpendicular to 
 the plane of the 
 paper whose centre 
 is G, and radius Fig. 7. 
 
 G P, and the centrifugal force acts along P H. Resolving 
 P H in the horizontal and vertical planes, we get P K and 
 P M. Again, let P E represent the force of gravity ; then 
 if P E, which must pass through O, is not in the plane of 
 C P D, we may resolve it in that plane and in A P B, as 
 P D and P L. Then if P, being free, is to remain at rest, 
 P L must equal P K. It is evident that if P E be perpen- 
 dicular to A P B i.e. if N P F be the arc of a circle there 
 can be no force P L. 
 
42 METEOROLOGY. 
 
 40. The Force of Gravity. It is important to note that g, 
 the force of gravity at any place, is represented by P D - P M. 
 Since forces are, in general, measured with reference to , it 
 is necessary to have a standard value, which is usually that 
 of the 45th parallel of latitude, and is equal to 9-8061 
 (metres). Its value at any other point is easily found from 
 the following empirical equation. 
 
 = 9-8061 (i- -00260 cos 2 9) (i- -ou^o)- 
 When 9 is the latitude, and h the height above sea level in 
 metres. 
 
 41. Relative Motion on the Earttts Surface. Now suppose 
 the point P to move on the earth's surface. Then since the 
 earth rotates from west to east, if P moves in an easterly 
 direction it is really moving faster than the earth, and if in 
 a westerly, slower ; hence in the first case the centrifugal 
 force will be greater than is counterbalanced by the ellipticity 
 of the earth, and P will tend to move towards the equator ; 
 while in the second the centrifugal force will be less, and the 
 horizontal component of gravity will tend to turn P towards 
 the pole. Thus the rotation of the earth always tends to 
 turn a moving body, which may of course be a current of air, 
 to the right in the northern hemisphere, and to the left in the 
 southern. This law is now often called Ferrel's Law. 
 
 42. Friction In concluding this chapter, we must remind 
 the student that a current of air experiences considerable 
 resistance from friction. On the earth's surface this is of 
 course least over the ocean, and greatest over a mountainous 
 country. For obvious reasons the coefficients of friction 
 cannot be determined once for all in the laboratory, but only 
 by observation in each particular case. It must not be 
 forgotten that there are also considerable frictional effects 
 between currents of air moving in different directions at 
 different altitudes. The operation of these frictional effects 
 will be more easily understood in the treatment of the 
 particular case given in the next chapter ; in the meantime 
 
FUNDAMENTAL PRINCIPLES. 43 
 
 it must be borne in mind, especially by the student of the 
 applications of mathematical formulae to meteorology, that 
 the difficulties in the way of absolute numerical determin- 
 ations of the coefficients are enormous, a fact which has 
 been rather lost sight of by critics of such applications in 
 this country. 
 
CHAPTER III. 
 
 CYCLONES. 
 
 43. Isobars and Gradients. In Chapter I, we attempted to 
 sift out certain facts concerning one particular type of dis- 
 turbance, as far as was possible from a study of horizontal 
 motions only. Let us proceed to inquire what distribution 
 of pressure will account for the observed horizontal circula- 
 tion, and then test our method by comparison with actual 
 readings of the barometer. 
 
 It will be necessary in this case to deal with distribution 
 of pressure over a large area, and therefore convenient to 
 employ the method of a synoptic chart by noting simul- 
 taneous readings of the barometer at different points. We 
 may then draw lines through series of points having the 
 same barometric reading, which gives us a system of contours 
 
 of heights of the barometer. 
 
 Fie. 8. 
 
 Let Fig. 8 represent a 
 series of such lines drawn 
 at intervals of o'i inch. 
 Then it is evident that by 
 drawing a line as A B 
 everywhere perpendicular 
 to the contours or isobars as 
 they are called, we can find 
 the steepness or gradient, 
 as A' B' in the lower part of 
 the figure, which represents 
 the variations of pressure 
 along A B. In the same 
 way as we measure a 
 railway gradient as i in 
 50, i in 200, and so on, 
 we say barometric gra- 
 
 44 
 
CYCLONES. 
 
 45 
 
 dients are so many thousandths of an inch in 15 miles, 
 or so many millimetres in one degree of the meridian or 
 in kilometres. 
 
 44. Gradients and Winds. Now, the force which causes 
 the atmosphere to move or produce the wind, evidently acts 
 " down " the gradient, from the points of higher pressure to 
 those of lower. Hence the problem of Chapter I. becomes, 
 What forms of isobars, or what direction of gradients will 
 
 Fig. 9. 
 
 give, under the known conditions, the observed wind circu- 
 lation? Let S P T (Fig 9) be any part of one of the thickwind 
 arrows of Figs, i to 4. At P the air is moving in the direc- 
 tion of S' P T. . The forces resisting this motion are two : 
 that due to the earth's rotation which, by Ferrel's law, tends 
 to make it move in a circle O P Q whose centre is R, and 
 that due to friction, which simply tends to prevent any 
 motion at all. These two forces, therefore, act in directions 
 P R, P S'; let them be represented by P R, P U. Then, if the 
 
46 METEOROLOGY. 
 
 velocity is uniform, the force of the gradient must be such 
 that the resultant P V of P R and P U is counteracted, and at 
 the same time there must be a force P K perpendicular to 
 S' P T proportional to the curvature of the path S P T at P, 
 the latter being required to counteract the centrifugal force 
 in S P T. Hence, taking P V backwards, the force required 
 must lie in some direction P L, the resultant of P V and P K. 
 That is, stand with your back to the wind, and the lowest 
 pressure lies to your left and in front. Or, since the isobars 
 are perpendicular to the gradients, we may say the wind 
 moves nearly along the isobars but tends to cross from the 
 higher to the lower ones. We observe that the force tend- 
 ing to make the angle between gradient and wind less than 
 90 is due to the component required to overcome the 
 friction on the earth's surface, and is therefore small in the 
 case described in Chapter I. where the system has come 
 from the ocean. We may, indeed, not be much surprised to 
 find the wind directions nearly parallel to the isobars. 
 
 45. Example of Low Pressure System or Cyclone. Figs. 
 10, ii, 12, and 13 show the distribution of pressure at the 
 times corresponding to Figs. 1-4. We observe in Fig. 10 a 
 fairly uniform barometric slope towards north-west, and as 
 expected, the wind blowing nearly along the isobars, but turn- 
 ing slightly inwards towards the lowest pressures. Fig. 1 1 
 exhibits a similar arrangement ; and Fig. 12 brings us to the 
 root of the matter by showing the nature of the who'e 
 system. The central calm area is the region of lowest 
 pressure ; the isobars are arranged concentrically around it, 
 and in accordance with our prediction, the wind blows 
 in a direction everywhere nearly parallel to the isobars, but 
 tending towards the lower pressure. Finally, we observe 
 from Fig. 13 that the system of lower pressure has moved ii 
 the same way as that of wind circulation, still further con- 
 firming the connection between the two. 
 
 We have thus arrived at what is evidently a much more 
 
CYCLONES. 
 
 47 
 
 Fig. 10 
 
 Fig. ii 
 
 19 6 90 
 
 F'l 12. 
 
 Fig. 13. 
 
48 METEOROLOGY. 
 
 satisfactory method of defining the system described in 
 Chapter I., for instead of merely recording the directions in 
 which the wind blows, we can ascertain the forces which 
 cause it to blow, and we now know something about how 
 these forces act. We may, therefore, in future begin by 
 inquiring as to the forms of the isobars and then discuss 
 particular types of weather associated with them. 
 
 The area of low pressure, and the whole system connected 
 with it, is called a " cyclone " or " depression ; " in America 
 simply a "low." 
 
 46. Intensity of Cyclones, It is evident that in a de- 
 pression the actual pressure or height of the barometer 
 is a matter of little moment ; the forces involved are due to 
 the gradients or differences of pressure, and are greater the 
 steeper the gradients. Hence a cyclone is of the mild type 
 described in Chapter!., or attains to a gale or hurricane just 
 as the gradients are gentle or steep. An excellent example 
 of this fact is given by the great storm which passed over 
 Edinburgh on January 24th, 1868. The lowest reading of 
 the barometer recorded was 29-29 inches, only o'6 in. below 
 the average for the month. The following table gives ap- 
 proximately the gradients for the day : 
 
 Hour. Gradient. 
 
 9 a.m., 
 Noon, 
 
 2 p.m., 
 
 3 P.m., 
 
 4 p.m., 
 
 8 p.m., 
 
 9 p.m., 
 
 inch in 3,333 miles. 
 
 , I2 4 
 
 78 
 
 84 
 
 107 
 
 196 
 
 513 
 
 A terrific hurricane raged from eleven till six o'clock, and 
 between noon and 4 p.m. much damage was done to trees 
 
CYCLONES. 
 
 49 
 
 and buildings, corresponding exactly with the times of 
 steepest gradient. 1 
 
 If we study the barometer changes at different points in 
 the path of a cyclone, it is at once obvious that in the 
 " front " of the depression the barometer is everywhere 
 falling, while in the " rear " it is everywhere rising ; and we 
 thus have the turning point along the line C D (Fig. 5), which 
 
 Fig. 14. 
 
 is called the u trough " of the cyclone. This leads to one 
 of the main difficulties in the use of the barometer at an 
 isolated station. I.f the observing station lies on the line 
 A B, the barometer will fall until the centre O passes, and 
 the line B O evidently lies directly along the gradient. But 
 if the station lies to one side of A B, the direction is not 
 along, but partly across, the gradient, and it is therefore 
 1 T. Stevenson in Good Words> 1868, p. 366. 
 
 D 
 
50 METEOROLOGY. 
 
 impossible to form an accurate estimate of the steepness. 
 Let A and B (Fig. 14) correspond to the same letters in 
 Fig. 5. Then if we draw isobars at intervals of o'i inch 
 we may take stations lying on paths A B, A' B', and A" 
 B." A B being perpendicular to the isobars, lies along the 
 gradient, but A' B' and A" B" lie more or less across it. 
 The lines a o b, a! o' P, a" o" b" show the record 
 that would be made by a self-registering barometer ; ob- 
 serve that in each case the barometer turns on the line o C, 
 the trough of the cyclone, but the fall is less as we recede 
 from A B. If we imagine the figure to represent a plan of 
 a cup-shaped hollow in the ground the isobars are simply 
 contour lines, and the barometer curves a 6, a b\ and a" b" 
 sections along A B, A' B 7 , and A" B". 
 
 Yet another difficulty must be taken into account. We can- 
 not judge of the steepness of the gradient merely by the rate at 
 which the barometer falls at any particular station ; for this 
 evidently depends on the rate at which the depression is 
 passing. In Fig. 14, the barometer at O stands about 0-3 
 inch lower than at B, but the rate at which it falls is deter- 
 mined by the speed with which O travels towards B, just as 
 .two men starting at the same time from the top of a hill may 
 reach the foot at the same moment, although one has climbed 
 slowly down a precipitous face, and the other run full speed 
 down a gentle slope. 
 
 47. Path of the Wind in Cyclones. In discussing the 
 forces which come into play when air is set in motion by a 
 difference of pressure, we found that the angle between the 
 wind direction and the gradient was slightly less than a right 
 angle, and that the deficiency from 90 was due to the 
 friction of the earth's surface, so that the greater the friction 
 the smaller the angle. We found, further, that in a system 
 travelling over the ocean the friction was so slight that the 
 wind direction seemed to follow a path in many places 
 nearly parallel to the isobars. So much is this the 
 
CYCLONES. 51 
 
 case in most cyclones passing over the British Islands that 
 rules drawn up for practical purposes usually ignore the de- 
 fect from a right angle altogether, and assume that the wind 
 blows perpendicular to the gradient, i.e. along the isobars. 
 Thus Buys Ballot's Law is usually stated, " Stand with your 
 back to the wind and the lowest barometer lies to the 
 left " ; nothing is said about the front. 
 
 But while we do not in this way introduce serious error in 
 
 B 
 
 Fig. IS- 
 
 practical calculations about the British Islands, the omis- 
 sion of the last clause would in many parts of the world 
 be simply disastrous ; and in all cases it is inaccurate. 
 
 Consider in detail the path of the wind when the isobars 
 are concentric circles, and the gradients therefore slope 
 radially towards a central point, and when the direction of 
 the wind makes a constant angle ^, less than 90, with the 
 
52 METEOROLOGY. 
 
 gradient. Let A B C D (Fig. 15) represent one of the isobars 
 (O the centre) ; then at any point P the air is moving in a 
 direction P/ making an angle ^ with O P ; it thus arrives, 
 say, at/. But at /it still travels at an angle ^ with of, and 
 so on. Hence the path is the curve called a logarithmic spiral. 
 We are accordingly led to conclude that in a cyclone the air 
 moves towards the centre in a spiral, and we observe again 
 that the angle ^ at which the incurving takes place depends, 
 other things being equal, on the friction between the air 
 currents, and the earth's surface, being smaller, the greater 
 the friction. The value of ^ must also be different at 
 different parts of the globe, since the deflecting force, due 
 to the earth's rotation, is not everywhere the same. If the 
 earth were at rest the wind would simply blow radially in all 
 directions towards the centre. 
 
 48. Influence of Circulation on the Gradient. We have 
 hitherto considered the gradient as a given force, which may 
 serve to explain the facts observed. But the wind circula- 
 tion in a cyclone must clearly react to a greater or less ex- 
 tent upon the gradient. We may compare the motion of 
 any particle of air in a cyclone to that of a stone whirled at 
 the end of a string, the string being meanwhile drawn in by 
 being wound round a stick. We know that as the string 
 gets shorter the stone revolves faster ; and Newton proved 
 that in such a case stone and string describe equal areas in 
 equal times. But the "centrifugal force " counteracted by 
 the tension on the string increases as the square root of the 
 cube of the increase of velocity. Hence, applying this to 
 the air in a cyclone, it appears that the air tends more and 
 more to keep away from the centre, and we must therefore 
 have a decrease of pressure, the gradient becoming steeper 
 and steeper near the centre. 
 
 49. Vertical Circulation in the Atmosphere: Condition 
 of Equilibrium. Granting that the air in a cyclone does 
 move in towards the centre, what becomes of it when the cen- 
 
CYCLONES. 53 
 
 tral area is reached ? Involved along with this is the further 
 question What causes the central force which gives rise to 
 the cyclone ? Evidently the only path open to the air from 
 the central area is in an upward direction. 
 
 The condition of rest, as stated in 20, is that the pres- 
 sures at any two points differ by the weight of a cylinder of 
 unit sectional area, whose length is equal to the perpen- 
 dicular distance between them. Since a column of air i sq. 
 cm. in section and i cm. in height at temperature o C. and 
 barometric pressure 760 mm. weighs '001293 grammes, we 
 can easily calculate the height of the atmosphere supposing 
 the density to be the same at all levels ; the average weight 
 is 1033-3 grammes, hence the height i^g- cm. = 7993 metres. 
 This is called the "height of the homogeneous atmosphere." 
 
 But the density varies as the pressure ; and since the 
 pressure in any horizontal layer is simply the weight of 
 the layers above it as there is no rigid enclosing shell, 
 the air must become less and less dense as we 
 ascend. Let fig. 16 represent a column of air 
 of i square metre section, whose total weight at the 
 earth's surface is P ; and let it be supposed that the 
 temperature is the same throughout. Then if/ x is 
 the pressure at a height of h metres, and p^ that 
 at (h i) metres, p p evidently represents the 
 weight of i cubic metre at the pressure / L to a 
 close approximation. Call this weight W x ; then 
 since the density is proportional to the pressure we 
 may put \V 1 = Cp 1 . Similarly, we have for a 
 height of// i metres W 2 = Cp 2 , for h 2 metres 
 \V 3 = Cp 3 , and so on. Hence we get Fig l6 
 
 * c -/,-/, -~ 
 
 / 2 
 
 and so on. and so on. 
 
54 METEOROLOGY. 
 
 Multiplying the left sides and right sides of the latter series 
 together, we get, remembering that P =/ a +/ 2 +/ a -I- "", 
 
 From the known properties 1 of damp air we can easily de- 
 termine the values of \\\ for different conditions of tem- 
 perature and moisture, whence we can find an expression 
 for C. In this way, the above formula becomes, after all 
 reductions have been made, 
 
 ^==18401-2 (i-f -00367 - ) (i-f'378 J-) 
 
 Z P 
 
 X (l X 'OO26 COS 2 </>) (l - - ) log - 
 
 5096720 / 
 
 where // = difference of levels between the two stations in 
 metres. 
 
 station. 
 
 e = pressure of vapour. 
 
 9 = the latitude. 
 
 Z = the mean of heights of the two stations above 
 
 sea-level. 
 
 This gives, then, the relation between pressures at differ- 
 ent elevations ; and by assuming the conditions at a par- 
 ticular time to be those of equilibrium, the formula may be 
 used to measure heights. 
 
 50. Ascending Currents, S/tiMe, Unstable, and Indifferent 
 States. Suppose now that from any cause an ascending 
 current is initiated at the earth's surface ; then we have seen 
 ( 21) that if the air is perfectly dry, it will be cooled nearly 
 i C. for every 100 metres of ascent. Hence there are 
 three possible cases with reference to the normal gradient 
 or rate of cooling of the surrounding air (p. 33). 
 
 1 See Sprung, Lfhrbttfh </fr Afettorologie, p. 67. 
 
CYCLONES. 55 
 
 i The fall of temperature maybe less than 1 C, per 100 
 metres. 
 
 2 It may be exactly i C. 
 
 3 It may be greater than i C. 
 
 In the first case, supposing the ascending current to start 
 from the earth's surface at the same temperature as the sur- 
 rounding air, when it reaches the height of 100 metres it 
 will be colder than the air surrounding it and therefore 
 heavier ; consequently the ascending motion will be 
 diminished. 
 
 In the second case, since the temperature in the ascend- 
 ing current falls at the same rate, the relative conditions 
 simply remain the same at all elevations. 
 
 In the third, the surrounding air having cooled faster than 
 the ascending current, the latter will be warmer than it at 
 the height of 100 metres, and the tendency to ascend will be 
 increased. 
 
 When the air is, as of course always happens, more or 
 less moist, the ascending current cools a little more slowly. 
 If the ascending motion continues until the temperature of 
 saturation is reached the rate of cooling suffers a further 
 sudden diminution, for the air is warmed by the latent heat 
 set free, and the ascending force correspondingly increased. 
 An initial higher temperature evidently still further favours 
 the continued ascent. 
 
 It appears, then, that if from any cause an ascending 
 current is established, there are certain definite conditions 
 depending partly on the temperature gradient and parti)-, on 
 the original strength of the ascending current, which deter- 
 mine whether the ascending motion is to continue and 
 increase or not. If it continues, it is evident that the 
 requirements of a cyclone are satisfied, for the removal of 
 air by the ascending current causes a diminution of pressure, 
 and the surrounding air tends to flow in from all sides, and 
 we see that a severe storm may easily be developed by the 
 
56 METEOROLOGY. 
 
 action of forces due to the earth's rotation on a quite incon- 
 siderable disturbance. When the normal gradient is such 
 that the ascending motion is checked, the atmosphere is said 
 to be in the stable state ; where the ascending motion is 
 encouraged, unstable ; and where the gradient is the same 
 as that in the ascending current, the condition is said to be 
 indifferent. 
 
 A striking illustration of the facts just explained may be 
 taken from the records of the Ben Nevis High and Low 
 Level Observatories, which is of special interest as it indi- 
 cates a line full of rich promise of a quite new method of 
 weather forecasting, which may one day supersede that of 
 synoptic charts. According to these observations the mean 
 rate of fall of temperature with increase of elevation is 
 o - 63 C. per 100 metres. Now the mean temperature of 
 the low level observatory is 8 C. ; hence a current of air 
 starting at a temperature of 10 C. will on an average be 
 denser than the surrounding air when it reaches a height of 
 about 850 metres, if we suppose that it cools o'9 C. for 
 every 100 metres of ascent. But the observations at the two 
 stations show that an unusually rapid decrease of tempera- 
 ture with height, especially when conjoined with an unusually 
 slow decrease of pressure with height, are " frequent con- 
 comitants and precursors of storms." 1 This means that 
 the conditions are favourable to the maintenance of ascend- 
 ing currents. 
 
 51. Condensation : Special Centres of Energy. We have 
 seen that if condensation begins while the ascending air is 
 still at a fairly high temperature, the tendency to ascend is 
 markedly increased. But as condensation goes on, and less 
 vapour is left in the air, the rate at which the temperature 
 falls increases again. And it is matter of observation that 
 at considerable heights above the earth's surface, the normal 
 temperature gradient becomes very much less. Hence the 
 1 Trans. ft.S.E. xxxiv., p. xlix. 
 
CYCLONES. 5/ 
 
 "indifferent" state must ultimately be reached, and that 
 probably at no very great height. It is worth noting that 
 there is in general a special centre of energy in the part of 
 the ascending column between the point where condensation 
 commences and that where the indifferent stage is reached. 
 This part of the column will be at a greater elevation the 
 drier the air supplied to the column originally. A peculi- 
 arity of the storms of Indian seas may probably be explained in 
 this way : it is found that in the rainy season when the air is 
 everywhere nearly saturated, storms which approach the land 
 are easily broken up and destroyed on meeting even a com- 
 paratively low range of hills ; while the storms preceding 
 and following the rains may cross a considerable tract of 
 country and emerge upon the ocean undiminished in 
 strength. 
 
 52. Descending Currents: Anti- Cyclones. Whatever quan- 
 tity of air ascends to the upper regions of the atmo- 
 sphere, an equal quantity must ultimately descend to take 
 its place. The continued up-draught of air must cause an 
 excess of pressure in the higher strata, and hence a tendency 
 to settle down. We thus have a system established which is 
 exactly the opposite of % cyclone ; an area of high pressure, 
 fed by the air flowing out from the ascending current, in 
 which the air tends to descend. In our preliminary in- 
 quiries about the cyclone we had the problem given 
 the circulation to find the forces : the present question 
 is given the force, find the circulation. If the earth 
 were at rest, the air would on reaching the earth's sur- 
 face flow directly from the area of high pressure to 
 that of lower ; but by Ferrel's Law it is in the northern 
 hemisphere deflected to the right : hence the circulation is 
 the opposite to that in a cyclone, the direction of motion 
 being with the hands of a watch, and at the same time tend- 
 ing away from the centre of high pressure. Systems of this 
 description are called anti-cyclones. 
 
58 METEOROLOGY. 
 
 53. Origin of Cyclones. It appears, then, that the essen- 
 tial part of the cyclone is a central ascending current, and 
 we have seen that under certain conditions the environment 
 of the current may be such as to materially strengthen it, 
 and under certain others to weaken it. Our information 
 with regard to the upper strata of the atmosphere is unfor- 
 tunately very incomplete ; it is derived chiefly from observa- 
 tions of the movements of clouds and the records of high 
 level observatories. The two most important series of cloud 
 observations are those of Ley and Hildebrandsson, which 
 agree in showing that the direction of the lower clouds is some- 
 what to the right of the surface winds, nearly parallel to the 
 isobars, while the highest or cirrus clouds show a marked 
 tendency to move away from the centre, their direction of 
 motion being still further to the right. Observations at 
 high level stations on the Alps have shown that in the 
 central area of a cyclone the temperature gradient closely 
 agrees with what we have been led to expect in an ascend- 
 ing current. 1 Again, Hann has found unmistakable 
 evidence of descending currents in anti-cyclones. 2 
 
 So much concerning cyclones we may therefore take as 
 established by observation ; but we are obviously left with 
 the formidable question What causes the ascending current 
 necessary for the cyclone? and it may be said that this 
 question is as yet unanswered. Until recently what is 
 known as the convection theory was believed to give a fairly 
 satisfactory solution, and in the hands of mathematical 
 meteorologists, and especially of the late Professor Ferrel, it 
 served to explain many important phenomena. The sub- 
 stance of it was briefly this : given an unstable vertical 
 temperature gradient, if at any place on the earth's surface 
 the air becomes slightly warmer than in the immediate 
 neighbourhood, it will ascend, and on account of the sur- 
 
 i Met. Zeit. y 1890, p. 332; 
 
 2/foV/., 1890, p. 227. 
 
CYCLONES. 59 
 
 rounding gradient the ascending motion will continue and 
 increase, until a cyclone is developed. The ascent will 
 continue until the gradients become indifferent, when on 
 account of the higher pressure the gyrations will be reversed 
 and become anti-cyclonic. From this anti-cyclone, colder, 
 denser air will radiate and slowly settle down on all sides of 
 the ascending current. This theory involves the con- 
 dition that the temperature gradient, in the area surround- 
 ing the cyclone, shall be unstable that the ascending current 
 shall be warmer than the normal, and the descending 
 current in the anti-cyclone colder. There can be no doubt 
 that when these conditions exist they are specially favourable, 
 and it must be admitted that with unstable conditions a 
 cyclone will develop more energy than ceteris paribus it 
 would be likely to do otherwise. But Hann 1 and others 
 have shown conclusively that cyclones frequently occur 
 when the gradient is far from being unstable, and not only 
 so, but that the ascending current in cyclones is often not 
 warmer, but colder than the descending current in anti- 
 cyclones ; facts in themselves fatal to the convection theory 
 in so far as it attempts to account for the origin of all 
 cyclones. We have, in addition, the fact that cyclones do 
 not occur with the greatest frequency in summer, when on 
 account of the warming of the earth's surface the gradients 
 are oftenest unstable and excessive heating over local areas 
 is most likely to take place. 
 
 It seems probable that the true cause of ascending and 
 descending currents is to be looked for in the great move- 
 ments of the atmosphere as a whole ; and in particular at 
 the interfaces between different main streams in the general 
 circulation. It is known that in different parts of the world 
 cyclones tend, at different seasons of the year, to follow 
 certain paths, and these paths appear to be directly or in- 
 directly connected with movements on a larger scale. 
 1 Si lib. der K. Akad. der Wiss. 1890. 
 
60 METEOROLOGY. 
 
 Helmholtz 1 has recently shown that a comparatively rare 
 upper current blowing along a dense under current, in the 
 same way as the wind blows over the surface of the sea, 
 will cause waves measuring many miles from trough to 
 trough and if the under current be supposed sufficiently 
 shallow, we can imagine that the air at the bottom, i.e. on 
 the earth's surface, may be seriously disturbed. The 
 mathematical difficulties involved in such problems are, 
 however, so great that a complete solution is still far out of 
 reach. 
 
 Having now discussed in some detail the mechanism of 
 an ideal cyclone, supposed exactly circular, and stationary 
 in position, we shall be the better able to understand the 
 bearings of further facts concerning the systems which 
 commonly occur. We have begun by satisfying ourselves of 
 the truth of certain facts, employing a method of probabilities 
 then by the help of known physical laws we have been able 
 to predict certain other phenomena, developing in this way a 
 theory of cyclones which although admittedly incomplete 
 has nevertheless shown itself useful in enabling us to under- 
 stand the main features of these disturbances. We return 
 to the method of probabilities, but are now in a position to 
 make a more accurate use of it by making numerous 
 measurements of the same thing and calculating the most 
 probable value : we may for example find the relative con- 
 ditions of temperature in the front and rear of cyclones by 
 observing the readings of the thermometer in a large number 
 of cases and taking the most probable value, which in such 
 a case is the average. 
 
 54. Form of Cyclones : Direction of Longest Diameter : 
 Size. Cyclones are not in general circular, but oval. 
 Loomis 2 found that on an average of three years the ratio 
 of the longest diameter to the shortest was 1-94 in the 
 
 1 Siizb. dzr Be.) liner Akademie. 1889, p. 503. 
 
 2 Contributions fc Meteorology ', 1885, pp. 7, 8 t 
 
CYCLONES. 6 1 
 
 United States and 1-70 in the Atlantic Ocean; and in more 
 than half the cases 1-5. Van Bebber T found that in 77 
 European cyclones the average ratio was 1*78, with a tend- 
 ency to be more circular in winter and more oval in summer. 
 According to the same investigators the longest diameter of 
 cyclones in the United States or in Europe lies in a direction 
 nearly W.S.W. to E.N.E. in a very large majority of cases, 
 while over the Atlantic the direction varies within consider- 
 able limits. 
 
 It is scarcely possible to fix a superior limit to the size of 
 cyclones ; the area covered may be several thousand miles 
 across. But when the dimensions become great, particularly 
 if the system is much elongated, the cyclone frequently 
 breaks up into two, three, or even more separate centres of 
 depression, each with its own particular system of wind 
 circulation, between which are belts of calms or light vari- 
 able winds. When such subdivision takes place one of the 
 centres usually increases in depth and intensity at the ex- 
 pense of the others, which fill up and disappear. Hence we 
 may have a violent storm of comparatively small area de- 
 veloped from a cyclone of small intensity but great extent. 
 Such changes may occur with great rapidity, so that it is 
 sometimes impossible to identify the same cyclone on two 
 synoptic charts drawn at an interval of only a few hours. 
 We have here obviously a fruitful source of failure in weather 
 forecasting; all the symptoms of an approaching cyclone 
 may be exhibited and the prognostics falsified by the filling 
 up of the depression or the weather of a cyclone front 
 may be experienced without being followed by the con- 
 ditions characteristic of the rear. It is further evident that 
 large cyclones must be modified in form and position by the 
 variations of the deflecting force due to the earth's rotation 
 in different parts of it, arising from difference of latitude. 
 
 1 Handbuch der Ausilbenden Witterunrjsku.nde, pt. II., p. 221, 
 
62 METEOROLOGY. 
 
 It may be said that the average size of cyclones is greater 
 the higher the latitude. Tropical cyclones are in general 
 smaller than those of the temperate zones, and more 
 symmetrical in shape ; their diameter does not often exceed 
 500 miles, and on an average the shape of the isobars is more 
 nearly circular. Small cyclones must not, however, be con- 
 founded with waterspouts or tornadoes, which are too small 
 to be sensibly affected by. the rotation of the earth, and may 
 revolve in either direction ; these are special phenomena of 
 a quite distinct nature. 
 
 55. Gradients and Observed Winds. In our discussion 
 of the mechanism of an ideal cyclone we saw that with any 
 given gradient there is associated a wind velocity whose 
 direction and speed depend on the deflecting force due to 
 the earth's rotation, and on the fractional resistance offered 
 by the earth's surface and by other masses of air. There 
 are therefore, a number of unknown quantities introduced in 
 the shape of coefficients of friction which make it difficult, 
 if not impossible, to calculate numerical values of wind 
 velocity from given gradients by the help of theoretical 
 formulae. Hence we merely state that the changes of 
 wind velocity with given changes of gradient are found under 
 favourable circumstances to justify the assumptions on which 
 the formulae are based, thereby confirming the correctness 
 of the theory and proceed to- the numerical results of 
 observation. 
 
 Loomis 1 gives the following results for cyclones over the 
 North Atlantic Ocean. The mean latitude of the centres 
 of these depressions was 58 0< 48', and the measurements were 
 made at the part of the cyclone where the gradients were 
 steepest. 
 
 i Contributions to Meteorology, p. 127. 
 
CYCLONES. 
 
 Isobars. 
 
 Latitude 
 
 N. 
 
 G. 
 
 Velocity. 
 
 Inclination. 
 
 715-720 
 
 58-o' 
 
 4-I7 
 
 I5-90 
 
 25 -24' 
 
 720-725 
 
 5736 
 
 3-87 
 
 l6'2I 
 
 26-30 
 
 725-730 
 
 57-12 
 
 371 
 
 16*14 
 
 27*54 
 
 730-735 
 
 56-48 
 
 3 '60 
 
 1 6-08 
 
 29-54 
 
 735-740 
 
 56-24 
 
 3'52 
 
 l6'I4 
 
 31-54 
 
 740-745 
 
 55"54 
 
 3'45 
 
 16*27 
 
 33'I8 
 
 745-750 
 
 55-30 
 
 337 
 
 16*27 
 
 33-42 
 
 750-755 
 
 55-0 
 
 3*24 
 
 15-84 
 
 34'6 
 
 755-760 
 
 54'3o 
 
 3-10 
 
 14-73 
 
 34H8 
 
 The first column gives the two isobars between which the 
 measurements are made, the second their mean latitude, 
 the third the gradient expressed in millimetres per degree of 
 the meridian (in kilometres), the fourth the velocity in 
 metres per second, and the fifth the angle between the 
 tangent to the isobars and the direction of the wind, 
 reckoned towards the centre. 
 
 The investigations of Ley 1 showed that in general a given 
 gradient is associated with stronger wind in summer than in 
 winter ; thus for Kew he obtained the following results : 
 
 Gradient . . .6 
 
 Velocity (m. per sec.) Winter r8 
 
 Do. do. Summer 2*9 
 
 9 
 
 2-3 
 3'9 
 
 12 
 
 3'4 
 
 15 
 
 5*o 
 6-4 
 
 18 
 6-3 
 7'5 
 
 21 
 
 24 
 
 8-0 
 9'4 
 
 Ley 2 further found that with a given gradient northerly 
 and easterly winds are markedly stronger than westerly or 
 southerly, results fully confirmed by Sprung 3 for stations on 
 the German coasts. The angle between the direction of 
 the wind and the isobars is given by Ley 4 as 13 for coast 
 stations, and 29 for inland stations, thus showing distinctly 
 the increased effects of friction on land. The mean of these, 
 
 1 Qu. Jour. Roy. Met. Soc.> III. p. 232. 
 
 2 Nature, vol. xxiv., p. 8. 
 
 3 Archiv der Deutschen Seeivarie, II., 1879. 
 tjoitr. Scot, Met. Soc., vol. iv. , 1873, P- 7 2 - 
 
6 4 
 
 METEOROLOGY. 
 
 21, agrees exactly with the value for Denmark given by 
 Hoffmeyer 1 , whose investigations also show a higher value 
 for southerly and easterly, i.e. land winds, than for northerly 
 and westerly. Loomis 2 gives for the eastern part of the 
 United States a mean inclination in cyclones of 47, the 
 greatest values being for north or north-west winds, which 
 are, in this case, the land winds. 
 
 The researches of Ley, Hildebrandsson, and others, have 
 shown that in the cyclones of Western Europe, at least, the 
 gradient is seldom, if ever, symmetrical round the centre, 
 being in four out of every five cases steepest in the quadrant 
 between S.E. and S.W. of the centre, and least in that be- 
 
 Fig. 17. 
 
 tween N.E. and N.W. Further, it is found that the in- 
 curving of the wind from the isobars towards the centre is 
 greatest in the front of the cyclone, and least in the rear. 
 Ley 3 divided the cyclone area into eight outer and eight 
 inner segments, as shown in Fig. 17, and gives the following 
 averages for the angles of incurving. The arrow shows the 
 cyclone's line of advance. 
 
 1 Met. Zeit., xiii., 338. 
 
 (?/ Met. Sec., 1877. 
 
 2 Anier. Jour, of Science ', i874 
 
CYCLONES. 65 
 
 Segment. Angle. Segment. Angle. 
 
 a x 11 a 3 13 
 
 b r 14 b 2 1 6 
 
 ci 24 c<, 26 
 
 d x 36 d a 35 
 
 e L 42 e 2 32' 
 
 fi 38 f a 37 
 
 gi 28 g 2 25 
 
 h t 10 ho 9 
 
 Mean 25'24' Mean 246' 
 
 It is now generally admitted that the warm moist 
 southerly winds in the front of the cyclones of Western 
 Europe are the principal source from which the system de- 
 rives its energy ; they are the real feeders of the ascending 
 current, and therefore it is not surprising to find a stronger 
 indraft there than occurs in the case of the colder, drier 
 northerly and north-easterly winds in the rear. 
 
 56. Distribution of Pressure at Higher Levels. This leads 
 to another consequence of the higher temperature of the air 
 flowing into the front of a cyclone than of that flowing into 
 the rear. Pressure must evidently decrease more rapidly 
 with height in the colder, denser air than in the warmer; and, 
 therefore, in any horizontal plane at, say, 1,000 metres eleva- 
 tion, the form of the isobars will not be the same as at the 
 earth's surface, but pressure will be relatively lower on the 
 northern and western sides, and the centre will be displaced 
 northwards. From observations of cirrus clouds, Ley 1 
 found that in the British Isles the centre was not only so 
 displaced but also lagged behind at higher levels. 
 
 57. Inclination of Winds at Higher Levels. We have 
 already stated (55) that observation confirmed the supposi- 
 tion that in the higher part of a cyclone the air flowed away 
 from the centre. The following table gives the average 
 angle between the isobars at the surface and the direction 
 
 1 Q. J. Met. Soc., 1887. 
 
Segment. 
 
 Angle. 
 
 a.. 
 
 
 
 bo 
 
 -39 
 
 Co 
 
 -17 
 
 do 
 
 12 
 
 p 
 
 45 
 
 f 2 : 
 
 40 
 
 66 METEOROLOGY 
 
 of the upper currents, shown by the motion of cirrus clouds, 
 as determined by Ley. The letters refer to the segments of 
 Fig 17, and the angles are this time reckoned positive away 
 from the centre. 
 
 Segment. Angle. 
 
 aj. 6 
 
 b x 11 
 
 CJL 34 
 
 dj. 56 
 
 e x 62 
 
 fi 73 
 
 gi -95 g 3 82 
 
 hV 9 ho 1 6 
 
 Mean i9 c '3o' i7 0< 24' 
 
 It appears from these numbers that, just as the inflow of 
 air at the surface is greatest on the front and right-hand 
 segments, so at the cirrus level the outflow is greatest in the 
 same segments. Ley concludes accordingly that the south- 
 easterly winds probably extend to only about half the eleva- 
 tion of the northerly winds on the left side of the cyclone, 
 and further explains the excessive outflow on the right side by 
 the fact that the great majority of our cyclones are associated 
 with an anti-cyclone lying to the right of them, the down- 
 current of which is fed by the air ascending in the cyclone. 
 We may note here the fact that the cirrus clouds observed 
 in this outflowing current are often the first indication given 
 of an approaching cyclone ; it is well known that mare's- 
 tail clouds moving from the north-west frequently precede 
 storms. 
 
 58. Distribution of Temperature 'and Rainfall. The dis- 
 tribution of temperature at the earth's surface in a cyclone 
 is in great part determined by the direction of the wind ; the 
 southerly winds in the front raise it above the average, and the 
 
CYCLONES. 
 
 6 7 
 
 northerly winds in the rear depress it. Along with this the 
 effect of the increased cloudiness in checking radiation must 
 be taken into account ; in summer the diminished sunshine 
 keeps the temperature below the average, while in winter 
 the loss of heat into space is in a measure prevented, and 
 temperature remains above the normal. 
 
 The position of the rain area and of the ring of low cloud 
 which surrounds it varies considerably, but the researches 
 of Loomis, Abercromby and others show decidedly that 
 
 Fig. 18. 
 
 while primarily related to the centre of the cyclone it gener- 
 ally extends furthest to the right front. The depression 
 described in Chapter I. affords a good example of this. 
 
 59. Summary : Diagram of Typical Cyclone. The best 
 summary of the knowledge we have so far gained concern- 
 ing cyclones may be given in Fig. 18 (modified from one due 
 to Moller 1 ), which represents a system moving eastward. 
 1 Ann. der Hydr., 1882. 
 
68 METEOROLOGY. 
 
 The heavy black lines indicate the isobars, the heavy arrows 
 the wind directions, and the dotted lines the isothermals at 
 sea level. The light lines and arrows give the isobars 
 and wind directions at the level of the cirrus clouds. The 
 shaded areas indicate the regions of lowest pressure in each 
 case. 
 
 60. Motion of Cyclones : Cyclone Tracks. Beginning 
 with the cyclones of Western Europe, the first and 
 most important fact established by Ley is, that by far 
 the greater number of cyclones move in an easterly 
 direction : when a westward motion occurs it is usually slow 
 and seldom continues for any length of time. The re- 
 searches of Abercromby, Koppen, and Van Bebber have 
 since shown that certain definite tracks can be laid down, 
 along which cyclone centres tend to move, and that 
 particular tracks are specially frequented during each 
 season of the year. The following is a summary of the 
 principal results obtained by the last-named investigator 
 from a discussion covering the period 1875-90. 
 
 Route I. begins to north-west of the Outer Hebrides and 
 runs in a north-easterly direction along the coast of Norway, 
 till it cuts the Arctic Circle, where it divides. This route is 
 much frequented at all times of the year, but especially in 
 autumn and winter. 
 
 Route II. runs from north of the Shetland Isles to the 
 Gulf of Finland. Not very frequently followed. 
 
 Route III. starts somewhat further north than II., and 
 runs either to the Skagerak or central Sweden. After reach- 
 ing the Baltic, cyclones following this track take either an 
 easterly or south-easterly course : and they almost invariably 
 occur between September and March. 
 
 Route IV. extends from the mouth of the English Channel 
 to the Gulf of Finland, and is chiefly followed by shallow 
 depressions during the summer months, although severe 
 i Met. Zeit. t 1891, p. 361, 
 
CYCLONES. 69 
 
 winter storms sometimes take this path, as for example the 
 great blizzard of March Qth and ioth, 1891. 
 
 Route V. also commences at the mouth of the English 
 Channel, and runs south-eastwards parallel to Route III. to 
 the gulf of Genoa, where it divides into three, one branch 
 turning northwards through Switzerland to the Baltic pro- 
 vinces of Russia, a second going eastwards to the Black Sea, 
 and a third curving southwards into the Mediterranean. 
 This track is most frequently followed during the winter 
 months. 
 
 These five principal routes vary somewhat in position 
 from month to month, and it must, of course, be under- 
 stood that there are, unfortunately for successful weather 
 prediction, a considerable number of depressions which do 
 not follow any of them consistently. Van Bebber, how- 
 ever, may be said to have finally established the two follow- 
 ing laws with reference to the general conditions prevailing 
 in the area affected by a depression, which practically em- 
 body the results obtained by Ley in 1872. 
 
 i The advance of a depression takes place in a direction 
 perpendicular to the line of steepest gradient, so that the 
 highest pressure lies to the right. Observe that this accords 
 with what we know of the motion of the upper currents 
 feeding anti-cyclonic areas to the right of the cyclone. 
 
 2 Depressions usually move so that the temperature is 
 highest to the right of their track. 
 
 The first law plays the more important part in winter, and 
 the second in summer. 
 
 6 1. Types of Weather. We can thus understand that the 
 weather of western Europe is in great part determined by 
 the relative pressures and temperatures over three great 
 areas ; the central part of the North Atlantic, the northern 
 part in the neighbourhood of Iceland and the continent of 
 Europe. It is found that the Iceland area is almost always 
 one of relatively low pressure, hence the tendency of cyclones 
 
70 METEOROLOGY. 
 
 to move eastwards, keeping the higher pressure on the right. 
 Again ( 39) pressure is relatively higher over the land than 
 over the sea in winter, and the reverse in summer, hence on 
 the whole depressions tend to skirt along the coast in 
 winter, and penetrate more easily inland in summer, the 
 path being, however, considerably modified by the distribu- 
 tion of temperature in the warmer season. Such consider- 
 ations have led to the classification of weather into a number 
 of types, each of which is distinguished by the prevalence 
 of cyclones following special tracks in succession. When, as 
 frequently happens, the general distribution of temperature 
 and pressure remains constant, a particular type of weather 
 may persist for a considerable period, giving rise to pro- 
 longed spells of weather ; and in these circumstances regions 
 lying outside the cyclone tracks may have settled fine 
 weather for weeks together while the districts traversed by 
 them experience a continued succession of disturbances. 
 
 62. Storms. The percentage of depressions in which 
 winds of dangerous violence are developed depends very 
 largely on the influence of land. We observe that nearly all 
 the main cyclone tracks tend to skirt along the coast or to 
 pass over inland seas, probably on account of the smaller 
 frictional resistance offered to their motion. But the air 
 drawn in by the depression from a land area must evidently 
 contain much less vapour than that over the ocean, and we 
 accordingly find a marked inclination to diminish in in- 
 tensity. Thus from an examination of 154 cyclones Van 
 Bebber T found that, while 65 per cent, occasioned gales in 
 the British Islands, only 32 per cent, retained dangerous 
 energy till they reached the coast of Germany. The records 
 of the British Meteorological Office ' 2 show that on the west of 
 Ireland, the north-west of Scotland, and the south-west of 
 England, all of which lie in the neighbourhood of great 
 
 1 Ilandbuch, pt. II., p. 265. 
 
 -Report of the Meteorological Council, 1887, p. 22. 
 
CYCLONES. 71 
 
 storm tracks, about twenty severe gales occur annually, 
 while in the north-east and east of England the number is 
 about eleven. On the whole, serious disturbances are most 
 likely to follow one of the main routes. 
 
 63. Rate of Motion of Cyclones. The rate of advance of 
 depressions varies within very wide limits. Van Bebber 
 gives, as the average of 1676 cases, a mean velocity of 27 
 km. per hour, the highest average occurring in October 
 (31 km.), the lowest in August (23 km.). He finds further 
 that at all times of the year depressions following the main 
 tracks usually move with considerably greater velocity than 
 " erratics," the highest averages being for Routes I., III., 
 and V. 
 
 It may be said that on the whole deep depressions move 
 faster than shallow, although the difference is not by any 
 means strongly marked. A much closer relation is observed 
 between the change of depth of a depression and its velocity 
 of translation. A cyclone which is increasing in depth and 
 intensity moves with increasing velocity, and conversely ; 
 and we have further the remarkable fact, that one which 
 maintains a constant depth moves in general slower than 
 one which is gradually becoming shallower,, unless the rate 
 of filling up is extremely rapid. 
 
 64. Relation of Winds to Rate of Advance. We have 
 stated in a former paragraph that the velocity of the wind 
 in a cyclone is not apparently appreciably affected by its 
 motion of translation. If we consider the case of a de- 
 pression travelling eastwards, we should expect an easterly 
 component in the winds throughout trie whole system which 
 might produce important effects in a shallow depression 
 moving with considerable velocity, The effect would 
 evidently be to cause outflowing winds in the front of the 
 system, and to increase the incurving in the rear ; again, 
 in the central calm area, we should find a wind correspond- 
 ing in direction and speed to the velocity of the system. 
 
72 METEOROLOGY. 
 
 But the incurving of the wind is greatest in front and least 
 in rear, and the wind required in the central area is never 
 observed. It therefore seems probable, as suggested by 
 Sprung, 1 that at least in the lower layers of the atmosphere, 
 the air is not moved forward in the cyclone in its progress, 
 but that the system is of the nature of a wave disturbance, 
 always affecting new masses of air. 
 
 65. Special Cyclone Areas. It may be laid down as a 
 general rule that cyclones generally move in an easterly 
 direction in all parts of the northern hemisphere north of 
 lat. 30. According to Loomis the average rate of motion 
 over the Atlantic Ocean is 29 kilometres per hour, and over 
 the United States, 457 kilometres, or nearly double that ob- 
 served in Europe. So far as is known, cyclonic storms 
 seldom or never originate within 5 of the equator. The 
 intermediate tropical belt is, however, the scene of the most 
 violent hurricanes known, especially in the neighbourhood 
 of the West Indies and in the Indian and China Seas. The 
 West Indian cyclones, most of which have their origin be- 
 tween 10 and 20 N. lat., usually move in a westerly or 
 north-westerly direction until they approach the thirtieth 
 parallel, when the path curves round by north towards east, 
 their average rate of progress westwards being 24 kilometres 
 per hour. In the Indian Ocean and China Sea the nature 
 and distribution of storms vary at different seasons of the 
 year in a manner too complex to be given in detail here. 
 The course of cyclones is, in general, westerly during its 
 first part, and it may or may not recurve eastwards, the 
 average rate of motion being still slower than in the West 
 Indies. 
 
 1 Lchrbuch) p. 246. 
 
CHAPTER IV. 
 
 ANTI-CYCLONES : OTHER FORMS OF PRESSURE AREAS. 
 
 66. Anti-Cyclones. In describing cyclonic systems we 
 had occasion to point out that the ascending currents must 
 be compensated by equivalent descending currents, and 
 that a wind circulation was associated with the descending 
 current, which was in every respect the antithesis of the 
 cyclone. 
 
 It is to be observed at the outset that anti-cyclones differ 
 very greatly from cyclones in intensity. The air is absorb- 
 ing heat and increasing in capacity for moisture as it de- 
 scends, and therefore no energy is available in the form of 
 heat. Again, the deflecting force of the earth's rotation acts 
 on a mass moving away from a centre instead of towards it ; 
 hence there is no tendency for the rotational velocity to 
 increase, as in the cyclone, on account of the principle of 
 equal areas. 1 Anti-cyclonic systems are therefore in general 
 ill-defined in shape, the barometric gradients are slight, and 
 the normal wind-circulation disturbed and disguised by local 
 or accidental causes. It is, therefore, rather unprofitable to 
 discuss such questions as the average inclination of the winds 
 to the isobars with the same detail as we have done in the case 
 of cyclones. We shall merely state a few leading features. 
 
 The characteristic circulation is, as already shown, exactly 
 the reverse of a cyclone ; in the same direction as the hands 
 of a watch, spirally outwards from the centre at the earth's 
 surface, and inwards towards it at the level of cirrus clouds. 
 Like cyclones, anti cyclones are not in general circular, but 
 
 '48. 
 73 
 
74 METEOROLOGY. 
 
 oval An area of high pressure, however, sometimes be- 
 comes very much elongated when placed between two 
 depressions. The most frequent direction of the longest 
 axis is in the United States nearly the same as in the case 
 of cyclones, N. 44 E. Over the Atlantic Ocean and Europe 
 the commonest direction is N. 75 E., nearly 40 more to 
 the eastward than that found for low pressure systems. 
 Just as in cyclones, a high pressure area when much 
 elongated has sometimes two or three distinct centres of 
 circulation ; but there is not the tendency for one of these 
 to grow in intensity at the expense of the others for which 
 cyclones are remarkable. 
 
 67. Permanent and Recurrent Anti- Cyclones. We have 
 seen ( 61) that in the North Atlantic pressure is per- 
 manently low over an area in the neighbourhood of Iceland, 
 and increases as we go southward. The maximum is 
 reached near latitude 32 N., and again decreases towards 
 the equator. There is in fact a permanent area of high 
 pressure over that ocean, usually called the " Atlantic Anti- 
 Cyclone," which varies in extent and intensity from month 
 to month in accordance with the principles explained in 
 39, attaining its greatest development in summer, and 
 least in winter. 
 
 On the south - eastern side of this system, there are 
 accordingly permanent north - easterly and easterly winds, 
 the north-east trades, extending from 30 N. lat. to about 
 7 N. lat. On its northern side, the Atlantic anti-cyclone 
 merging into the low pressure area near Iceland, gives rise 
 to the prevailing westerly winds of the North Atlantic and 
 Western Europe, and frequently penetrates over the Iberian 
 Peninsula and Southern Europe, especially during the 
 summer months, sometimes extending northwards over the 
 British Isles. 
 
 Again, over the vast land surface of Asia and Eastern 
 Europe pressure is excessive during winter, and circulation 
 
ANTI-CYCLONES. 75 
 
 anti-cyclonic. This immense system sometimes includes 
 the whole of north-western Europe, but in general is only 
 connected with the Atlantic anti-cyclone by a ridge of high 
 pressure in Southern Europe known to meteorologists as 
 the "Continental axis." 
 
 68. Relation of Anti-Cyclones to Adjacent Cyclones. 
 It appears that there are two large areas of high pressure, 
 the centres of which are practically immovable, but which 
 vary in extent and intensity according to the season of the 
 year, one being strongest and most fully developed when 
 the other is weakest ; and these combine to give a north- 
 westerly gradient directed towards a stationary low centre 
 near Iceland. We have already seen how these areas 
 govern the motions of cyclones, which tend to skirt round 
 their borders, following one another in succession, and 
 breaking into them with difficulty. It is important to 
 notice that there may thus be a number of cyclones " feed- 
 ing " one high pressure area at the same time; and further, 
 that while the one system compensates the other in the 
 matter of transferring air from one level to another, there 
 is no mutual relation in the sense of cause and effect. 
 Cyclones of great intensity are not in general associated 
 with areas of extraordinarily high pressure, and an unusually 
 high barometric maximum most frequently corresponds to a 
 minimum of only moderate depth. Loomis, 1 to whose 
 investigations this result is due, suggests as a reason for it 
 the fact, that the average breadth of high-pressure areas is 
 nearly proportional to the barometric gradient,' from which 
 it would follow that air does not flow more rapidly from a 
 system in which pressures are excessive, than from one in 
 which they are moderate. 
 
 69. Moving Anti-Cyclones. While the weather of western 
 Europe is almost completely controlled by the great high 
 
 1 Contributions to Meteorology, p. 89. 
 
76 METEOROLOGY. 
 
 pressure areas we have described, smaller systems are not 
 infrequently developed, and these are perhaps more strictly 
 the antitheses of the true cyclone than the larger ones, which 
 are really integral parts of the general circulation of the 
 atmosphere as a whole. These systems vary greatly in extent 
 and position, and move slowly and irregularly from place to 
 place, generally, however, in an easterly or south-easterly 
 direction. Loomis gives their mean velocity at 13 kilometres 
 per hour,, but they not infrequently remain stationary for 
 days together. They may indeed be treated as merely off- 
 shoots separated from the large high pressure systems, with 
 which they usually reunite sooner or later. 1 
 
 In the United States the great controlling areas of high 
 pressure are absent, and the anti-cyclones therefore stand 
 in more direct relation to depressions. Loomis found 
 that in a great majority of cases anti-cyclones were pre- 
 ceded and followed by areas of low pressure the average 
 distance from the high centre to the lo\v being 2371 miles 
 on the east side, and 2381 miles on the west, and the 
 gradient twice as steep in the former case as in the latter. 
 The average breadth of the anti-cyclones measured was 2587 
 miles, and the mean direction of their path S. 40 E. with a 
 velocity of about 34 kilometres per hour; more to the 
 southward and less rapid than in the case of cyclones. 
 
 70. Weather in Anti-Cyclones : "Radiation" Weather. 
 The weather conditions accompanying anti-cyclones are 
 determined by the descending current. The air which dur- 
 ing its previous ascent was to a great extent deprived of its 
 aqueous vapour, and probably also washed free of a large 
 proportion of its dust, is in its descent warmed by compres- 
 sion, and therefore at the same time dried. It accordingly 
 becomes very transparent to radiations of all kinds, which are 
 freely transmitted from the sun to the surface of the earth 
 or the thin layer of air lying in contact with it, and back 
 1 See Van Bebber, Lehrbuch der Meteoroloyie, p. 277. 
 
ANTI-CYCLONES. 77 
 
 again into space. Hence we have in the anti-cyclonic area 
 clear, cloudless skies, and at the earth's surface wide ranges 
 of temperature. In winter, when high pressure areas attain 
 t-ieir fullest development over land, the temperature of the 
 lowest strata of the atmosphere is extremely low. Loomis 
 found that in 74 cases over Eurasia the average temperature 
 was 28'3 C,, at 7 a.m. Washington time, corresponding 
 to the warmest part of the day ; or io'6 C. below the mean 
 temperature. The persistence of these anti-cyclones over 
 Siberia gives rise to some of the lowest temperatures observed. 
 At Verkhoyansk the average during the month of January is 
 48-8 C. Similarly, on the American continent, Loomis' 
 investigations show a mean depression of temperature i3'9 
 C. below the average, the defect being greater the higher the 
 pressure, and markedly less near the Atlantic coast than in 
 inland regions. 
 
 The effects of strong radiation under anti-cyclonic condi- 
 tions are also apparent in summer; the unusual power of 
 the sun's rays causes a great rise of temperature during the 
 day, and the heat thus received is freely parted with during 
 the night. The net result, at least in inland districts, is 
 simply to enormously increase the daily range ; the heat 
 gained during the day is lost during the night, and hence 
 on the whole anti-cyclones are not associated with tempera- 
 tures above the mean in summer although the extra 
 quantity of heat received from the sun prevents the great 
 depression below the average which occurs in winter. In 
 the neighbourhood of the sea, as in north-western Europe and 
 on the Atlantic seaboard of America, summer anti-cyclones 
 are accompanied by higher mean temperatures, probably 
 because the nocturnal radiation is checked by the presence 
 of vapour brought in by the sea breeze during the day. 
 
 It must, of course, be borne in mind that we are here dealing 
 only with a thin stratum of the atmosphere in contact with 
 the surface of the earth. While the main body of the 
 
78 METEOROLOGY. 
 
 atmosphere in an area of high pressure has its absorbing 
 power reduced, this does not hold good close to the earth- 
 where the air is not likely to contain less dust than usual, 
 and where the supply of moisture is kept up from the 
 ground or still more from the sea. This bottom layer may 
 indeed be partly regarded as the body which absorbs and 
 radiates heat. The lowering of temperature in it frequently 
 leads to condensation, hence in fine summer weather valleys 
 and low lying grounds are frequently enveloped in fog 
 during the night, especially in the neighbourhood of lakes 
 and streams where the supply of vapour is abundant. This 
 cloud-covering serves to protect the earth's surface and 
 vegetation growing on it ; for when the deficiency of vapour 
 is excessive even in the lowest layer of air the whole of the 
 heat radiated is derived from the earth itself, and we find 
 copious deposition of dew or even hoar frost, and growing 
 plants are 4f nipped " or blackened when the air temperature 
 is comparatively high. In insular and coast regions where 
 the atmosphere never becomes very dry, or where the supply 
 of rainfall is sufficient to ensure constant evaporation from 
 the soil, these "ground frosts " are rather the exception than 
 the rule, and never occur with great severity. It is obvious, 
 however, that they are more likely to occur the greater the 
 intensity of the anti-cyclonic system and the longer its 
 duration, a matter of common experience in Britain, as 
 appears from popular proverbs e.g., " Heavy dews in hot 
 weather indicate a continuance of fair weather, and no dew 
 after a hot day foretells rain." " If mists rise in low ground 
 and soon vanish, expect fine weather." 
 
 " When the mist comes from the sea, 
 Then good weather it will be." 
 
 The low lying mists and fogs are in summer dissipated as 
 the sun's rays gain power with the progress of the day, but 
 
ANTI-CYCLONES. 79 
 
 in winter while the intensity of nocturnal radiation is greater 
 the sun's heat is much less, and the mists frequently become 
 permanent. Thus in Western Europe anti-cyclonic weather 
 is in winter often characterized by cold wet mist or stratified 
 cloud sometimes resting on the ground, sometimes rising a 
 few hundred feet above it; all exposed objects are covered 
 with ice-spicules condensed from the mist, which assume the 
 most fantastic shapes, and the damp cold is extremely trying 
 to all living creatures. Co-existing with these conditions we 
 find in the upper atmosphere brilliantly clear weather with 
 temperatures above the normal, and most intense dryness, 
 characteristic of the true descending current. 
 
 These phenomena are familiar to anyone who has ascend- 
 ed a hill on a fine morning to see the sunrise. The sides of 
 the hill up to a certain level are enveloped in thick mist, 
 from which one suddenly emerges into a clear cloudless 
 atmosphere and if the peak be the highest in the neigh- 
 bourhood no land may be visible from the summit, only a 
 great undulating plain of cloud. After the dawn the cloud 
 becomes tinged with red, and the colour grows rosier and 
 warmer till the sun rises above the horizon and the whole 
 expanse becomes dazzlingly bright. Then it slowly rises 
 and disappears, to redescend in the cool of the evening. 
 
 71. Temperature Gradient in Anti-Cyclones. From a 
 scientific point of view the evidence obtained from high level 
 observatories is still more striking. During winter anti- 
 cyclones, when weather in plains and valleys is intensely 
 cold, and, if there be enough vapour, misty and wet, 
 temperature at mountain summits is far above the average 
 for the time of year, and often above that at low levels, the 
 sun shines with burning heat, and the drought is comparable 
 to that of the deserts of Sahara or Arabia. Numerous 
 examples of this have been discussed by Hann, Woeikof, * 
 and others, comparing the records of observatories such as 
 iSeel/e/. Ztit., 1892, p. 361. 
 
80 METEOROLOGY. 
 
 Santis, Pic du Midi, Puy de Dome, etc., with corresponding 
 observations at stations in plains and valleys ; the most 
 notable cases recorded are perhaps those of 23rd January to 
 3rd February 1876, and 2oth and 30th December, 1879 of 
 which the former was remarkable for mistiness and dampness, 
 at low level, and the latter for intense cold. Similar condi- 
 tions are also of common occurrence in the British Isles, 
 as is shown by the records of the Ben Nevis Observatory. 
 
 The formation of ground fogs and low clouds in anti- 
 cyclonic weather is a matter of vital importance to the 
 farmer, since they serve to protect the surface of the ground 
 from excessive cooling by nocturnal radiation. At certain 
 seasons of the year warmth and sunshine are essential to the 
 flowering or filling of grain crops, and these conditions are 
 most fully realised in an area of high pressure. But if on 
 account of long drought or other cause, the lower strata of 
 the atmosphere are excessively deficient in aqueous vapour, 
 saturation only occurs at extremely low temperatures, and 
 radiation is not checked by the formation of cloud until 
 vegetation is blighted by frost. Hence the most promising 
 crops are often destroyed by ground-frost after a season of 
 the finest weather, and the cause is often quite local, 
 occurring most frequently in soils which do not retain 
 moisture well. As an example of this we may quote the 
 following sentence by Dr. Buchan, referring to Scotland : 
 "On the night of August 3ist to September ist of 1885, the 
 potato crop was totally destroyed, and grain crops in flower 
 were also destroyed by frost in various localities in the valleys 
 of Tweed, Nith, Clyde, and Spey. Over middle and upper 
 Speyside, temperature on the ground fell to 15 F., with 
 results most disastrous to these crops ; whereas at Dalna- 
 spidal (1414 feet above the sea), in the district immediately 
 contiguous, potatoes were scarcely, if at all blackened." 1 
 We may notice in this connection the familiar saying that 
 1 Trans. jR. S. E., vol. xxxiv., p. xxiii. 
 
ANTI-CYCLONES. 8 1 
 
 when sheep go to the hill-tops at night the weather will be 
 fine, obviously because the nights are there drier and warmer 
 than in low lying ground. 
 
 72. -Fohn-winds. The peculiar phenomenon known as 
 Fohn-wind depends on principles analogous to those just 
 described. When a cyclone, or succession of cyclones, 
 follows a track in the neighbourhood of a range of mountains, 
 the air required for the supply of the ascending currents is 
 drawn from the slopes and valleys, creating a partial vacuum, 
 into which air from the higher levels descends as a hot dry 
 wind, sometimes of great violence. Since cyclones are in 
 general more frequent and more intense in the colder 
 months of the year, Fohn-winds occur oftenest in winter, and 
 since the lower strata of the atmosphere are relatively coldest 
 at that period the difference of temperature is most marked, 
 rising to 10 or 15 C. above the normal. The districts 
 most noted for these winds are the northern and southern 
 slopes of the Alps, the northern slopes of the Pyrenees, 
 the southern shores of the Caspian, the valley of the Missouri, 
 the southern island of New Zealand, and Greenland. 
 
 73. Intermediate Systems, Since cyclones and anti- 
 cyclones are in general oval, it follows that between any 
 two such systems there must be an intermediate space, for 
 which we have not as yet fully accounted. We adhere to 
 our method of description from the distribution of pressure 
 or the shape of the isobars, and although no other system of 
 fundamental importance like the cyclone or anti-cyclone is 
 met with, these areas are of considerable practical conse- 
 quence, inasmuch as they concern transition stages between 
 one great system and another. We accordingly find an un- 
 usually large number of sayings and prognostics relating to the 
 weather in these periods, especially where proximity of a 
 low pressure area is involved. These have been very fully 
 discussed for the British Islands by Abercromby, 1 and 
 
 1 Weather. International Scientific Series. (Kegan Paul & Co.) 
 
 F 
 
82 METEOROLOGY. 
 
 we shall confine ourselves chiefly to a summary of his 
 results. 
 
 74. Straight Line Isobars. Loomis remarks that between 
 an area of high pressure and a neighbouring area of low 
 pressure, the isobar of 30 inches, which may be taken to 
 represent the average of pressure between the two systems, 
 generally assumes the form of a straight line, and if these 
 areas are much elongated, it may extend with little curvature 
 for several thousand miles. We have, as a rule, a consider- 
 able area over which at least two or three straight isobaric 
 lines (supposing them to be drawn for differences of o'i in. 
 of pressure) are closely parallel to each other. In accordance 
 with laws already known to us the wind at the earth's surface 
 blows towards the lower pressure, making a nearly constant 
 small angle with the isobars, but is usually fitful and gusty. 
 On the outer limits of a straight isobar area, the weather 
 has on the one side many characteristics of a cyclone, and 
 on the other of an anti-cyclone ; but between the two, in 
 the straight isobar region proper, some peculiarities are ob- 
 served, which probably depend on the fact that in the 
 lower layers of the atmosphere warm air is being transferred 
 from the anti-cyclone to the cyclone, and acquiring moisture 
 from the earth by the way ; and in the upper layers, cold 
 air is moving in the opposite direction. Chief amongst 
 these is the formation of very " hard "-looking stratified 
 cloud, which sometimes entirely overcasts the sky, but is 
 frequently broken by chinks through which the rays of the 
 sun stream down, presenting the appearance known as 
 "streamers" or "the sun drawing water." On account of 
 vapour in the cloud strata it commonly happens, especially 
 in winter, that chiefly red rays are transmitted through these 
 openings, and the sky becomes wild and lurid. At the 
 same time the atmosphere is very clear, distant objects are 
 unusually distinct and well-defined, and the contrast of light 
 and shadow unusually sharp. It seems likely that the 
 
ANTI-CYCLONES. 
 
 S3 
 
 occurrence of these conditions in the early morning is the 
 cause of the so-called " false dawn," recorded by observers 
 from Omar Khyyam : " Before the phantom of false 
 morning died " to the negroes of North America, " If de 
 sun git up berry early in de morning, and go to bed be- 
 fore he git up, it's a sign it rain before soon." 
 
 It must be distinctly understood, however, that we do 
 not have any sequence of weather-changes, because an area 
 of straight isobars cannot progress as a whole over a par- 
 ticular spot, except in the unheard of case of a cyclone and 
 anti-cyclone moving in exactly the same direction, and with 
 exactly the same speed. Lurid chinks, " streamers," 
 " visibility," etc., associated with straight isobars, are 
 prognostics of bad weather merely because of the proximity 
 of a cyclone. 
 
 75. Wedges. When two cyclones are close together a 
 narrow wedge-shaped area of high pressure moves along 
 between them in which the isobars are as represented in 
 Fig. 19. One side of this is the rear of the retreating 
 cyclone, and the other the front of the one following it ; 
 hence in the former the 
 barometer is everywhere 
 rising, and in the latter 
 falling, and the two are 
 separated by a crest or 
 wave -line, the locus of 
 highest pressure. We 
 may, so far as weather 
 is concerned, regard the 
 two flanks of the wedge 
 as rear and front of 
 cyclones, and the wedge 
 itself as a projecting 
 
 tongue of an anti-cyclone Fig. 19. 
 
 situated to one side. Observe that along the line of 
 
84 METEOROLOGY. 
 
 the crest, where the barometer begins to fall again, the 
 wind " jumps " or shifts suddenly to almost the opposite 
 direction. Hence the saying, " A nor'-wester is not 
 long in debt to a sou'-wester." 
 
 In the rear of the retreating cyclone we have the clearing 
 up described in Chapter I. The atmosphere is, as in the 
 straight isobars, unusually transparent, and distant objects 
 look near. Hence the proverb, " The further the sight 
 the nearer the rain/' and Sir Patrick Spens' 
 
 " I saw the new moon late yestreen, 
 Wi' the auld moon in her arm ; 
 And if we gang to sea, master, 
 I fear we'll come to harm." 
 
 The central area is distinguished by the " radiation weather " 
 described under anti-cyclones : but since the two depressions 
 are always moving onward, the fin ; e weather is only tem- 
 porary, and is popularly described as " too fine to last," or 
 "a pet day." The wide end of the wedge near the crest 
 is often associated with fog, and the narrow end with 
 thunderstorms or showers. 
 
 As the second depression approaches, the weather 
 changes to that of a cyclone front halo-bearing sky, restless 
 animals, etc., but it is important to note the formation of the 
 high cirrus clouds revealing the out-flow from the south-east 
 side of the cyclone towards the high pressure. These often 
 take the form of stripes of fleecy cloud commonly called a 
 "weather-head" or "Noah's Ark." In the British Isles, 
 when the anti-cyclone lies well to the south, so that the de- 
 pression moving eastward passes over the land, the weather- 
 head will lie north-west to south-east ; but if the anti-cyclone 
 extends well to the northward so that the depression skirting 
 it is deflected to the north-east before reaching the coast, 
 
ANTI-CYCLONES. 85 
 
 the axis of the wedge, instead of lying north and south, as 
 in the first case, will be turned round towards west and 
 the weather-head will consequently extend in a north-east 
 south-west direction- Hence with reference to " Noah's 
 Ark " we find 
 
 "North and south the sign of drought, 
 East and west the sign of blast." 
 
 and amongst Scottish fishermen the N.W.-S.E. and N.E.-S.W. 
 directions are given, 
 
 76. V-Shaped Depressions. The antithesis of wedge- 
 shaped isobars occurs between two contiguous anti-cyclones 
 in the form of a V-shaped depression, which again is of the 
 nature of a tongue projecting from a cyclone situated to one 
 side. On the advancing side of the V, with falling barome- 
 ter, the weather is that of a cyclone front halo-bearing sky, 
 and increasing cloud, followed by a rain area which is 
 bounded by the line of lowest pressure along the axis of the 
 V, a line characterised by heavy squalls and clearing 
 showers, as in the trough of a cyclone. The retreating side 
 is simply the rear of a cyclone, detached clouds and then 
 blue sky, with rising barometer. Along the line of the 
 trough the wind "jumps " as in the wedge, suddenly revers- 
 ing its direction. 
 
 77. Line-Squalls and Secondaries. But V depressions are 
 not necessarily bounded on each side by anti-cyclonic 
 systems ; they are, as it were, a specialised form of cyclone 
 or of part of a cyclone, and as such frequently occur within 
 a cyclone of the ordinary or primary type. There is not, 
 however, as in the former case, the same definition of form 
 imposed ; hence the V sometimes becomes modified on the 
 one hand into a long, narrow strip, or on the other into a 
 smaller cyclone which is occasionally sufficiently developed 
 to have a complete circulation, and is then called a 
 
86 METEOROLOGY. 
 
 " secondary." The weather of such disturbances is to a 
 certain extent that of the V first described, superposed upon 
 that of the primary cyclone. The normal distribution of 
 pressure in the latter may or may not be seriously altered as 
 a whole, but gradients generally become steep and irregular, 
 and in the immediate neighbourhood of the trough (which 
 becomes the centre in a complete secondary) the wind 
 usually blows in violent gusts, rain falls in torrents, accom- 
 panied by thunder and lightning the last especially in 
 summer, at which season these irregularities are most 
 common. 
 
 In the attenuated or steep form the flanks of the V are 
 greatly diminished, and the axis or trough becomes a kind 
 of surge or wave, which is propagated at a speed varying 
 within wide limits one part sometimes moving with four 
 times the velocity of another. Such belts are usually known 
 as line-squalls or line-thunderstorms, and their energy is 
 occasionally excessive. They give rise to a large proportion 
 of the summer thunderstorms of Central Europe, 1 and are 
 common in some parts of America, but comparatively rare 
 in the British Isles, the most remarkable case amongst our 
 records being that of 24th March, 1878, which struck and 
 capsized H.M.S. Eurydice. 
 
 V-shaped depressions, secondaries, line-squalls, and 
 thunderstorms are in general most erratic in all their 
 movements, and their occurrence is, in consequence, exces- 
 sively difficult to predict. The rapidity with which they 
 sometimes travel, and the violence of rain, hail, or wind 
 often developed within them, render them a source of great 
 danger to life and property. 
 
 78. Modifications Occurring in these Systems. We have 
 now so far discussed the various forms of atmospheric move- 
 ment depending upon the distribution of pressure, and have 
 been able to connect with these most of the more important 
 
 1 See Von Bezold, Beob. d. Meteorol. Stat. in Kreiche. Bayern. 1880. 
 
ANTI-CYCLONES. 8/ 
 
 phenomena of weather. We must, however, remember that 
 the circulation in all the systems described up to this point, 
 is determined by the deflecting force due to the rotation of 
 the earth, acting according to Ferret's law. It must be 
 understood that, in describing these atmospheric move- 
 ments, the average general path only is concerned, and that 
 small local disturbances and eddies have not been con- 
 sidered. Thus, for example, if we state that in cyclones 
 the direction of the wind makes a given angle with the 
 isobars, we mean that the statement is true on the average 
 of a large number of cases, and more nearly true for any one 
 case the greater the intensity of the cyclone for in a deep 
 depression the forces due to the gradients are greater in 
 proportion to local disturbing forces. In anti-cyclones, 
 again, the gradients are, as a rule, so slight that large parts 
 of the area are often distinguished by calms, or very light 
 variable winds ; but on the whole there is no doubt about the 
 general circulation. For if we take a sufficiently large num- 
 ber of cases extending over a sufficiently large area it is 
 clear that local disturbances or distortions are just as likely 
 to take place in one direction as in another, and hence they 
 will, in the long run, balance each other, and leave the real 
 circulation as the " most probable " result. 
 
 We are not at present concerned with the direct effects 
 produced in special districts by the conformation of the 
 earth's surface ; these fall more properly under the heading 
 of Climate^ since their effect is constant ; but we may briefly 
 consider a few of the more important modifications which 
 occur in the larger pressure systems. 
 
 79. Squalls, Derechos. First comes the simple squall, 
 which varies in intensity from the puff of wind whose pro- 
 gress is marked by " catspaws " on the surface of the sea, to 
 the Bden of Germany, or the derechosj- the "straight 
 blow of the prairies " of America. In their fully developed 
 1 See Hinrichs, Amer. Met. Jour., vol v., p. 306. 
 
88 METEOROLOGY. 
 
 state, squalls are remarkable by the presence of a dense 
 black cloud at no great elevation, which may be seen 
 approaching sometimes so quickly that it is impos- 
 sible to shorten sail in the time between the first obser- 
 vation of the cloud on the horizon, and the ship bein^ 
 struck. The cloud, as seen from the advancing side, is in 
 general of an arched form hence the name " arched 
 squall," and under the arch the wind blows in gusts, some- 
 times of great violence, and usually accompanied with 
 deluges of rain, sleet, or hail, and much thunder and 
 lightning. Within the squall pressure is greatly in excess, 
 the barometer rising as much as i mm. or more, and falling 
 again as the squall passes off. If the weather preceding 
 the occurrence of the squall be warm and sunny, temperature 
 usually suffers a sudden considerable fall but if cloudy, 
 little change is observed. Taking these facts together, and 
 along with them the fact that the direction of wind in 
 squalls is usually that of the lower cloud motion, there can 
 be little doubt that squalls are caused by sudden down- 
 rushing of the air in the upper strata : and it seems probable 
 that this is due to a sudden heavy fall of rain, bringing 
 quantities of air down with it, simply by mechanical friction. 
 The air thus brought down is dynamically warmed and 
 dried, hence, in thunder showers of this type, it is not un- 
 common to find the air comparatively dry, even during a 
 downpour of rain. 
 
 How the sudden excessive condensation arise.? is not so 
 obvious, unless we accept the suggestion of Von Bezold, 
 already referred to, that a mass of air may become super- 
 saturated through the want of dust nuclei. Another ex- 
 planation is afforded by supposing an extremely " stable " 
 vertical gradient of temperature, such that cold, dense, 
 upper layers of air break through lower warmer layers lying 
 underneath, condensing the vapour in the latter, and causing 
 them to ascend thereby still further favouring the descent 
 
ANTI-CYCLONES. 89 
 
 of cold air. It would seem that some process of this kind 
 sometimes does take place without giving rise to condensa- 
 tion, as in the case of what are called " white squalls," 
 which are occasionally sufficiently violent to dismast or even 
 capsize vessels. 1 Captain Toynbee remarks that these 
 squalls most frequently come from the direction of the 
 upper current thus at the northern limit of the S.E. trade 
 wi.ids, they often come from north-east, moving with the 
 upper clouds, although the surface wind is S.E. or S. 
 " White squalls " are not accompanied by rain or electrical 
 disturbance violent gusts occur suddenly without warning 
 in fine weather. 
 
 This theory would also explain the squally, gusty winds 
 characteristic of the trough and rear of eastward-moving 
 cyclones ; the sudden intrusion of cold polar winds over an 
 area which has just experienced warm southerly or equatorial 
 winds causing great condensation by the mixture of the 
 two. 
 
 80. Tornadoes. In contrast to the various kinds of 
 squall attributable to a descent of upper currents, we have 
 the sudden ascent, under similar conditions, of masses of 
 warm air. This phenomenon, in its various degrees of 
 intensity, gives rise to the familiar dust-whirls on a dry road, 
 to the dust-storms and water-spouts of the tropics, and to 
 the destructive tornado or "twister of the prairies." These 
 whirlwinds are quite distinct from cyclones, inasmuch as 
 their height greatly exceeds their diameter, and compared 
 with cyclones, their diameter is extremely small ; the smallest 
 known cyclone extends over many times the area covered 
 by the largest tornado, and there is no record of any 
 phenomenon of an intermediate type. It is obvious that 
 the rotation cannot therefore be due to the motion of the 
 earth, and it is probably caused by accidental or local 
 
 1 See Jahncke, Q. J. Met. Soc., April, 1874. 
 
9O METEOROLOGY. 
 
 deflection of the air in its movement towards the centre ; 
 the "principle of equal areas," referred to in connection 
 with cyclones, sufficiently explains how a very rapid 
 rotation may be originated by a quite insignificant initial 
 deflection, and it is easy to understand that air moving 
 over the earth's surface is not likely to preserve an ab- 
 solutely straight path for any length of time. Hence we 
 find that in this class of disturbances, where an ascending 
 current is involved, the rotational movement is a much 
 more important element than in the squalls or " derechos " 
 where the air has descended and escaped the rotational im- 
 pulse due to surface irregularities. Indeed the increased 
 wind-velocity in the squall is probably to a large extent 
 simply the velocity of the upper cu;rent, which is normally 
 greater than that of the under current, unimpeded by friction 
 with the earth's surface. 
 
 The conditions of instability most favourable to the 
 occurrence of whirlwinds differ to some extent from those 
 most commonly associated with squalls. In the latter case 
 it would seem that the process is chiefly mechanical, air 
 being either dragged down by falling rain, or forcing itself 
 down by its own inertia ; while in the former the action par- 
 takes more of the nature of a convection current. Hence 
 squalls are most frequent where cold air is brought into 
 contact with warm through the independent action of dis- 
 tribution of pressure, as in the trough and rear of a cyclone, 
 but whirlwinds are most commonly experienced in the front 
 of shallow depressions and during the warmest hours of the 
 day, where there is little or no wind and the weather is 
 sultry and oppressive. Thus we have in this country the 
 well-known prognostic of dust-whirls preceding rain, and in 
 the great tornado-regions of the United States the death- 
 like stillness and silence is one of the surest precursors of 
 these storms. 
 
 In this case, however, as in that of cyclones and anti- 
 
ANTI-CYCLONES. 9 1 
 
 cyclones, the question of the cause of ascending and descend- 
 ing currents is one of great difficulty, and has not yet re- 
 ceived an adequate answer. 
 
 Waterspouts are frequently seen on the Atlantic. A low 
 cloud of great density is seen to bulge downwards on its 
 lower side in the form of an inverted cone, and at the same 
 time the surface of the sea underneath is thrown into violent 
 commotion as if boiling. The conical cloud continues to 
 descend, and the boiling sea appears to rise up to meet it, 
 till the two columns join, and the spout appears suspended 
 from the main body of the cloud above. 
 
 The apparent descent of the cloud in a waterspout is 
 caused by the great reduction of pressure by centrifugal 
 force within the whirling vortex, which allows the vapour 
 inside the column to condense. When the rotation dim- 
 inishes and the spout is about to bre^ik up, the pressure 
 again increases, and the cloud evaporates, the column 
 becoming transparent during the process. Waterspouts are 
 usually succeeded by torrents of rain. 
 
 Dust or sand storms are of constant occurrence in the 
 deserts of Africa and Arabia. They are simply spouts in 
 which the waters of the sea are replaced by the sand of the 
 desert. In the absence of either, the whirlwind becomes a 
 tornado. 1 
 
 The tornado is best known as the terrible scourge which 
 devastates certain regions in the United States notably 
 Missouri, Kansas, and Georgia. It occurs at all times of the 
 year, but most frequently in May ; and always in the south- 
 east quadrant of an area of low pressure, generally at a 
 distance of 300 to 500 miles from the centre. Preceded 
 by a continuance of southerly winds and gradually increas- 
 ing heat and moisture, or by a sultry oppressive atmosphere, 
 the formation of the tornado begins with "the sudden appear- 
 
 1 See Finley, Report on Six Hundred Tornadoes ; also Amer. 
 Met. Jour., vol. vii., (1890), p. 165 
 
2 METEOROLOGY. 
 
 ance of ominous clouds, first in the south-west, then almost 
 immediately in the north-west and north." " If the clouds are 
 light, they resemble smoke rising from a burning building ; 
 if dark, they present a deep greenish hue, which appears to 
 increase in intensity as the storm advances. The motions 
 of the clouds are peculiar, in that they appear to be rushing 
 from every quarter towards a common centre, marking the 
 incipient stages of a gyratory motion in the cloud region." 
 Then follows the formation of the funnel-shaped cloud. The 
 development of the storm is sometimes arrested here, and 
 the only result is a severe hailstorm ; sometimes, again, it is 
 confined to atmospheric layers at some elevation, and the 
 tornado can be seen pursuing its course overhead with but 
 little effect at the earth's surface : and even tornadoes which 
 do reach the ground sometimes rise and redescend, leaving a 
 track interrupted here and there by a series of breaks in 
 which no damage is suffered. 
 
 Concerning tornadoes which do reach the ground, Finley 
 states that the length of the path of destruction varies from 
 300 yards to about 200 miles, its average width being 
 about 450 yards. The prevailing direction cf progressive 
 movement is north-east, with an average velocity of 44 miles 
 an hour, which, however, varies within the wide limits 7 and 
 100. It would b^ easy to multiply harrowing descriptions 
 of the destruction caused by these storms, of trees uplifted 
 and driven down into the ground like stakes, of houses ex- 
 ploded by the air in their rooms expanding in the partial 
 vacuum of the central whirl. But the final statements by 
 Finley are sufficiently suggestive : " No structure that rises 
 above the earth, however made, can wholly resist the viol- 
 ence of the tornado." " No building can be made suffi- 
 ciently large, strong, high, or low, or of any material whatever, 
 to resist the force of the tornado's vortex." The approach 
 of the tornado is accompanied by a noise resembling that of 
 heavy machinery, which increases in intensity till the final 
 
ANTI-CYCLONES. 93 
 
 crash, which comes with the suddenness of an explosion. 
 The average duration of the storm at any one point is little 
 over a minute, but there is no way of escape except in " caves 
 and holes in the earth." 
 
 8 r. Prediction of Squalls and Tornadoes. The special 
 prediction of squalls and whirlwinds is obviously, from the 
 nature of the case, an impossibility. We can predict that a 
 given rock by the. seashore will be covered by the tide on 
 certain days and at certain hours, but it is hopeless to pre- 
 dict how the advancing water will swirl and eddy round each 
 projecting angle or loose stone. All that can be done is to 
 give due warning of the unstable state favourable to squalls 
 and whirls ; the rest must simply be watched for on the 
 spot. 
 
CHAPTER V. 
 
 WEATHER FORECASTING. 
 
 82. Methods of Forecasting. It will be understood from 
 the facts outlined in the preceding chapters that the best 
 method of predicting weather over any considerable extent 
 of country is to study the general conditions of pressure, 
 temperature, wind, etc., at a particular time, and then, by 
 the help of past experience, to form a conjecture as to the 
 changes most likely to follow. From what we know of the 
 motions of cyclones and anti-cyclones, of their changes of 
 form and intensity, the sudden development of secondaries 
 and squalls, and so forth, and from our profound ignorance 
 of the causes producing these changes, it is evident that 
 anything like reliable prediction can only be attempted for a 
 very short period. An expert forecaster can sometimes form 
 a fairly accurate notion of the sequence of weather for two 
 or three days in advance, but in practice it is found impos- 
 sible to issue regularly anything like trustworthy predictions 
 for much more than twenty-four hours. We may say that, 
 given the synoptic chart representing the conditions over a 
 certain area, say at 9 a.m. on a particular day, it is possible 
 to predict what will be the general features of the chart till 
 the same hour of the following day, with, under favourable 
 conditions, a likelihood of being right five times out of six. 
 
 Under these circumstances, it is necessary for a forecast- 
 ing service in any country that material for constructing a 
 synoptic chart should be collected at a central office at 
 least once daily, and since forecasts based on such charts 
 
 94 
 
WEATHER FORECASTING. 95 
 
 are only valid for about twenty-four hours from the time at 
 which the observations at the various stations are made, it is 
 obvious that no time must be lost; for the time occupied in 
 tabulating and reducing the observations, telegraphing the 
 results to headquarters, plotting the data from all stations 
 on the chart, drawing isobars, isothermals, etc., deducing the 
 forecast from the indications of the chart and transmitting 
 the forecast to the districts concerned, must all be deducted 
 from the time for which the prediction holds good. 
 
 83. Forecasting by Synoptic Charts. It is difficult, for 
 obvious reasons, to lay down any hard and fast rules for 
 predicting weather by means of charts. Experience alone 
 can make a successful forecaster, and even the most expert 
 cannot always assign full reasons for his predictions, so 
 much is it a matter of simple judgment. We may, however, 
 indicate a few general principles. 
 
 The first point to be looked at is the extent, form, and 
 intensity of the regions of highest pressure. For, in the first 
 place, where we find anti-cyclonic conditions, changes take 
 place most slowly, and again, the high pressure areas to a 
 great extent control the paths of the cyclones. By com- 
 parison with previous charts we can ascertain whether the 
 anti-cyclones are extending and increasing in intensity, or 
 retreating and breaking up. In the former case we can 
 predict for the anti-cyclonic areas continued " radiation " 
 weather, light winds, warm days, arid dewy nights in summer, 
 and cold or fog in winter ; special attention being paid to 
 the probability of ground-frosts likely to damage crops. In 
 the latter case, we must observe carefully the breaking up of 
 the high pressure system, noting the chances of its being 
 penetrated by cyclones, and its effect on the cyclone tracks 
 in its neighbourhood. Further, we must, when possible, 
 determine the relations of the high pressure areas to the 
 general circulation. Thus, for example, in Western Europe 
 it is important to know when an anti-cyclone is a part of the 
 
96 METEOROLOGY. 
 
 Atlantic high pressure system, or of the Eurasian, whether it 
 is merely a ridge connecting the two, or a small system dis- 
 tinct from both. This is technically known as determining 
 the type of weather, for it is found that the arrangements of 
 the high pressure areas can be classified into a number of 
 distinct types according to the relative extentand intensity 
 of the different elements, and each of the great national 
 weather offices employs a classification suitable to its 
 special requirements. 
 
 The main features of the high pressure areas determined, 
 the next point is to look for cyclones. These tend to 
 follow certain routes in succession ; hence, by referring to 
 former charts, we note the tracks of previous cyclones, and 
 inquire whether, on account of alterations in the anti-cyclonic 
 areas or other cause, we have any reason to expect that suc- 
 ceeding depressions will take a different path. In this con- 
 nection we may make use of, for example, Van Bebber's 
 cyclone tracks, or for the British Isles we may give weight to 
 the fact that cyclone centres have a tendency to move up 
 the English Channel, round the north-west coasts of 
 Scotland and Ireland, along the line of the Caledonian 
 Canal, or the Forth and Clyde Canal, or, more generally, 
 that they select the line of least resistance. This done, we 
 seek for signs of a cyclone actually approaching, reports of 
 cirrus clouds, halos, sun-dogs, rising temperature, falling 
 barometer, and so on ; and if we find any such, we obtain 
 all possible data for forming an opinion as to its size, depth, 
 position, direction of path, and rate of motion. This is, of 
 course, the most important part of all forecasting, for we 
 must ascertain whether the depression is of sufficient energy 
 to merit the name of a storm, and the districts likely to be 
 endangered must have timely warning. After its arrival, the 
 cyclone must be carefully watched, and the districts over 
 which it passes kept informed ?s to the changes of wind and 
 weather corresponding to their position with respect to the 
 
WEATHER FORECASTING. 97 
 
 centre. We must further determine whether the system is 
 increasing in depth or filling up, and be on our guard 
 against sudden changes of direction or speed, which would 
 falsify predictions and leave certain coasts unwarned. Again, 
 a constant look-out must be kept, especially during the 
 passage of shallow depressions in the warmer months of the 
 year, for the formation of V depressions or secondaries, or 
 rather for conditions favourable to their development. 
 This is at once the most difficult and one of the most im- 
 portant points to be attended to in order that some warn- 
 ing may be given of squalls or severe hail and thunder 
 storms. 
 
 As the cyclone passes off, we look carefully for indica- 
 tions of another following it, a check in the fall of 
 temperature accompanying the rear, the unusual clearness, 
 "refraction," or radiation characterising a "pet" day, the 
 reappearance of cirrus clouds, or a check in the rise of the 
 barometer, with a backing wind. Here the British Isles stand 
 at a very great disadvantage in the matter of forecasting, for 
 cyclones almost invariably approach from the westward, and 
 travel with considerable velocity. But immediately to the 
 west is the Atlantic, from which we obtain no sign the 
 first notice we can receive of an approaching storm is from 
 Valentia, only 400 miles west of London ; hence a rapidly 
 moving storm may be upon us almost before it is possible 
 to give warning. 
 
 84. Storm Warnings. The work of storm warning is in 
 practice to a great extent distinct from that of ordinary con- 
 tinuous forecasting. In the British Meteorological Office, 
 on the approach of a cyclone of dangerous intensity, special 
 messages are sent out warning the coasts likely to be affected. 
 The warnings thus received are made public by the exhibi- 
 tion of certain signals. In this country these take the form 
 of a cone, which is hoisted at seaports on receipt of the warn- 
 ing message. When hoisted with its apex downwards, the 
 
 G 
 
98 METEOROLOGY. 
 
 cone signifies that strong winds may be expected, at first 
 from southward or south - westward. 
 Hoisted apex upwards, it means that 
 strong winds from northward or north- 
 eastward may be expected. These sig- 
 nals are at best merely a caution of the 
 most general kind, and further details of 
 the approaching disturbance must be 
 CONE" obtained from the warning telegram itself, 
 Fig. 20. which is (or ought to be) always posted 
 
 up for public inspection. 
 
 The student who has fairly mastered the contents of 
 the preceding chapters will find a study of the Daily 
 Weather Reports of the Meteorological Office, or of 
 the Deutsche Seewarte, the most profitable meteorologi- 
 cal exercise to which he can address himself. Let 
 him take a chart for a given day, and after comparing 
 it with those for say a week previous, attempt a forecast 
 for the twenty-four hours following ; after which let 
 him compare his forecast with that issued by the office, and 
 both with the report of actual weather issued the following 
 day. Where either or both of the forecasts is found to have 
 failed, let him carefully inquire in what the failure principally 
 consists temperature, wind, rain, etc., and seek for hints 
 to be useful in future. 1 
 
 85. Forecasting for Isolated Observers. We have seen 
 that the forecaster at a central station labours under 
 great disadvantages from the unavoidable incompleteness of 
 
 1 The British Daily Weather Reports may be obtained from the 
 Secretary, Meteorological Office, 63 Victoria Street, S.W., London, 
 post free as issued for ^"i per annum. For further practical information 
 in forecasting see Abercromby, Weather, International Scientific 
 Series, 1887, or Principles of Weather Forecasting also by Aber- 
 cromby, published by the Meteorological Council (Stanford, 1885). 
 Also R. H. Scott, Weather Charts and Storm Warnings, jrd edition, 
 Longmans, 1887. 
 
WEATHER FORECASTING. 99 
 
 the data from which he has to work ; and that at best his pre- 
 dictions can only be of the most general character. But on 
 the other hand the practical man, who wants to foretell the 
 weather at the particular station at which he lives and moves, 
 labours under the disadvantage of not being able to get the 
 general view afforded by the synoptic chart. Hence the 
 sailor on board ship, and the fisherman or farmer in remote 
 districts, must depend wholly on his own observations to 
 furnish him with the means of prediction. We propose now 
 to give some hints which may be helpful in applying the 
 principles already explained, premising again that nothing 
 can replace long and careful observation many an old salt 
 blissfully ignorant of all scientific meteorology is nevertheless 
 a practical forecaster of the greatest proficiency. 
 
 First, and most important, we have for the northern hemi- 
 sphere Buys Ballot's rule, for enabling us to form a notion 
 of the distribution of pressure : 
 
 "Stand with your back to the wind, and the centre of 
 lowest pressure lies about two points in advance of your left 
 hand." 
 
 We shall not in general find much difficulty in recognising 
 anti-cyclonic weather the high barometer, the direction and 
 force of wind, and the te radiation " effects described by so 
 many prognostics, are unmistakable. It is then only neces- 
 sary to keep a strict watch on the progress of an anti-cyclone, 
 so as not to miss the first signs of its breaking up. During 
 a spell of fine weather the barometer rises, perhaps a little 
 irregularly, but on the whole steadily, till a maximum is 
 reached, after which it begins to fall again slowly ; showing 
 that the anti-cyclone is either retreating or breaking up. 
 This gradual dissolution may go on for days, with little 
 change in the weather beyond a slightly increased cloudiness, 
 but the first indications of an approaching cyclone must be 
 looked for, a sudden increase in the rate at which the 
 mercury falls any of the signs of increased humidity de- 
 
IOO METEOROLOGY. 
 
 scribed in the prognostics or the appearance of " mares' 
 tails," or windy cirrus clouds ; the direction and rate of 
 motion of these last being of great importance. Cirrus 
 moving rapidly from west or north-west may be accepted, in 
 Britain at least, as a very reliable warning. 
 
 After satisfying ourselves in various ways that a cyclone is 
 actually approaching, the next step is to find the position of 
 its centre, its direction and rate of motion, and its extent. 
 This can usually be done with a fair approximation to accuracy 
 by a careful study of the changes of the wind, with the help 
 of Buys Ballot's Law and constant observation of the 
 barometer ; a close watch being kept all the time on the 
 appearance of the sky. We note that when the barometer 
 falls rapidly and the direction of wind does not change, we 
 may expect the centre to pass directly over us ; when the 
 barometer falls rapidly and the wind veers quickly, we are 
 near the path of the centre, of a deep depression if the winds 
 are violent, and of a rapidly moving one if they are only of 
 moderate strength ; and when the barometric fall is gradual 
 and the change of wind direction is slow, we are at a consider- 
 able distance from the path of the centre, the wind again 
 veering faster the faster the depression is travelling. 
 
 The next change to be looked for is the passage of the 
 line of the trough marked by the barometer rising again- 
 after which, of course, we may predict the weather charac- 
 teiistic of the rear of a cyclone. Observe, however, that all 
 these rules for a single isolated station may be greatly modi- 
 fied by a cyclone of elongated form ; the changes at the two 
 ends are then very sudden, and very gradual in the inter- 
 mediate parts. 
 
 As before, the tendency of cyclones to follow each other 
 in succession must be kept in mind, and after one passes off 
 the conditions associated with a " wedge " or K straight 
 isobar " must be examined for the smallest indications of 
 another depression. 
 
WEATHER FORECASTING. lOl 
 
 It is unnecessary to go into further detail, which would 
 simply involve repetition of much that has been already ex- 
 plained. We may simply give as general cautions : 
 
 (i). Although cyclones occur most frequently when 
 barometric pressure is low generally, and the converse, the 
 changes from time to time are the most important factors to 
 be observed. The words "rain," " change," " fair," en- 
 graved on the barometer are always misleading, and alto- 
 gether wrong when the instrument is used at any place other 
 than that at which it was constructed, on account of the 
 change of level. 
 
 (2). In general the rule 
 
 " Long foretold, long last, 
 Short notice, soon past," 
 
 holds good. 
 
 (3). A southerly wind is much more likely to veer to the 
 northward with the sun than a northerly to the southward. 
 
 (4). All the ordinary interpretations of barometric indica- 
 tions, sky appearances, wind changes, etc., are liable at any 
 time, but especially in summer, to be falsified by a sudden 
 change of course in a cyclone, or by squalls, secondaries, 
 V depressions and such disturbances. 
 
 A good deal of help may often be obtained from local 
 prognostics. On the west coast of Ireland and Scotland a 
 heavy ground swell or breaking surf is accounted a sure 
 precursor of a storm, because the sea disturbance oc- 
 casioned by a cyclone frequently outstrips the cyclone itself. 
 Again, we have innumerable local prognostics depending on 
 humidity, visibility, etc. In Lancashire, we find the saying, 
 
 " If Riving Pike do wear a hood, 
 Be sure the day will ne'er be good ; " 
 
 and at Plymouth, if from the Hoe the Breakwater light 
 (2 miles distant) and the Eddystone light (14 miles distant) 
 
102 METEOROLOGY. 
 
 seem of the same colour or brightness, it is a sign of rain next 
 day. 
 
 The most successful forecasting may, of course, be ac- 
 complished by combining local observation with a study of 
 synoptic charts. The knowledge of general conditions de- 
 rived from the latter can then be modified to suit local 
 peculiarities. We know, as it were, what to look out for, 
 and can at once recognise the earlier stages of the different 
 changes. 
 
CHAPTER VI. 
 
 METEOROLOGICAL INSTRUMENTS AND OBSERVATIONS. 
 
 86. Need for more Accurate Measurements. We have 
 found that, as stated at the outset, it is possible to observe 
 many of the meteorological elements with sufficient accuracy 
 for describing weather changes, without the aid of in- 
 struments. Up to this point, we have employed only the 
 barometer to measure changes of pressure otherwise imper- 
 ceptible, and the thermometer to enable us to distinguish 
 temperature effects from those of moisture. 
 
 But when, instead of the main characteristics of change 
 from day to day, we come to consider the average result of 
 all the variations from month to month and year to year, to 
 compare the weather of say, one summer with that of 
 another or with the average of a long succession of summers, 
 or to contrast the average weather of different places during 
 the same periods, then these methods become inadequate. 
 We must increase the accuracy of our observations, and by 
 instrumental or other aids transform them into quantitative 
 measurements which can be compared amongst themselves, 
 and subjected generally to the ordinary methods of mathe- 
 matical treatment. 
 
 The making of meteorological observations presents for 
 the most part no great difficulty the essential qualification 
 is merely a capacity for doing a small piece of routine work 
 at stated times without losing interest in it, and so becoming 
 careless. 
 
 87. Observations must be Comparable. In observational 
 work connected with climatology one point must always be 
 kept in view that the principal value of the records consists 
 in their bearings on those of neighbouring stations, The 
 first essential, therefore, is that the observations at different 
 
 103 
 
104 METEOROLOGY. 
 
 places must be comparable, the instruments used must be 
 similar in form, and must be exposed in a similar way, and 
 the errors peculiar to them and to the observer must be 
 known. Thus in some cases it is not advisable to use the 
 best known instrument for observing a certain phenomenon, 
 because although some observers may have skill to use it, 
 many may not, and the few sets of more accurate observa- 
 tions are not comparable with those made at the majority of 
 stations. We shall confine ourselves to describing instru- 
 ments and methods in general use in this country, and we 
 may state generally that all instruments should have their 
 errors of construction determined by comparison with the 
 recognised standards at Kew Observatory. 
 
 88. Observations of Air Temperature. We have already 
 discussed the variations of air temperature with some detail, 
 and are therefore in a position to understand the nature of 
 the quantitative measurements required. These are, in 
 general, the mean daily temperature and the daily extremes. 
 The latter are measured by self- registering maximum and 
 minimum thermometers,, and the former either by averaging 
 a number of observations at fixed hours, or by applying a 
 correction to the mean of the maximum and minimum. 
 (See 35.) 
 
 Now, it must be remembered that with the exception of 
 pressures, meteorological observations are not measurements 
 of precision like, for example, electrical measurements. In 
 determining air temperatures or humidities or the direction 
 and force of winds, we are not concerned to make measure- 
 ments of great accuracy which refer only to the air in the 
 immediate neighbourhood of the instruments at the moment 
 of observation. What we want is a fairly correct estimate of 
 the general conditions. 
 
 Take one example. If we observe air temperatures with 
 a thermometer of great sensitiveness, fully protected from 
 radiation effects by a silvered thimble on the bulb, we shall 
 
METEOROLOGICAL INSTRUMENTS. 10$ 
 
 find constant variations from minute to minute, sometimes 
 amounting to a degree or more, whether the air is calm or 
 not. 1 Hence it appears that there are incessant local 
 variations, and in order to get the average temperature near 
 a given station, say at 9 a.m., we should take readings every 
 minute from say five minutes before the hour till five 
 minutes past, and find the mean. But in practice this is 
 much more easily done by using a less delicate thermometer, 
 which, as it were, averages the small variations for itself, and 
 gives the result of all the minute rises and falls of tempera- 
 ture to which it is exposed. 
 
 89. The Thermometer. The thermometer most usually 
 employed has a spherical bulb J to f inch in diameter, 
 although some observers prefer a cylindrical form ; and 
 its accuracy is tested to o'i (F. or C.). Note that what- 
 ever be the form of the attached scale which may be of 
 wood, porcelain, or zinc the degrees must be engraved 
 on the stem, otherwise we have no guarantee that the zero 
 points do not shift. The scale most commonly used in this 
 country and in America is that of Fahrenheit, of which the 
 chief merits are that it does not involve the use of minus 
 quantities for temperatures below freezing, and that the de- 
 grees are of convenient length. On the other hand, it has 
 the disadvantage of requiring to be converted into the cen- 
 tigrade scale whenever any theoretical investigations or com- 
 parisons with Continental work are involved. The matter 
 is, however, unimportant, as the difficulties presented by the 
 conversion are not formidable. This and all other meteoro- 
 logical computations are best performed by means of tables, 
 to be found in the official Instructions in the Use of 
 Meteorological Instruments of the Meteorological Office, 
 or Hazen's Tables, or the somewhat bulky edition of 
 Guyot's Tables issued by the Smithsonian Institution. 
 
 90. Reading the Thermometer. In reading a thermometer 
 
 1 Aitken, Proc. #. 8. JE. t xii., p. 687. 
 
106 METEOROLOGY. 
 
 the principal error to be guarded against is that due to 
 parallax, or not having the eye directly opposite the point 
 at which the liquid stands. When the liquid in question is 
 mercury, this is easily avoided by making the degree mark 
 on the stem coincide with its reflection in the mercury column. 
 The limit of accuracy required in the readings is one- 
 tenth of a degree (o'i). The marks on the stem usually re- 
 present whole degrees, and the tenths are obtained by esti- 
 mating the position of the top of the column with reference 
 to the degrees above and below. It may be remarked that 
 the five degree marks, and frequently the freezing-point 
 mark, are longer than the others, and care must be taken in 
 determining the whole degrees of the reading not to confuse 
 these longer marks with each other. The most careful ob- 
 server will sometimes record 42 or 34 for 37; he sees 
 that the reading is 2 above a long mark, and mistakes the 
 35 mark for 32 or 40, and so on. 
 
 91. Self-registering Thermometers. The forms of self-regis- 
 tering thermometers are numerous. We refer the reader to 
 larger treatises for detailed descriptions. The maximum ther- 
 mometer most usually employed is that of Phillips, which is 
 merely an "ordinary" mercurial thermometer with its stem 
 placed horizontally, and the mercury column broken by a 
 small bubble of air. 
 
 The minimum thermometer in almost universal use was 
 devised by Rutherford in 1794. The liquid employed is 
 alcohol, and in the alcohol column is placed a small glass index. 
 
 92. Thermometer Exposure. The proper exposure of 
 thermometers so that they may indicate the true tempera- 
 ture of the air is a matter of great difficulty. The conditions 
 required are two a constant circulation must be kept up 
 round the thermometer bulbs, and in its passage to the in- 
 struments the air must not have its temperature changed by 
 passing over hot or cold surfaces, and the thermometer 
 bulbs must be protected not only from the direct rays of 
 
METEOROLOGICAL INSTRUMENTS. 
 
 lO/ 
 
 the sun. but from radiations of all kinds from surrounding 
 objects. These conditions are probably most nearly realised 
 by the sling thermometer, which is attached to a cord some 
 two feet in length and swung round like a sling, the opera- 
 tion being of course performed in the shade. There are, 
 
 Fig. 21. 
 
 however, many objections to observations of this kind, and 
 various forms of thermometer shelter have been devised, 
 none of them altogether satisfactory. That used in this 
 country is called after its inventor the " Stevenson " screen, 
 and except perhaps on unusually calm hot days gives fairly 
 
IOS METEOROLOGY. 
 
 accurate results. It consists merely of a wooden box, as 
 shown in Fig. 21, the sides of which are composed of a 
 double row of louvre boards placed opposite ways. The top 
 is solid and the bottom entirely open. 
 
 The screen should be placed on a level grass plot of con- 
 siderable extent, and well clear of buildings, trees, rising 
 ground, or anything likely to interfere with free circulation 
 of air. It should be mounted on legs so that the bulb of 
 the ordinary thermometer is just 4 feet from the ground. 
 
 93. Calculation of Mean Temperature. As already ex- 
 plained, if we accept the average of twenty-four hourly 
 observations of temperature as the true mean for the day, 
 we can obtain a close approximation to this value for any 
 given day without actually making the hourly observations, 
 either by observing the temperature at suitable hours, or by 
 applying a proper correction to the mean of the maximum 
 and minimum. We give the following hours as affording 
 the best averages 9 a.m. and 9 p.m. ; 10 a.m. and 10 p.m. * 
 3 a.m. and 3 p.m. ; 6 a.m., 2 p.m., and 10 p.m. ; any four 
 hourly intervals. The mean of maximum and minimum 
 may be taken as about o'5 F. above the average temperature 
 for nearly all climates, but more accurate results may be ob- 
 tained for the British Isles by employing the following 
 method. " Multiply 1 the difference between the observed 
 maximum and minimum by the proper factor obtained from 
 the following table, and add the product to the minimum " 
 
 Month. Factor. 
 
 January and December, . . 0*520 
 
 February and November, . . 0-500 
 
 March and October, . . . 0*485 
 
 April and September, . . . 0*476 
 
 May and August, .... 0*470 
 
 June and July, .... 0*465 
 
 94. Accumulated Temperature. We shall have occasion 
 later on to make use of another quantity deduced from 
 1 Weekly Weathzr Report, 188}, p. 3. 
 
METEOROLOGICAL INSTRUMENTS. 109 
 
 temperature observations the accumulated temperature. 
 It has been found that when the air is warmer than a certain 
 neutral temperature, the progress of vegetation is accelerated, 
 and hence it becomes important, having fixed that point, to 
 know how much and for how long the temperature is above 
 or below it. The investigations of A. de Candolle have 
 shown that in the case of most plants vegetation does not 
 actively commence till a temperature of somewhere about 
 6 C. or 42 F. is reached, and this temperature is therefore 
 accepted as the base. But now there arises the difficulty 
 of computing in a convenient way the amount of heat at a 
 temperature above 42 F., received by a plant. We want a 
 means of expressing the amount by which the temperature 
 exceeded 42 F., and the time it continued above that point. 
 
 This problem has been solved by General Strachey's con- 
 ception of a "day-degree." One day-degree is a difference 
 of i F. from the base temperature continued for 24 hours. 
 Thus an average temperature of 43 F. for 24 hours means 
 one day-degree positive, or to the good for vegetation ; an 
 average of 41 for the same period means one day-degree 
 negative, or lost. Hence for each kind of crop we can be- 
 gin at seed-time, add all the positive day-degrees, omit all 
 the negative, and estimate the total required to ripen the crop. 
 
 We are at present only concerned with the practical 
 question of calculating the number of day-degrees for a given 
 period. The following rules are taken from the Weekly 
 Weather Report for 1884: 
 
 Rules for computing for a Weekly Period the Accumulated 
 Temperature above or below 42 F. from the observed 
 Maximum and Minimum. 
 
 1. Obtain the mean temperature, from the means of the 
 seven observed maxima and minima, in the manner already 
 described for daily values. ( 93.) 
 
 2. In obtaining the accumulated temperature four cases 
 may occur, to which the following rules will apply : 
 
no 
 
 METEOROLOGY. 
 
 
 
 
 C 
 
 s E , 
 
 s 
 S|.4 
 
 |1 |1 
 
 rj- 
 
 1 
 
 I 
 
 
 
 1 
 
 w G O 
 
 Hi 
 
 III 
 
 Si|a s 
 
 Si "3 g 
 
 
 
 c 
 
 d 
 g 
 
 s 
 
 
 (U 
 
 T3 S 
 
 'r^ 'O ^ 
 
 OJ 
 
 a 
 
 M 
 
 H 
 
 .^0-5 
 
 ojj= ^fJ.c 5 
 
 
 S 
 
 
 
 rr " > 
 
 "^ t- rC 'M "^ 
 
 "o 
 
 _< 
 
 
 
 ^ ^2 
 
 ^C - 1 - 1 -^- o 
 
 <3 
 
 1 
 
 
 
 g^l 
 
 o c| ^ 
 
 CO 
 
 "3 
 
 
 
 *0 '.S 
 
 s^'i-5 
 
 
 j 
 
 
 c 
 
 - ||? 
 
 |l 
 
 
 Js 
 
 
 <L 
 
 |'||l 
 
 (U p 
 
 
 g 
 
 
 X 
 
 J " " -^ % ^" 
 
 
 '3 
 
 ptj 
 
 -k- 1 
 
 11 rC "^ ^O 
 
 >~< ^1" 
 
 s 
 
 Jo 
 o 
 
 o 
 
 Th 
 
 1 
 
 Sill 
 
 Igl 
 
 g 
 
 o 
 
 G 
 
 H 
 
 > 
 
 
 
 ^ x S ^ . 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 rr3 T3 O Cj 
 
 T3 3 O 
 
 o 
 
 
 < 
 
 $ 
 
 w JH J fe r | 
 
 .SJS 
 
 H 
 
 
 
 "o 
 
 *5 "" ^c u 
 
 CT 1 rt "5 
 
 
 
 
 2 
 
 s^- J"S 
 
 2 S^ 
 
 
 
 
 ^ 
 
 1 ill s 
 
 ^.c-3 
 
 
 
 
 C/D 
 
 
 r-H -^ 
 
 
 
 
 
 S -c ^ -5 
 
 O "CXi 
 
 
 
 o 
 
 1 
 
 (L> 
 
 o 
 
 "rt 
 
 i 
 
 PtJ 
 
 "S 
 
 N 
 
 1 
 
 v> 
 
 J 
 
 ri 
 
 1 
 
 g 
 
 "^ 
 
 *! 
 
 r|- 
 
 "^* 
 
 o 
 
 
 
 
 
 H 
 
 o 
 
 .^2 
 S 
 
 ^ 
 
 i^ 
 
 W3 
 
 c 
 .S 
 
 |ri 
 
 II 
 
 gl 
 
 Jri 
 
 
 
 .s 
 
 o ^ 
 
 
 'O 
 
 S tj- 
 
 S c 
 
 s g 
 
 S ^ 
 
 1 
 
 i 
 
 II 
 
 KH 1> 
 
 2 
 
 li 
 
METEOROLOGICAL INSTRUMENTS. I I I 
 
 This gives the average daily temperatures for the week in 
 each case, and must be multiplied by 7 to give the totals for 
 the week. 
 
 As an example let the mean of seven maxima for a week 
 in February be 47*3 and of the corresponding minima 
 34-6 : then the mean for the week is 4i'o. By third line in 
 the table, day-degrees above 42 F. 
 
 = (47-3 -42*0) x 0-4 = 2-1 average per day; 
 and below 42 F. 
 
 = (41 -0-34-6) -2- 1 =4-3 average per day. 
 
 Hence for the week we have 
 
 Day-degrees to the good = 2-1x7 = 1 4* 7 
 lost =4-3x7 =30-1. 
 
 Collecting these figures week by week, we get a kind of de- 
 bit and credit account as the season advances. 
 
 95. Temperature of Radiation. Radiation thermometers 
 are commonly employed to afford a measure of the intensity 
 of the heat radiations received from the sun or given off by 
 the surface of the earth. 
 
 The black bulb thermometer is an ordinary maximum 
 having the bulb covered with lamp black, a substance which 
 absorbs practically all heat rays falling upon it, and enclosed 
 in a glass jacket, which is exhausted of air and hermetically 
 sealed. When freely exposed to the action of the sun's rays 
 the thermometer rises until the heat received by the bulb in 
 unit time is just equal to that lost by radiation to the en- 
 closing jacket, which latter remains approximately at the 
 temperature of the air. The maximum recorded by the 
 black bulb minus the maximum temperature of the air is 
 taken as a measure of the maximum heating intensity of the 
 sun's rays for the day. 
 
 It cannot, however, be said that the indications of the 
 black bulb thermometer are of much value. In the first 
 place the sun's rays do not in general attain their greatest 
 power at the hour of maximum air temperature, but 
 
112 METEOROLOGY. 
 
 considerably earlier, an.1 we should therefore subtract from 
 the black bulb reading not the maximum, but the actual 
 air temperature at the time the black bulb reaches its 
 highest point. Again, it is almost impossible to con- 
 struct instruments exactly alike in the covering of the 
 bulb, the formation of the jacket, and the amount of the ex- 
 haustion. Finally, supposing the instrument perfectlyadapted 
 to its purpose, it is doubtful if the data afforded by it would 
 be of much value. Here again, what is really wanted is a 
 measure of total heat received, not of maximum intensity at any 
 instant. The actinometers of Pouillet, Crova, Langley, and 
 others, which to a certain extent give this, are, unfortunately, 
 not suited for general use ; but much may be learned from the 
 duration of direct solar radiation, even without attempting to 
 estimate its intensity. 
 
 96. Sunshine. This is measured by means of sun- 
 shine recorders, of which various forms have recently 
 been devised, that known as the Campbell-Stokes, in which 
 the sun's rays are concentrated by means of a glass ball and 
 made to leave a charred record on a strip of prepared paper, 
 being probably the most reliable. 
 
 97. Terrestrial Radiation. The terrestrial radiation 
 thermometer is an extremely sensitive Rutherford minimum. 
 The bulb is modified so as to present the greatest amount 
 of surface relatively to its contents, either by making it in 
 the form of a hollow cylinder, or by drawing it out and 
 bending it back upon itself. The instrument is exposed on 
 a surface of close short grass, being supported upon two 
 forked sticks so that the bulb just touches the tips of the 
 blades. 
 
 If we grant, what is extremely doubtful, that the reading 
 of this thermometer really represents the lowest temperature 
 reached by the surface of the grass during the night, it is 
 obvious that we have not gained much information, for that 
 temperature does not depend on radiation alone, but on the 
 
METEOROLOGICAL INSTRUMENTS. 113 
 
 temperature and moisture of the soil underneath, and on the 
 length and closeness of the grass ; and hence not only the 
 readings at different places, but even those at the same place 
 on different nights are not comparable. Still, the method 
 has a certain value, inasmuch as by subtracting the minimum 
 on grass from the minimum temperature of the air we get a 
 rough estimate of nocturnal radiation, \vhich may be useful 
 in agricultural matters. 
 
 98. The Barometer. Since the strictly local variations of 
 pressure in the atmosphere are, for the most part, extremely 
 small, the barometer in its application to meteorology may 
 be accurately called an instrument of precision. It consists 
 essentially of two parts the glass tube containing the 
 mercurial column and the scale for measuring the length 
 of that column. 
 
 The only corrections which must be applied to the height 
 of the mercurial column are due (i) to capillarity, and (2) 
 to change of density in the mercury. Since mercury does 
 not wet the interior of the tube, the capillary forces always 
 depress the column, making it shorter than it would other- 
 wise be by an amount which is greater the smaller the tube. 
 Hence also the mercury surface in the tube is convex up- 
 wards, and we measure the column from the central summit 
 not from the point of contact with the glass. Again, the 
 density of mercury depends upon its temperature, and the 
 less the density the longer will be the column necessary to 
 balance the pressure of the air. It is usual to accept the 
 density of pure mercury at 32 F. as a standard, and to re- 
 duce all observations of the barometer to that temperature. 
 Both of these corrections are in practice included with others 
 referring to the scale, so we need not discuss them further 
 at present. 
 
 The scale, which must be made of brass if the observa- 
 tions are to yield results of scientific value, is (except where 
 extreme accuracy is required) engraved on a tube which 
 
 H 
 
I 14 METEOROLOGY. 
 
 forms the case of the instrument. Any error in its construc- 
 tion must be ascertained by comparison with a standard. 
 Now, we must clearly understand what exactly is the quantity 
 we want to measure. Observe that the mercurial column 
 may be of any length ; we may fill the whole tube, however 
 long, by inclining it. The pressure is not measured by the 
 length, but by the vertical distance between the top of the 
 mercury in the tube and the surface of it in the cistern. 
 Hence 
 
 (1) The sca^e must be exactly vertical. This is easily 
 arranged by allowing the barometer to swing freely, either 
 by a hook or on gimbals. 
 
 (2) Since when the mercury column falls some mercury 
 flows out of the tube into the cistern, and when it rises out ot 
 the cistern into the tube, the level of the mercury surface in 
 the cistern varies with every change of pressure, and some 
 means must be adopted of adjusting either the zero end of the 
 scale to the mercury surface, or the mercury surface to the 
 zero end of the scale ; or else of making proper allowance 
 for the error introduced in the absence of such adjustment. 
 
 A modified application of the first method is the siphon 
 barometer much used on the Continent. The " cistern " 
 proper is done away with, and the lower end of the tube 
 bent through two right angles, so as to form a siphon. The 
 positions of the mercury in both legs of the siphon are then 
 measured, and the difference of level gives the height of the 
 barometer. 
 
 The second method, of adjusting the mercury surface to 
 the zero point of the scale, was first applied in practical 
 form by Fortin, and is now generally used in the better 
 class of instruments. The bottom of the cistern is made of 
 a flexible material such as leather, and is kept in position 
 by an external piston. The piston can be moved by means 
 of a screw, and the surface of the mercury in the cistern 
 raised and lowered at will. The zero point of the scale is 
 
METEOROLOGICAL INSTRUMENTS. 1 15 
 
 usually an ivory index to which the mercury is adjusted 
 either directly or by a float the stem of which bears a mark 
 corresponding to one on the index. 
 
 In the third class of instruments, where no special adjust- 
 ment is made, the zero of the scale is placed at a neutral 
 point, and a " capacity correction " is applied to all readings 
 above cr below that point. The amount of this correction 
 is sometimes marked on the instrument, but more usually 
 incorporated in the scale, which is then divided into inches 
 (or millimetres as the case may be) plus or minus the capa- 
 city correction. 
 
 (3) Since the length of the scale varies with temperature 
 all measurements made with it must be accompanied by a 
 statement of the temperature of the instrument at the time 
 of observation, so that they may be reduced to a standard 
 temperature at which the scale is correct. It is usual to 
 combine this correction with the temperature correction 
 required for the mercury column, and to express the joint 
 result in a table for reducing barometric observations to 
 32 F. Observe that except in special cases these tables 
 refer to barometers with brass scales. 
 
 The temperature "of the instrument is determined by 
 means of an " attached thermometer," the bulb of which is 
 placed inside the brass case. In making an observation 
 this should be read first, and the barometer should always 
 be hung where sudden great changes of temperature do not 
 occur so that both mercury column and brass scale shall 
 be at a fairly uniform temperature throughout. 
 
 At the upper end of the scale the brass tube forming the 
 case has two slits cut in opposite sides of it, exposing the 
 barometer tube inside and enabling us to observe the rise 
 and fall of the mercury. In the slits two small brass slides 
 are placed, with their lower edges-- in exactly the same hori- 
 zontal plane ; these slides can be moved up and down by a 
 rack and pinion arrangement, and the plane of their lower 
 
Il6 METEOROLOGY. 
 
 edges marks the level to be measured by the scale. Hence to 
 " set " the barometer, we bring the lower edges of the two 
 slides into an exact straight line with the highest point of 
 the mercury in the centre of the tube. Fig. 22 represents a 
 section through the tube, A A the two slides. In making 
 the adjustment it is necessary to have a strong light behind 
 
 the instrument, so as to be 
 able to see when the ray 
 C D is just cut off by 
 
 ^ e to p ^ e mercurv B 
 
 r* B an d the slides. 
 
 Having set the slides 
 in position we have next 
 to obtain the reading on 
 the scale, which we may 
 repeat is the distance be- 
 tween the ivory zero or 
 the " neutral point" as 
 the case may be, and the 
 lower edge of the slide. 
 
 99. The Vernier. 1\& degree of accuracy usually re- 
 quired is one five-hundredth (-002) of an inch, and in order 
 to distinguish so small divisions, we employ the device in- 
 vented by Peter Vernier in (fa) 1630, and known by his name. 
 
 Along the side of one slit is engraved the fixed scale of 
 inches divided into tenths (-100) and half-tenths (-050). 
 The vernier or movable scale is screwed to the slide which 
 moves in the same slit, and its zero coincides with the lower 
 edge, the point to be measured. The whole length of the 
 vernier corresponds to twenty-four divisions of the fixed 
 scale, i.e. twenty-four half-tenths (24 x -050= 1-200), and 
 that length is on it divided into a number of parts greater 
 by one twenty-five ; hence one division of the vernier is 
 equal to -fa x 1-200 = -048 or one five-hundredth of an inch 
 (002) less than one division of the fixed scale. 
 
METEOROLOGICAL INSTRUMENTS. Ii; 
 
 If, then, the zero line of the vernier is at the same level 
 as (say) the line of 29-000 inches on the fixed scale, so that 
 these two appear as one straight line, then the first division 
 of the vernier will be '002 inch below the first of the fixed 
 scale, the second -004 inch below the second, and so on ; 
 and therefore conversely if the first line of the vernier 
 corresponds with the first of the fixed scale, then the 
 zero of the vernier must be '002 above the line of 29-000 
 inches on the fixed scale. But the zero of the vernier is the 
 point we wish to measure, hence the reading is 29-000 
 inches + '002 inches = 29*002. Similarly, if the second line 
 of the vernier corresponds to the second of the fixed scale, 
 the zero of the vernier is '004 above the 29*000 inches line, 
 and the reading is 29-004 inches, and so on. Hence, gener- 
 ally, having set the slide : 
 
 (1) Note the division on the fixed scale which comes next 
 below the zero point of the vernier. 
 
 (2) Starting from the zero point of the vernier, count the 
 divisions upwards till that one is reached which is in corre- 
 spondence with any line on the fixed scale. Multiply the re- 
 sulting number by '002, and add the product to (i). The 
 sum is the reading required. 
 
 In practice, every fifth line on the vernier, which, of 
 course, marks one hundredth of an inch (5 x "002 -oio) 
 bears a figure, and the intermediate lines obviously mark 
 002, -004, "006, *oo8 respectively ; hence, after ascertaining 
 which line corresponds to one on the fixed scale, we may 
 read off directly, obtaining the hundredths from the first 
 figure below the line so found, and the five-hundredths by 
 allowing "002 for each intermediate line on the vernier. 
 This avoids the counting and multiplication of (2) 
 above. 
 
 100. Summary of Reductions in Reading Barometer. 
 We may summarise the steps in observing the pressure at 
 the cistern of the barometer as follows : 
 
1 1 8 METEOROLOGY. 
 
 (a) Read attached thermometer. 
 
 (b) Adjust mercury in cistern to zero point of scale, if the 
 instrument is constructed on Fortin's principle. 
 
 (c) Set and read vernier. 
 
 (d) Apply capacity correction where adjustment (b) cannot 
 be made. 
 
 (e) Apply Kew correction for scale error and capillarity. 
 (/) Reduce result of foregoing to standard temperature, 
 
 32 F. 
 
 These are necessary and sufficient for making all observa- 
 tions at exactly the same place comparable with each other, 
 and we may remark in passing that pressure observations 
 should be published in this form, without further reduction. 
 
 101. Reduction to Sea-Level. When we come to compare 
 the readings of the barometer at one station with simul- 
 taneous readings at another, a correction for difference of 
 level in the two instruments must be applied. The most 
 convenient method for general use is to reduce the ob- 
 servations at all stations to sea-level ; and it is therefore 
 most important that the height of the barometer cistern 
 above mean sea-level should be accurately known. It is 
 not too much to say that the value of barometer observa- 
 tions for scientific purposes largely depends on accuracy 
 in this matter. The height being known we reduce to sea- 
 level by applying the formula of 49. Observe that this 
 involves the assumption of certain stable conditions, and 
 also a knowledge of the vertical temperature gradient, and 
 that in consequence it is not always accurate. 
 
 102. Calculation of Mean Pressure. From what has been 
 said in 36 it will be understood that the reduction of 
 barometer observations taken at any hour, or set of hours, 
 to the true daily mean is by no means so simple a matter 
 as in the case of temperature observations. The great 
 variations in different seasons r.nd at different places both 
 
METEOROLOGICAL INSTRUMENTS. 119 
 
 in the form and amplitude of the phases of the curves make 
 it impossible to give a general rule for the correction for 
 daily range. For special inquiries in which the records of 
 a large number of stations in different parts of the globe have 
 to be compared, involving observations at different sets of 
 hours, corrections calculated from the hourly observations 
 of some standard station having a similar climate are ap- 
 plied to each individual case ; but great care must be exer- 
 cised in selecting the standard for comparison. We must 
 not, for example, employ the data of a station having an 
 insular climate to correct observations from one situated in 
 a continental region, for the daily curves of the two may be 
 almost exactly opposite in form. The following table, 
 calculated from data given by Buchan, 1 shows the average 
 corrections to be applied to the means of observations 
 at 9 a.m. and 9 p.m., 6 a.m., 2 p.m., and 10 p.m., and 
 9 a.m., 3 p.m., and 9 p.m. at a number of typical stations 
 in order to reduce them to the true mean of the day. They 
 refer only to the whole year, and in most cases vary con- 
 siderably from month to month : 
 
 Corrections to be applied to mean of observa- 
 tions at (thousandths of an inch). 
 Hours. Hours. Hours. 
 
 Atlantic 
 
 
 
 9:9 
 
 6: 2 : 10 
 
 9:3:9 
 
 Long. 2o-3< 
 
 *., 1 
 
 DPW., . r 
 
 -30 
 
 o 
 
 -9 
 
 Valentia, 
 
 . 
 
 - 5 
 
 o 
 
 -3 
 
 Kew, . 
 
 . 
 
 - 9 
 
 + i 
 
 2 
 
 Culloden, 
 
 . 
 
 - 5 
 
 
 
 -I 
 
 Ben Nevis Obsei-vatory, 
 
 - 3 
 
 o 
 
 -3 
 
 Gries, . 
 
 . 
 
 -16 
 
 - 1 
 
 + 4 
 
 Pola, . 
 
 . 
 
 - 6 
 
 + 1 
 
 -2 
 
 Moscow, 
 
 
 2 
 
 
 
 O 
 
 Batavia, 
 
 . 
 
 -36 
 
 -I 
 
 -4 
 
 Sitka, 
 
 . 
 
 
 
 o 
 
 
 
 We note that the average of observations at 6 a.m., 2 p.m., 
 1 Challenger Reports, "Atmospheric Circula'.ion," p. 13. 
 
I2O METEOROLOGY. 
 
 and 10 p.m. gives a close approximation to the true mean in 
 all the widely different cases included in the table ; hence this 
 makes an excellent combination of hours for barometric 
 observations. The correction to be applied to the average 
 of 9 a.m. and 9 p.m. varies between the limits 'ooo at Sitka, 
 and - '036 at Batavia, being greatest for what we may call 
 oceanic climates, and least in insular climates or places 
 situated on the western confines of a continental area, 
 where the " morning maximum " is delayed. In the 
 British Isles the correction for these hours varies between 
 '005 at Valentia, and '009 at Oxford and Greenwich, 
 averaging about *oo6 over all. 
 
 103. Non-Mercurial Barometers. It is unnecessary to 
 give here any description of the various delicate pressure 
 gauges used as barometers. The aneroid, metallic baro- 
 meter, etc., while undoubtedly of great value for many 
 purposes, cannot replace the mercurial barometer for regular 
 observations, not because their errors are large, but because 
 they are variable. Observations with the best of aneroids 
 must be constantly checked by comparison with those of a 
 mercurial barometer. 
 
 Special mention must, however, be made of self-recording 
 barometers and thermometers. The most accurate instru- 
 ments, which continuously record their readings by means 
 of photography, are only suitable for specially equipped 
 observatories, but in recent years cheap barographs and 
 thermographs have been introduced, notably by Richard 
 Freres of Paris. These instruments, particularly the 
 barograph, give tracings which are extremely valuable when 
 combined with the ordinary eye-observations. Using two 
 daily readings of the mercurial barometer to determine the 
 error of the barograph, the latter can be made to yield data 
 sufficiently accurate for the determination of the diurnal 
 barometric curve, besides affording a great deal of informa- 
 tion about changes of pressure generally, and on the other 
 
METEOROLOGICAL INSTRUMENTS. 121 
 
 hand the automatic records afford a ready means of detect- 
 ing errors in the eye-observations. 
 
 104. Wind Observations : Direction. In determining the 
 direction of the wind, it is to be observed that the true, not 
 the magnetic, direction is to be recorded,, and care must be 
 taken in beginning observations that the cardinal points are 
 accurately determined once for all. When the observations 
 are made from a vane, this should be perfectly clear of trees, 
 buildings, or anything likely to deflect the course of the 
 wind, and the vane must have constant attention in case of 
 working stiffly or " sticking," especially during frost. When 
 possible it is better on the whole to make the observations 
 directly from a flat roof or similar exposed situation. Draw 
 a "compass-card/' twelve or fourteen feet in diameter, on 
 the ground or roof, and standing in the centre, turn so as to 
 face the wind ; then read off the direction from the 
 " compass-card." 
 
 Anyone possessing a little mechanical skill can easily 
 construct a self-recording wind-vane for himself. A small 
 vane is attached to a suitable spindle an old " jenny " 
 spindle will answer the purpose which revolves in bushes 
 fixed in the ends of a long narrow box, the end carrying the 
 vane protruding. Inside the box the spindle carries a 
 cylinder about four inches in diameter, and extending 
 through nearly its whole length. An upright fixed beside 
 the cylinder is grooved so as to allow a small pencil or 
 siphon pen to work up and down, moved by a simple 
 adaptation of the " works" of an American clock. A piece 
 of smooth-surfaced paper is attached to the cylinder after 
 the manner of a newspaper envelope, and revolves with 
 changes of wind-direction, the spindle being moved by the 
 vane. In contact with the paper the pen is drawn upwards 
 by the clock at a given rate, and a record of every change is 
 traced. The paper can be removed by slitting it longitud- 
 inally. The record will, in most cases, be found to be a band 
 
122 METEOROLOGY. 
 
 of varying width, according to the unsteadiness of the wind ; 
 the average direction may be obtained by drawing a line 
 down the centre of the band. 
 
 In making observations of wind direction it is well to 
 record to the nearest point : and then in calculating the 
 mean direction to halve all the " by " directions (N. by E., 
 N.E. by N., etc.) between the two points on each side 
 counting half the N. by E.'s to N. and half to N.N.E., half 
 the N.E. by N. to N.N.E. and half to N.E., and so on. 
 This reduces to sixteen points, and the process may be 
 repeated so as to reduce to eight points N., N.E., E., etc. 
 
 105. Wind Pressure and Velocity: Anemometers. Many 
 and various appliances have been devised for measuring the 
 pressure and velocity of the wind, but none have as yet 
 proved altogether satisfactory, either on account of defects in 
 the instruments themselves, or of the difficulty of interpreting 
 the records. Of velocity anemometers, that of Robinson is 
 on the whole the most practically useful, being strong and 
 durable in construction, and also independent of the direc- 
 tion of the wind. It consists of a vertical spindle, arranged so 
 as to rotate with the least possible friction, which carries at 
 its upper end four radial arms at right angles to each other, 
 and at the extremity of each arm a hemispherical cup. The 
 cups are arranged with their convex sides pointing in the 
 same direction, and hence, whatever the direction of the 
 wind, two opposite cups present opposite faces to it. But 
 the pressure against the concave side of one cup is greater 
 than against the convex side of the one opposite to it, hence 
 the whole system rotates, and always in the same direction, 
 the convex side of the cups advancing. A counter records 
 the number of turns made by the spindle. Apart from the 
 fact that this instrument does not work satisfactorily in light 
 winds, and that the frictional resistance offered by the 
 spindle varies considerably with the temperature and the 
 condition of the weather, a good deal of doubt still exists 
 
METEOROLOGICAL INSTRUMENTS. 123 
 
 about the reduction of the data given by it. The ratio of 
 the velocity of the wind to that of the cups varies with the 
 size of the instrument, and it is therefore necessary to adopt 
 certain standard dimensions ; those officially recognised in 
 this country are Radius of cups, 4-5 inches ; length of 
 arms (centre of cup to centre of spindle), 24 inches. It was 
 long supposed that the numerical value of the ratio for an 
 anemometer of this size, and indeed for all sizes, amounted 
 to 3-0, but recent investigations 1 have shown pretty conclu- 
 sively that this value is considerably too great, 2*1 being 
 much nearer the truth. It may be pointed out that the 
 wind velocities published in most meteorological records are 
 affected by the erroneous value of this factor, and must 
 therefore be handled with caution in any inquiry involving 
 their use. 
 
 In the course of the researches just referred to, Dines 
 describes what promises to be the first really satisfactory 
 pressure anemometer. The fact that wind blowing against 
 the mouth of an open tube causes an increase of pressure 
 within it was many years ago applied by Lind to the 
 measurement of wind pressure. In Lind's anemometer, 
 which has undergone a number of modifications, the mouth 
 of a tube is kept facing the wind by a vane, and the increase 
 of pressure in it measured by a column of water in a siphon. 
 But the indications of such an instrument are unsatisfactory 
 on account of the changes of pressure at the other end of 
 the tube. The quantity to be measured is the difference of 
 pressure between the end exposed to the wind and the 
 other, which cannot be maintained at an exactly constant 
 amount. Dines solves the difficulty by exposing the other 
 end also to the wind, but in such a manner that the pressure 
 is reduced instead of increased thereby not only rendering 
 the instrument independent of accidental cnanges of 
 pressure not due to the wind, but increasing the differences 
 1 See Dines: Quart. Jour. Met. Soc., vol. xviii., p. 182. 
 
124 METEOROLOGY. 
 
 caused by it, and making them easier to measure accur- 
 ately. 1 
 
 From experiments with the tube and other anemometers, 
 Dines gives the following equation connecting P, the wind 
 pressure in Ibs. per square foot, with V, its velocity in 
 miles per hour. 
 
 P = -oo3 V 2 . 
 
 1 06. Estimation of Wind Force. From what has been 
 said it will be apparent that exact instrumental measure- 
 ments of air movements are not easily obtained, and in 
 practice they are usually dispensed with, and a system of 
 estimation based on the effects produced by winds of 
 different strengths is substituted. It would appear at first 
 sight that this method is extremely crude, and likely to give 
 widely discordant results with different observers ; but ex- 
 perience has shown that a much better agreement is obtained 
 than we might expect, closer indeed than is shown by most 
 instrumental methods. 
 
 The scale most commonly employed is that devised by 
 Sir F. Beaufort in 1806, usually called the " Beaufort 
 scale." It ranges from o to 12, the former figure represent- 
 ing a calm, and the latter a hurricane of such violence as to 
 reduce a ship to scudding under bare poles. The inter- 
 mediate numbers were originally determined by the amount 
 of sail a full-rigged ship would carry in a wind of the force 
 to which they correspond, but it is scarcely necessary to 
 detail them in that fashion here. A little practice will 
 enable the observer to estimate the wind force simply by the 
 "feel " of the wind as he faces it. The following hints may 
 be useful : 
 
 Force o Calm. 
 
 1 Light air. 
 
 2 Very light breeze. 
 
 3 Light breeze. 
 
 1 For details see Q. J. Met. Soc., xvi-, p. 2c8, and xviii., p. 168. 
 
METEOROLOGICAL INSTRUMENTS. 125 
 
 4 Moderate breeze. 
 
 5 Fresh breeze. 
 
 6 "Half a gale." 
 
 7 Moderate gale. 
 
 8 Fresh gale. 
 
 9 Strong gale. 
 
 10 "Whole gale." 
 
 1 1 Violent storm. 
 
 12 Hurricane. 
 
 Various other scales, differing in the range of numbers 
 employed, are in use, such as the English land scale 0-6, 
 the American international scale o-io, and others. 
 
 107. Diurnal Variation of Wind. If we bear in mind 
 the relations of the upper to the lower currents explained in 
 a former chapter we shall have no difficulty in understand- 
 ing the diurnal variations in the direction and force of the 
 wind, first fully accounted for by Sprung. 1 The upper 
 current moves faster than the lower, and in the northern 
 hemisphere follows a direction more to the right in the 
 southern more to the left. Now during the day the heating 
 of the lower strata causes them to ascend, thereby increas- 
 ing the friction between upper and lower currents, and 
 reducing the difference of velocity. The lower current, 
 under the influence of the upper, is deflected to the right 
 in the northern hemisphere, to the left in the southern, and 
 has its speed increased in both cases, while the upper, 
 under the influence of the lower, is deflected in the opposite 
 direction, and has its speed diminished ; and this effect is 
 the more marked the greater the diurnal heating of the 
 lower strata relatively to the upper, and the greater the 
 normal angle between the two currents caused by friction 
 of the lower on the earth's surface. Hence 
 
 (i). Near the equator, and over the open sea, there is little 
 diurnal variation either of direction or speed. 
 
 1 Met. Zeit., 1884, p. 16. Lehrbuch der Meteoroloyie, p. 342. , 
 
126 METEOROLOGY. 
 
 (2). On plains and similar land surfaces, even at great 
 elevations, the wind shifts with the hands of a watch and 
 attains its maximum strength during the afternoon, backing 
 and diminishing again at night in the northern hemisphere. 
 The changes of direction are reversed in the southern 
 hemisphere. 
 
 (3). On mountain peaks the wind shifts against the hands 
 of a watch, and diminishes in strength during the afternoon, 
 and veers and increases again at night. The changes of 
 direction are reversed in the southern hemisphere. 
 
 1 08. Hygrometry. We have already hinted at two 
 methods of determining the amount of aqueous vapour in 
 the atmosphere ; one by actually abstracting the moisture 
 in the form of water and weighing it, and another by cooling 
 a bright surface till the temperature of the dew-point is 
 indicated by the deposition of moisture. The first of these 
 may be dismissed as useless for regular observations, and 
 the second as requiring more time and manipulative skill 
 than most observers have at their disposal. It accordingly 
 becomes necessary to fall back on indirect methods, and of 
 these we need only describe the one in almost universal 
 use. 
 
 The psychrometer, the invention of which is variously 
 ascribed to Hutton, Leslie, Gay-Lussac, and August, enables 
 us to calculate the dew-point from the temperature of the 
 air and the "temperature of evaporation." Two thermo- 
 meters of exactly similar construction are placed side by 
 side in the Stevenson screen. One of these is the common 
 thermometer already described, and the bulb of the other is 
 enveloped in a covering of thin muslin (perfectly clean and 
 free from starch) kept wetted by a piece of cotton wick at- 
 tached to it, which dips into a vessel of water. 
 
 As long as the air is dry, water evaporates from the wet 
 muslin covering, and the heat required for the evaporation 
 is drawn from the bulb of the thermometer, which accord 
 
METEOROLOGICAL INSTRUMENTS. I2/ 
 
 ingly has its temperature lowered to what is called the 
 " temperature of evaporation." A steady condition is 
 reached, and from the difference between the readings of the 
 dry and wet bulbs we find the dew-point. 
 
 The problem of calculating the dew-point from the 
 temperatures of the air and of evaporation has been at- 
 tacked by Ivory, August, Belli, Apjohn, Regnault, Kamtz, 
 Clerk-Maxwell, Sworykin, Pernter, Ferrel, and Everett, from 
 theoretical considerations, but none of the solutions hitherto 
 offered are completely satisfactory. There seems to be 
 little doubt that the equation required is of the form 
 
 f f" = (t t')K E 
 where f ' = pressure of vapour at temperature of wet-bulb. 
 
 i" = . dew-point. 
 
 t = temperature of dry bulb. 
 
 t' = wet bulb. 
 
 p = barometric pressure at time and place of observa- 
 tion. 
 
 P = a standard barometric pressure. 
 
 But K involves serious difficulties. For values of (t t') such 
 as are usually found in insular climates, it is best to use the 
 empirical tables of Glaisher, 1 provided |- is nearly equal to 
 unity. These tables cannot, however, be trusted for very 
 low humidities, and are not applicable to observations made 
 at high level stations. 
 
 109. Absolute Humidity. The pressure of vapour actually 
 existing in the atmosphere at any time, is, of course, the 
 saturation pressure corresponding to the temperature of the 
 dew-point, and is commonly known as the absolute humidity. 
 The pressure indicated by the barometer is that due to the 
 air and vapour together, but it must be observed that it is 
 not the simple sum of the two pressures unless we assume 
 that the air and vapour are completely mixed throughout, a 
 
 1 6lh edition, 1876. 
 
128 METEOROLOGY. 
 
 condition seldom if ever realised. 1 Absolute humidity 
 shows two distinct types of diurnal variation. Over the 
 ocean the curve has a single maximum and single minimum, 
 closely coinciding with those of temperature, and this simple 
 form occurs with little modification in damp climates or 
 during rainy periods generally. In continental climates a 
 double maximum is found, one in the forenoon and another 
 in the late afternoon. The minimum between is probably 
 caused by the supply of fresh vapour from the ground being 
 insufficient to replace that removed by the strong ascending 
 currents set up by the diurnal rise of temperature. 
 
 no. Relative Humidity. Of greater practical importance 
 is the relative humidity, or ratio of the absolute humidity 
 to saturation. It is usual to express this numerically as a*per- 
 centage, saturation being represented by 100, i.e. by dividing 
 the vapour pressure corresponding to the dew-point by that 
 corresponding to the temperature of the air, and multiplying 
 by 100. Since the diurnal variation of absolute humidity is 
 small in amount, the curve of relative humidity is practically 
 the inverse of that of temperature, and accordingly retains 
 almost the same form at all times and in all climates. 
 
 in. Clouds: Classification of Different Kinds. The 
 condensation of vapour into visible form takes place in such 
 a manner as to give rise to an almost infinite variety of 
 cloud-shapes, and the classification of the different kinds of 
 clouds is a matter of great difficulty. We may take it that 
 there are two distinct ways in which condensation most 
 frequently occurs either a layer of the atmosphere is cooled 
 bodily to near its dew-point, or a body of moist air is in- 
 truded into a mass which is cold and dry. The first may 
 happen through cooling by radiation, as in the case of air 
 lying in a valley during the night, or at an interface between 
 upper and under currents; and the result is a stratified 
 
 iSee Strachey, Proc. 7?. S., vol. xi., p. 182. 
 
METEOROLOGICAL INSTRUMENTS. 1 29 
 
 sheet of greater or less extent, forming a cloud which re- 
 ceives the generic title of stratus. The second, again, is 
 usually the result of an ascending current caused by heating 
 of the lower strata, in which the temperature falls till the 
 dew-point is reached. As the ascending motion becomes 
 weaker and weaker the air spreads out on all sides like the 
 head of a sheaf, and the cloud takes a heaped or cumulus 
 form. 
 
 Endless combinations of both stratus and cumulus forms 
 are found at all levels, but beyond a certain elevation, pro- 
 bably that where the cloud-particles are frozen, special 
 characteristics are observed, the structure of the cloud- 
 masses taking a finer texture usually compared to hair or 
 wool. All forms of this, type are designated cirrus. 
 
 It is obvious that we may subdivide each of these great 
 classes to any extent, so as to obtain any required degree of 
 minuteness in description ; but however useful an elaborate 
 classification may be to specialists, it is necessary for 
 ordinary observing purposes to limit the number of terms 
 employed. 
 
 Howard, in 1803, proposed the following cloud-nomencla- 
 ture, which is still in general use 1 : 
 
 1. CIRRUS. Parallel, flexuous, or diverging fibres, extensible 
 
 by increase in any or all directions (" mares' tails "). 
 
 2. e CUMULUS. Convex or conical heaps, increasing upward 
 
 from a horizontal base. 
 
 3. STRATUS. A widely extended, continuous, horizontal 
 
 sheet, increasing from below upward. 
 
 4. CIRRO-CUMULUS. Small well-defined roundish masses, 
 
 in close horizontal arrangement or contact (the 
 familiar " mackerel sky "). 
 
 5. CIRRO-STRATUS. Horizontal or slightly inclined masses 
 
 attenuated towards a part or the whole of their cir- 
 
 1 E c say on the Modifications of Clouds. Phil. Mag., vol. vii. 
 
 I 
 
1 30 METEOROLOGY. 
 
 cumference, bent downward, or undulated ; separate, 
 or in groups consisting of small clouds having these 
 characters. 
 
 6. CUMULO-STRATUS. The cirro-stratus blended with the 
 
 cumulus, and either appearing intermixed with the 
 heaps of the latter, or super adding a widespread 
 structure to its base. 
 
 7. CUMULO-CIRRO-STRATUS OR NIMBUS. The rain cloud. 
 
 A cloud, or system of clouds, from which rain is 
 falling. 
 
 This classification is undoubtedly meagre, and although 
 none of the efforts to improve upon it have proved entirely 
 satisfactory, the following, proposed by Abercromby and 
 Hildebrandsson I in 1887, was recommended for general use 
 by the International Meteorological Conference at Munich in 
 1892. 
 
 High Clouds. 
 
 1. CIRRUS. Curl cloud. Average height 9,000 metres. 
 
 2. CIRRO-STRATUS. Average height 9,000 metres. 
 
 3. CIRRO-CUMULUS. Average height 6,500 metres. 
 
 Middle Clouds. 
 
 4. STRATO-CIRRUS. Average height ? metres. 
 
 5. CUMULO-CIRRUS. Average height 4,000 metres. 
 
 Low Clouds. 
 
 6. STRATO-CUMULUS. Average height 2,000 metres. 
 
 7. CUMULUS. Average top 2,000 ; base 1,500 metres. 
 
 8. CUMULO-NIMBUS OR CUMULO-STRATUS. Average top 
 
 3,000; base 1,500 metres. 
 
 9. NIMBUS. Average height 1,500 metres. 
 
 10. STRATUS. Average height 600 metres. 
 
 The numbers give the approximate average height at 
 which each form of cloud is observed, based on the measure- 
 1$. /. Met. Soc., vol. xiii., No. 62, 1887. 
 
METEOROLOGICAL INSTRUMENTS. 131 
 
 ments of Ekholm and Hagstrom made at Upsala, and in 
 the mountains of Jemtland. The observations of Clayton 
 and Ferguson at Blue Hill Observatory, Massachusetts, 
 give on the whole substantially the same figures for eastern 
 America. The classification is similar to that of Howard, 
 extended so as to distinguish an intermediate layer of clouds, 
 Nos. 4 and 5. It would serve no good purpose to describe 
 each in detail; the student is referred to Abercromby's 
 photographs in the paper already quoted. 1 
 
 A general classification of clouds becomes possible only 
 in view of the fact established by Abercromby, that cloud- 
 forms are identical in all parts of the world, which simply 
 means that the processes of condensation take place every- 
 where in the same way. This, however, does not mean that 
 the same cloud-forms are everywhere associated with the 
 same kind of weather. 
 
 112. Amount oj Cloud. The amount of cloud is usually 
 registered by considering that part of the sky more than 20 
 above the horizon, and estimating the fraction obscured to 
 the nearest tenth ; thus overcast sky is recorded as ten, half 
 covered as five, and so on. Recent investigations at 
 Pavlovsk 2 have shown that the error of this method may 
 amount to 10 per cent., even with careful observation, and 
 it seems desirable that greater accuracy should be obtained. 
 The proportion of sky covered does not vary greatly at dif- 
 ferent hours of the day, but there is, in general, a tendency 
 towards a maximum of stratiform clouds near the hour of 
 minimum temperature, and of cumulus clouds towards the 
 afternoon ; the actual resulting variation depending, of 
 course, on the local conditions favouring the formation of 
 each kind. The observations of H.M.S. Challenger 5 show 
 
 1 See also Nature, vol. 37, p. 129. Cloud Atlas, by Neumayer, 
 Koppen, and Hildebrandsson, Hamburg, 1890 ; also K. Singer, Cloud- 
 forms, Ackermann, Munich, 1892. 
 
 2 Repertorium filr Meteorologie, Bd. xiii. 
 
 3 Challenger Reports, " Atmospheric Circulation," p. 28. 
 
132 METEOROLOGY. 
 
 a double maximum at, or shortly after, sunrise, and early 
 in the afternoon ; and Liznar 1 gives the following four types 
 of curve at different places : 
 
 Madrid. One maximum at noon, minimum in evening. 
 Los Angeles. One morning maximum, minimum at noon. 
 Vienna. Two maxima and minima ; principal maximum 
 
 morning, principal minimum evening. 
 Tiflis. Two maxima and minima j principal maximum 
 
 noon, principal minimum evening. 
 
 113. Height of Clouds. According to Ekholm and Hiig- 
 strom, the height at which all forms of clouds occur increases 
 from morning to night, except the summits of cumulus clouds, 
 which are highest about 1-30 p.m. The observations of 
 Clayton, at Blue Hill, indicate that the base of cumulus is also 
 highest in the early afternoon, although both observers agree 
 that the vertical thickness is greatest at that time. The 
 daily change of level is less the greater the elevation. 
 
 114. Precipitation. Under the term precipitation is in- 
 cluded the total aqueous deposit on the earth's surface from 
 the atmosphere, whether as rain or in the various wholly or 
 partially frozen forms of snow, hail, or sleet. The physical 
 processes involved in the formation of the different kinds of 
 deposition are by no means well understood, and it is 
 usual in making measurements to class all together as 
 " rainfall " ; snow, hail, etc., being melted into water and 
 measured as such. 
 
 The amount of rain falling on the earth's surface at any 
 place is measured in terms of the depth to which that surface 
 would be covered supposing none of the rain to be absorbed, 
 and the object of the rain gauge is to expose a certain 
 limited area so that it shall receive the same amount of rain 
 as the surface of the ground in its immediate neighbourhood, 
 no more and no less, and retain all the water so received 
 for measurement. The depth of water collected over that 
 Met. Zeit., 1885, p. 342. 
 
METEOROLOGICAL INSTRUMENTS. 
 
 133 
 
 area represents the rainfall, and accuracy of measurement is 
 attained by transferring the water to a narrow vessel. If, for 
 example, the area of the rain gauge is twenty square inches, 
 and the rainfall "05 of an inch, we avoid the necessity of 
 measuring this inconveniently small quantity by pouring the 
 water into a measuring glass two square inches in section, 
 which will be filled to a depth of half an inch, a length 
 easily measured to the required degree of accuracy by means 
 of a scale engraved on the side of the vessel. 
 
 The most convenient rain gauge for practical pur- 
 poses consists of a copper or japanned tin cylinder, at the 
 upper end of which is fixed a turned brass ring with a sharp 
 edge whose diameter is accurately known. Some six inches 
 below the ring the cylinder narrows to a funnel, and a tube 
 leads to a collecting vessel (Fig. 23). The ratio of the 
 area of the brass ring to the sectional 
 atea of the measuring glass must, of 
 course, be accurately known. Wild 
 has shown that the size of the gauge does 
 not appreciably affect the results, at 
 least within the limits of four and 
 twenty-four inches diameter. The sizes 
 usually employed in this country are 
 five and eight inches. 
 
 The rain gauge should be exposed on 
 a level piece of ground, clear of all 
 objects whose height is greater than 
 their distance from the gauge. The 
 rim of the gauge, which must be per- 
 fectly level, is usually placed at a height 
 of one foot above the ground, an ad- 
 justment of great importance, as the 
 rainfall diminishes rapidly as the Fig. 23. 
 
 height above ground increases, at least until sixty feet is 
 reached. 
 
1 34 METEOROLOGY. 
 
 In describing the climate of any locality, it is important 
 to know not only the total amount of rain, but the rate at 
 which it fell. Two stations may have almost the same 
 annual rainfall, but while one receives the bulk of it in a few 
 weeks of torrential rains, the other has it spread over months 
 of misty drizzle. For the accurate determination of the 
 rate of fall it would be necessary to employ a continuously 
 self-recording gauge, but for most purposes it is sufficient 
 merely to record the number of " rainy days." Meteorolo- 
 gists are unfortunately not unanimous in defining a rainy 
 day. Symons, to whom the rainfall organisation in Britain 
 in great part owes its existence, adopts a minimum record of 
 o'oi inch as characteristic of a rainy day; Hann 1 suggests 
 i mm. (-04 in.). The matter is simply a question of how 
 much is to be allowed for accidental deposits, as dew or 
 hoar-frost, and for errors of observation. 
 
 Of all meteorological phenomena, rainfall is the most 
 variable and uncertain. In determining the average fall at 
 any place, it is necessary to deal with observations extend- 
 ing over a long period, and hence in any general discussion 
 great caution must be exercised. Until within the last few 
 years, the number of stations furnishing rainfall observations 
 oftener than once a day was extremely small, and it is im- 
 possible to speak in general terms as to the hourly variation. 
 According to Hellman, 2 the diurnal variation of rainfall, 
 like that of cloudiness, can be classified according to a 
 number of typical curves. An afternoon maximum occurs 
 in many places, especially in summer, corresponding to the 
 hour of maximum frequency of thunderstorms ; and another 
 maximum late at night or in the very early hours of the 
 morning is connected with peculiarities in the diurnal march 
 of pressure, temperature, and wind-velocity. The latter 
 phase is characteristic of Western Europe at all seasons, and 
 
 1 Met. Zeit., 1888, p. 39. 
 - Mtt. Zeit., 1889, p. 271. 
 
METEOROLOGICAL INSTRUMENTS. 135 
 
 is always specially strong in winter. Symons 1 gives as ten- 
 tative results from twenty years' observations in London 
 (i) In winter the nights are wetter than the days. (2) In 
 spring and autumn there is not much difference. (3) In 
 summer nearly half as much again falls by day as by night. 
 
 115. Evaporation. The amount of water returned to the 
 atmosphere by evaporation from the earth's surface is an 
 element of the greatest importance in meteorological investi- 
 gations, especially in its bearings on agricultural problems. 
 Its measurement is, however, a matter of very great difficulty, 
 and no simple instrument has as yet given sufficiently good 
 results to justify its coming into general use. 
 
 1 British Rainfall, 1885, p. 25. 
 
CHAPTER VII. 
 
 THE ELEMENTS OF CLIMATE. 
 
 116. Method of Inquiry. In order to gain an adequate con- 
 ception of the climate of any part of the earth's surface, it is 
 necessary in the first place to know what types of weather 
 most frequently occur at different seasons, their average in- 
 tensity and persistency, and the probable deviations at any 
 given time from the usual state of things. We have already 
 learned something of the conditions, so far as they are 
 known, which favour the formation of one or other of these 
 types, and it remains now to consider more particularly 
 what are the general effects produced. Given the average 
 type of weather, what are the principal modifications to 
 which it is subject ? 
 
 Now, in answering this question, the method to be 
 followed is still the same. In dealing with weather, we 
 treated it as the variable quantity, regarding everything else 
 as constant. We now propose to regard the weather as a 
 constant or a known variable, and to inquire what other 
 factors go to make up climate, whether they are variable or 
 not, and within what limits. But since in most cases the 
 vicissitudes and variations of weather extend through a wide 
 range, we shall not be justified in regarding any of its ele- 
 ments as constant except by employing carefully determined 
 averages covering a very large number of cases. So great 
 are these variations that we were able to describe, and in 
 part to account for many of the phenomena observed with- 
 out the presence of other variables (tacitly assumed, for the 
 
 136 
 
THE ELEMENTS OF CLIMATE. 137 
 
 time, to be constant), being even suggested ; and as soon 
 as other variables were looked for, it became necessary to 
 employ more refined methods of observation in order to de- 
 tect them. 
 
 This much was gathered originally from one exception, 
 which, as it were, proved the rule. In lower latitudes, 
 where the sun's heat is of great intensity, the phenomena of 
 diurnal variation attain so full a development that instead of 
 being completely disguised by the weather disturbances, as 
 in higher latitudes, the case is frequently reversed. The 
 diurnal variations in temperate zones, although not in general 
 so well-marked either absolutely or in relation to the weather 
 disturbances, are nevertheless important elements of climate. 
 We have already had occasion to discuss them as fully as 
 limits of space permit while describing the methods and 
 instruments employed in their measurement. In still higher 
 latitudes, as the sun's power diminishes, the diurnal march 
 of phenomena becomes smaller and smaller in extent, until 
 a point is reached beyond which at certain seasons day and 
 night do not occur, and any small horary variation, if ob- 
 served, must be due to the indirect influence of other parts 
 of the atmosphere. 
 
 Hence the inquiry is still further reduced to the question 
 If over a certain area the average weather and the average 
 daily variations are the same, is the climate uniformly the 
 same at all points ? 
 
 117. Land and Sea. We have hitherto divided the sur- 
 face of the earth into land and water simply with reference 
 to the broad distinctions between the two in respect of 
 thermal properties ; we have regarded the sea as a fixed 
 surface having certain peculiarities of behaviour as to 
 radiation, absorption, reflection, etc., and the land as a 
 similar and similarly-situated surface having certain others. 
 This division may suffice as long as we are dealing with 
 general weather systems affecting large areas, but it ob- 
 viously leaves many considerations out of account which, 
 
IjS METEOROLOGY. 
 
 reacting on these systems, must seriously modify the effects 
 produced by them in different places. 
 
 Let us examine, then, the effects produced by winds 
 blowing across different kinds of surfaces, and how these 
 effects react upon the atmosphere. In discussing the in- 
 curving of wind in cyclones we compared the friction of air 
 moving over the sea and of that moving over land, and we saw 
 how the effects of such friction became less marked at small 
 elevations, disappearing gradually, and showing quite clearly 
 the friction between one layer of air and those immediately 
 above and below it ; or in other words, the viscosity of the air. 
 All the friction effects observed in the atmosphere are pro- 
 bably due to viscosity. There is always a very thin layer 
 of air touching the earth's surface over which the others 
 slide, and this contact layer is practically fixed to the surface, 
 so that there is no slipping friction properly so called. The 
 next layer drags over the fixed one, and again the next over 
 that, and so on. 
 
 The effects of this process on a land surface are obvious 
 enough. Over a plain country there will be a uniform 
 dragging, which, over a broken or mountainous surface, will 
 show an irregular increase, and extend on the whole to con- 
 siderably higher levels. 
 
 1 1 8. Oceanic Climates. But over the sea the case is 
 markedly different. The dragging force on the surface pro- 
 duces a movement of the water corresponding to its amount 
 and persistency, and ultimately gives rise to a surface 
 current ; and since the water has more inertia and greater 
 specific heat than the air, the current penetrates further 
 than the wind which causes it, and acts as a kind of brake 
 on the fluctuations of temperature of the winds passing over 
 it. Hence where the prevailing winds blow landwards from 
 the sea they bring with them the climate of the region 
 whence they came, retaining the oceanic temperature 
 characteristics of small range and deferred maxima and 
 minima. Again, when the prevailing winds are seawards 
 
THE ELEMENTS OF CLIMATE. 139 
 
 from the land, the surface water is driven away from the 
 coast, and bottom water, varying little in temperature, comes 
 welling up from below. 1 
 
 It is matter of observation that these two causes acting 
 together supplement the transference of heat brought about 
 by the winds. Warm water is carried from lower latitudes 
 to higher, and cold from higher to lower, with the result 
 that on the whole in the tropical zones the continents are 
 warmer than the oceans, while the reverse is the case in 
 higher latitudes. 
 
 Since the annual range of air temperature over the sea 
 is much less than over the land, it follows that, if the mean 
 of the former for the whole year is the higher, the greatest 
 differences must occur in winter. In the temperate zones 
 the excess of cold over continental, compared with oceanic 
 areas, in winter, is greater than the excess of heat in sum- 
 mer. The following average temperature for each month 
 at Valentia in the extreme west of Ireland, and at Voronezh 
 in Central Russia, in almost the same latitude, may be taken 
 as an example : 
 
 Valentia. Voronezh. Diff. 
 
 Lat. 5i'5s' N. Lat. 51 44' N. 
 
 January, . . 45'3 F. i6'4F. +28'9 
 
 February, . . 45-5 167 -t-28'8 
 
 March, . 46-3 36-9 +19-4 
 
 April, . . . 49'o 44*4 + 4'6 
 
 May, . . . 52*8 60*4 7'6 
 
 June, . . . 56-3 68-3 -12-0 
 
 July, . . . 587 70-6 -12-2 
 
 August, . . . 59-5 67-9 8*4 
 
 September, . . 56*6 57*1 0*5 
 
 October, . . 52*0 44^3 +77 
 
 November, . . 47*6 34'! +i3'5 
 
 December, . . 45-4 22*5 +22*9 
 
 Year, . 51-3 44-2 + 7-1 
 
 1 Scottish Geographical Magazine, 1 888, p. 345. 
 
140 
 
 METEOROLOGY. 
 
 119. The Atlantic Ocean. The best illustration of these 
 facts is afforded by the ocean most familiar to us the 
 North Atlantic. The prevailing winds are there determined, 
 as already stated, by the "Atlantic anti -cyclone," and by an 
 area of low pressure whose centre remains stationary in the 
 neighbourhood of Iceland : and of these the former attains 
 its fullest development in summer, and the latter in winter. 
 
 Fig. 24. February. Fig. 25. August. 
 
 NORTH ATLANTIC OCEAN. 
 
 Sea o to 5 F. 
 warmer than Air. 
 
 \\\ 
 
 5 to 
 10 F. 
 
 \\\ 
 
 Air o to 5 F. 
 wanner than Sea. 
 
 Hence the North Atlantic forms an immense anti-cyclonic 
 vortex, weakened and deformed in winter by the depression 
 to the north, which receives cold north-westerly winds from 
 the great winter anti-cyclone over Canada, and gives south- 
 westerly winds (the great cyclone track) to western Europe. 
 The chart recently published by H.S.H. the Prince of 
 Monaco 1 shows that the surface currents in the North Atlantic 
 
 Comptes rendus, t. cxiv., 8th Feb., 1892. 
 
THE ELEMENTS OF CLIMATE. 
 
 141 
 
 are in great measure determined by these prevailing winds, the 
 average velocity for the whole of the ocean amounting to 4^ 
 nautical miles in twenty-four hours, and the centre of 
 rotation coinciding closely with that of the Atlantic anti- 
 cyclone in the region of the Saragasso Sea. 
 
 The corresponding temperature effects are well shown in 
 Figs. 24 and 25, which divide the North Atlantic into 
 regions where the air is warmer or colder than the sea, or 
 where both have the same temperature, during the two 
 extreme months February and August. We observe, 
 generally, that in accordance with the principles just ex- 
 plained 
 
 (1) When wind and current move east or west, air and sea 
 
 have the same temperature, 
 
 (2) When both move northwards, the air cools faster than 
 
 the water, and the sea is warmer than the air. 
 
 (3) When both move southwards, the air heats faster than 
 
 the water, and the sea is cooler than the air. 
 
 Fig. 26 shows how these conditions determine the annual 
 range of air tempera- 
 ture, i.e. the difference 
 between the hottest 
 and coldest months 
 (August and Febru- 
 ary). Between 45 
 and 55 N. lat. the 
 prevailing westerly 
 winds coming from 
 the American conti- 
 nent are slowly cooled 
 in summer and warmed 
 in winter, the range "" Fig. 26. 
 
 decreasing to 20 F. in 49 W. long., and to 15 in 30* 
 
142 METEOROLOGY. 
 
 VV. long. Still further east, where the winds are more 
 southerly, the drift current from the Gulf Stream carries 
 a line of minimum range close to the coast of Europe, 
 the increase caused by the proximity of the Continent 
 being scarcely felt 200 miles from land. The retardation in 
 the annual changes of temperature, produced by the 
 same cause as that which diminishes their amplitude, 
 is well shown by the fact that, if we draw a line from 
 Valentia through Mullaghmore to the Pentland Firth, thence 
 to Heligoland, and through Danzig, Riga, and St. Petersburg 
 to Archangel, almost all stations north and west of this 
 line have the lowest monthly average of temperature in 
 February, while all to south and east have it in January ; 
 and in the former area the annual maximum is markedly 
 later than in the latter. 1 
 
 On the eastern seaboard of America the distribution is 
 considerably complicated by the presence of the Arctic or 
 Labrador current, which penetrates southward as far as 
 Cape Hatteras between the Gulf Stream and the coast. 
 
 If we follow the meridian of 65 W. from the coast of 
 Nova Scotia southward, we find a rapid decrease of annual 
 range, chiefly due to the warming of the north-westerly 
 winter winds of Arctic America; and after the Gulf Stream 
 is crossed the decrease becomes gradually slower. South 
 of 30 N. lat, in the trade wind region on the southern side 
 of the Atlantic anti-cyclone, the winds are always easterly, 
 and the currents take the same direction ; the smaller range 
 is simply caused by the greater uniformity of the seasons in 
 the lower latitudes. 
 
 Along with these chiefly mechanical effects we must take 
 into account the evaporation which goes on from the surface 
 of the sea. Oceanic winds, having an unlimited supply of 
 moisture, are never very far from the point of saturation ; 
 
 1 See also G. Schwalbe : Ueber die Maxima und Minima der 
 Temperatur. Berlin, 1892, 
 
THE ELEMENTS OF CLIMATE. 143 
 
 according to the Challenger observations the average relative 
 humidity in the North Atlantic is about 80 per cent. Hence 
 a very slight reduction of temperature, whether by radiation 
 or by contact with a cold body, produces fog or mist. Con- 
 densation of this kind occurs most frequently where a warm 
 air-current meets a colder, or passes over a region where 
 the surface of the sea is abnormally cold, or again where it 
 comes in contact with a cold land surface. We shall have 
 occasion later to discuss the last case in detail ; for the 
 others we need only mention such examples as the great 
 fog-banks of Newfoundland, the Norwegian coast, and the 
 coast of Peru. 
 
 The effect of strong evaporation upon the sea itself, in 
 relation to its influence on currents through increase of the 
 density of the water, belongs more properly to the province 
 of oceanography than of meteorology. 
 
 1 20. Continental Climates. The source of the great 
 differences between oceanic and continental climates lies of 
 course in the much greater variations of temperature to 
 which the air over a land area is exposed. But it must be 
 borne in mind that these excessive fluctuations are limited 
 to the surface of the earth and to the strata of air lying im- 
 mediately upon it. If we ascend an isolated peak, which 
 enables us to gain a considerable elevation, but does not 
 present sufficient horizontal surface to materially disturb the 
 temperature gradients as does a high plateau or range of 
 mountains, we find a gradual decrease in the variations 
 of temperature, and a transition to the climatic conditions 
 found over the ocean. For example, at the low level ob- 
 servatory of the Puy de Dome, situated in a valley where 
 radiation effects are strongly marked, the mean annual 
 range of temperature is 32'4 F., while at the peak, 3,500 
 feet higher, it amounts to only 24'3 F. 1 
 
 Since, then, we are dealing only with a thin layer next 
 iWoeikof, Met. Zeit., 1892, p. 373. 
 
144 METEOROLOGY. 
 
 the surface, it is evident that so long as the surface is level 
 there will be little tendency towards horizontal motion ; all 
 the temperature changes at the surface will simply go to 
 produce more or less local convection currents like the 
 familiar whirls on a dusty road on a calm hot summer day. 
 When the heating and cooling is unequal, pressure will be 
 on the whole less over the warmer than the colder areas, 
 and a movement of air will begin from the latter towards 
 the former. Along a coast line we find the diurnal 
 phenomenon of land and sea breezes : during the day the 
 land is hotter than the sea ; the air over it rises, and its place 
 is taken by air brought in by the sea breeze : again, during 
 the night the land is colder than the sea ; the air over it sinks 
 and flows towards the lower pressure as the land breeze. 
 Similarly, a land area, liable to be greatly heated and cooled 
 by radiation, shows the seasonal alternations analogous to 
 land and sea breezes, usually termed monsoons ; when the 
 land is hotter than the sea surrounding it, a constant wind 
 blows in towards it, and when colder, away from it. 
 
 121. Effect of Mountains. As we have said, little intensity 
 can be developed in either of these cases as long as the land 
 surface is level. In the northern parts of Central Africa, 
 where the power of the sun's rays is enormous and some of 
 the highest known temperatures have been recorded, there 
 is little circulation of a definite character ; at such stations 
 as Murzuk, Shimmedru, Ghadames, and Kuka almost every 
 second day is calm, while at Cairo calms average one day in 
 four. Although within the trade wind region, Alice 
 Springs in the heart of Australia averages more than one 
 calm day in three. 
 
 But if the land surface is elevated so as to form a cone, 
 or high plateau with long sloping sides, the thin stratum 
 lying next the surface will, when cooled, tend to flow down- 
 wards and outwards, and will flow faster the steeper the slope. 
 Again, when this stratum is heated it will tend to rise. We 
 
THE ELEMENTS OF CLIMATE. 145 
 
 know that the temperature of the whole mass of the atmo- 
 sphere decreases as we ascend. But the heating effect of 
 the sun's rays on a land surface is nearly the same at all 
 elevations, hence the greater the elevation the greater the 
 difference of temperature between the layer of air next the 
 ground and the great body of the atmosphere at that eleva- 
 tion, and therefore the greater the tendency to ascend. For 
 example, suppose the layer next the ground to be 100 feet 
 thick. Let the normal temperature at sea level be 40 F., 
 then, assuming an average vertical temperature gradient of 
 1 F., in 300 ft. for the whole of the lower atmosphere, the 
 normal temperature at a height of 3,000 feet will be 30 F. 
 If now the 100 feet layer be heated to 60 F. throughout, at 
 sea level it will be 20 F. warmer than the surrounding air, 
 but at a height of 3,000 feet on a mountain slope the differ- 
 ence will be 30 F., and therefore the ascending tendency 
 correspondingly greater. 
 
 Ferrel 1 aptly compares the case of warm and rarefied air 
 ascending a slope to that of warm air rising in a flue. If 
 the flue is horizontal there is little circulation, no matter how 
 hot the air in it compared with that outside ; but even a slight 
 inclination sets up a strong current, and the current is 
 stronger the greater the difference of temperature inside and 
 outside, as we know from the fact that the draught in a 
 chimney is best in cold weather. With the same tempera- 
 ture inside the flue, i.e. in the stratum of air next the 
 ground, we obtain the " cold weather " conditions at high 
 levels. 
 
 Hence, we find invariably that land and sea breezes 
 and monsoon winds never attain a marked development un- 
 less there are mountains or high lands in the immediate 
 neighbourhood. The strongest land and sea breezes known 
 are experienced at Port Royal in Jamaica, in the vicinity of 
 the Blue Mountains ; and, notwithstanding the compara- 
 
 1 Popular Treatise, on the Winds, p. 197. 
 
 K 
 
146 
 
 METEOROLOGY. 
 
 lively weak insolation and the many disturbing causes, the 
 high lands of Dartmoor produce marked land and sea 
 breezes at Plymouth during favourable anti-cyclonic condi- 
 tions. The north-east or winter monsoon of the Indian 
 Ocean, although really superposed on the trade winds, never 
 rises above a moderate breeze, while the south-west monsoon, 
 drawn towards the Himalayas, is usually almost a gale. 
 
 122. Mountains and High Land as Obstacles to Air 
 Currents. Assuming the existence of a wind blowing from 
 the sea to the land, we come to the effect produced on the 
 air current by meeting an obstacle, such as a range of hills. 
 Looking first at the purely mechanical problem, let us sup- 
 pose that the wind advances exactly at right angles to the 
 coast-line, and meets a straight line of cliffs, from the 
 summit of which a table-land extends for some distance in- 
 land. Here we have a case analogous to that of water 
 rushing out of a dock over the " sill." The main stream, on 
 meeting the sill, jumps clear of it, and travels for some dis- 
 
 ^^ tance before losing 
 
 its upward deflec 
 tion. Fig. 27 re- 
 presents a vertical 
 section through 
 the stream. On 
 the face of the sill 
 Fig . 27 . at A, pressure is 
 
 increased and the motion diminished, and above, at B, pres- 
 sure is diminished and the rate of motion increased. At C 
 matters begin to assume their normal course. The space B 
 forms a kind of slack water in which rotary or vortex motion 
 is set up, the vortices gradually "tailing off" towards C. 
 After passing C a certain amount of undulatory motion is 
 retained for a time, and waves C B' C', C B" C", etc., 
 similar to A B C, succeed each other, rapidly diminishing in 
 size. 
 
THE ELEMENTS OF CLIMATE. 147 
 
 Suppose, now, that the wall of the dock, instead of being 
 vertical as at A, slopes, as shown by the dotted line A'. 
 The upward deflection will then evidently be diminished, 
 and the undulations will be smaller the more A' is inclined. 
 Again, if the sill, instead of being horizontal, slopes down- 
 wards as at B', the undulating motion will be greatly in- 
 creased, and the slack water space at B, with its rotating 
 vortices, correspondingly enlarged. These conditions re- 
 present the case of wind passing over an irregular undulat- 
 ing country, or over the crest of a range of mountains. 
 Every inequality on the earth's surface deflects the air up- 
 wards from its weather side, and forms a region of confused 
 irregular whirls on the lee side. Whence we may form 
 some idea of the enormous difficulties in the way of study- 
 ing the general motions of the atmospheric ocean where our 
 observations are largely restricted to the eddying move- 
 ments round pebbles at the bottom. 
 
 The justness of the analogy between the motions of air 
 and water must be established by observations of the former. 
 These have been made by Cleveland Abbe x and others 
 with the help of kites, and much may be learned by ob- 
 serving the frequent smoke vortices on the lee side of a tall 
 chimney. But the most abundant data are afforded by the 
 winds themselves. 
 
 123. Effect of Mountains on Rainfall. When a current 
 laden with moisture from the sea is forced upwards by an 
 obstacle in its path its temperature is at once reduced and 
 the aqueous vapour it contains condensed, first into cloud, 
 then into rain. Hence, universally, the regions of greatest 
 rainfall are found where a prevailing wind from the sea meets 
 high land at a short distance from the coast. The British 
 Islands, where with prevailing winds from the Atlantic and 
 a watershed running nearly north and south, the annual 
 
 1 See his elementary discussion of this subject. Repor. of Chief 
 Signal Officer, 1889, Appendix 15. 
 
148 METEOROLOGY. 
 
 rainfall on the western side averages nearly one half more 
 than that on the eastern, are a familiar example. Still more 
 striking are the effects of the Western Ghats and the Khasia 
 mountains on the south-west monsoon. The mountains 
 being excessively steep, the monsoon current has to rise 
 through a great height while crossing only a small tract of 
 country, and we accordingly find such phenomenal falls as 
 250 inches at Mahableshwar, and 474 inches at Chirra Punji. 
 We have here obviously another force strengthening 
 the ascending currents in mountainous regions. The 
 latent heat set free by condensation goes to raise the 
 temperature of the air, and therefore to increase the tendency 
 to ascend. 
 
 After the obstacle is surmounted the air, largely deprived 
 of its moisture and no longer forced to ascend, becomes dry, 
 and it is not, in general, possible to follow its motion by 
 means of cloud. The region of vortex motion correspond- 
 ing to B (Fig. 27) can, however, frequently be recognised ; 
 for, since pressure is there somewhat reduced, the air coming 
 over the summit is drawn in, and being cooled by expansion 
 again becomes saturated, and a thin wreath of cloud is 
 formed which presents the appearance of a "smoking 
 mountain." 
 
 124. Helm and Helm Bar. When, after crossing a 
 mountain ridge, the current begins to descend on the other 
 side, it is sometimes unable to dislodge the mass of air lying 
 below, which then acts like the dock sill and continues the 
 undulating motion. The current reflected upward is thus 
 again cooled, sometimes until another cloud is formed. 
 Hence the phenomena of the " helm and helm bar " seen on 
 Cross Fell and the Eden Valley, and the "Table Cloth" 
 and bar of Table Mountain. 1 The application of this 
 principle may be further extended by supposing the obstacle 
 analogous to the dock sill to be wholly air, and we are 
 iSee Quart. Jour. Met. Soc., xv., p. 103. 
 
THE ELEMENTS OF CLIMATE. 149 
 
 brought back to the frictional effects of an upper current 
 moving upon an under, and the formation of waves and 
 ripples on the surface of the latter as shown by " mackerel 
 skies," "Noah's Arks," and other forms of cirro-cumulus 
 cloud. 
 
 125. Obstacles not at Right Angles to Currents. For the 
 sake of clearness we have restricted ourselves to the case 
 of currents meeting an obstacle exactly at right angles. 
 In the general case where the obstacle is more or less in- 
 clined, part of the current is forced upwards, and part 
 deflected so as to flow along the side of the obstacle, the 
 relative proportions of the two parts depending on the angle 
 of inclination ; just as in a stream dammed by a weir part 
 of the water flows over the weir and part is deflected into 
 the mill-race. The rainfall of India during the south-west 
 monsoon affords an excellent opportunity of studying such 
 effects, little or no rain falling in regions where the current 
 runs parallel to the mountains at a time when excessive 
 amounts are recorded where the mountains are at right 
 angles to it. 1 
 
 We conclude then, generally, that the effect of high land 
 or mountain ranges, towards which winds blow from the sea, 
 is greatly to increase the cloudiness and rainfall on the 
 weather side, and, as it were, discharge the oceanic qualities 
 of the current so that on the lee side the climate becomes 
 more that of a continent. The air, deprived of its moisture, 
 sinks down under the shelter of the mountains, and especi- 
 ally in valleys remains comparatively undisturbed. This 
 fact is at once apparent from the increase of solar and 
 terrestrial radiation, as indicated by greater variations of 
 temperature. In Fig. 28 lines are drawn through those parts 
 of the British Isles having the same average annual range of 
 temperature, and these show that, wherever the prevailing 
 south-westerly winds meet elevated land, the range of tem- 
 1 See Scottish Geographical Hay., 1892, p. 248. 
 
ISO 
 
 METEOROLOGY. 
 
 perature rapidly increases, and the maximum range in- 
 variably occurs in valleys and low country on the lee side, as 
 in Salisbury Plain, the Central and Eastern Plains, the 
 Plain of York, and Strathmore in Scotland. Again, where- 
 ever the belt of high land is interrupted the oceanic charac- 
 
 Fig. 28. 
 
 teristic of small range is retained much further eastward, as 
 in the Vale of Severn, the Cheshire Plain, the district be- 
 tween the Pennine Chain and the Cambrian Mountains, 
 and the Plain of Forth and Clyde. 
 
THE ELEMENTS OF CLIMATE. 151 
 
 
 
 126. Variability of Temperature. Another method of 
 discussing changes of temperature, introduced by Hann, 
 shows the differences of climate in another way. If we cal- 
 culate the mean difference of the temperature of each day 
 from that of the next, we get the variability of the tempera- 
 ture, or probable change from day to day. Now, Hann's 
 investigations show that the variability of temperature in- 
 creases from the coast inland; 1 and, applying this to the 
 British Isles, we find a variability of i'9 F. at Valentia and 
 Falmouth, 2*4 at Aberdeen, 2 0> 5 at Armagh, Glasgow, and 
 Stonyhurst, 2"j at Kew, 3-! at Oxford, and 3 -4 at 
 Makerstoun. 2 The two last named stations show a varia- 
 bility equal to that of such truly continental places as Cassel, 
 Prague, or Budapest. 
 
 127. Variations in the Nature of Surface: Snow-cover- 
 ings. The distribution of land and sea, and the configura 
 tion of the land surfaces are, of course, constant quantities 
 so far as our present purposes are concerned. We come 
 next to variations due to different kinds of land surfaces. 
 In a former chapter it was found that the conditions of 
 temperature and moisture varied in different soils and 
 exposures, and under different kinds of vegetation. These 
 will be more conveniently discussed at a later stage, for they 
 involve considerations outside the domain of meteorology 
 proper. 
 
 The researches of Woeikof 3 and others have shown that 
 apart from its effect on the soil the presence of a snow 
 surface during a considerable part of the winter has a 
 marked influence on climate. According to Zenker, a snow 
 surface reflects about \ of the sun's rays falling upon it, 
 while clear ground reflects only about -^ hence snow is 
 much less warmed during the day ; and since during winter, 
 
 t. Zeit., 1892., p. 46. 
 See R. H. Scott, Proc. R. S., vol. xlvii., p. 303. 
 3 See Dzr Einflms einer Schneedecke: Geog. Abhan-]., Wien, 1889. 
 
I $2 METEOROLOGY. 
 
 under the anti-cyclonic conditions favourable to sunshine, 
 the air is exceptionally dry and free from dust, little heat is 
 absorbed by it. Again, from its feathery structure, dry snow 
 contains large quantities of air, and it is therefore an ex- 
 tremely bad conductor of heat ; and at the same time, from 
 the nature of its surface, it radiates heat more freely at night 
 than the surface of the ground. Thus with the same weather 
 conditions the surface has a lower temperature when covered 
 with snow, and the deficiency being communicated by con- 
 duction to the layer of air in contact with it, tends to 
 increase the reversal of the vertical temperature gradient 
 noticed in winter anti-cyclones ( 71). The cold layer, on 
 account of its weight, lies close to the snow surface, and as 
 long as the snow remains, temperature seldom rises to the 
 melting point even with strong sun. The rough surface of 
 the snow offers considerable resistance to horizontal motion 
 of air, and hence warm winds coming from other regions are 
 weakened, and tend to rise above the cold surface layer. 
 Snow is therefore favourable to the development and per- 
 sistence of anti-cyclonic conditions, and a thaw is usually 
 due to strong external influences, powerful enough to break 
 up the anti-cyclone and dislodge the masses of cold air. In 
 northern continental areas, such as Russia and Siberia, the 
 extensive snow-sheets, lasting all through the winter, no 
 doubt considerably increase the intensity of the anti-cyclones 
 and the severity of the cold. 
 
 When the temperature of the air rises, the limitation im- 
 posed by the fact that the snow cannot become warmer than 
 32 F. makes itself felt. The warm air blowing from the 
 open sea or from land free of snow, is cooled by contact 
 with the snow, and still more by the abstraction of heat 
 required for melting. Further, as the melting goes on the 
 snow changes into ice ; it no longer contains vast numbers of 
 air spaces, and therefore becomes a better conductor of heat. 
 Again, as the snow disappears, the ground is saturated with 
 
THE ELEMENTS OF CLIMATE. 1 53 
 
 ice-cold water, and more heat is required to raise the 
 temperature of the surface. Thus on the whole a thick 
 snow-covering tends to intensify and continue the weather 
 of winter, and to prevent the warmer weather of spring 
 taking effect. An exceptionally cold winter is not so likely 
 to be followed by a late spring as one in which the snowfall 
 is unusually large. 
 
 128. Recurrent Changes of Weather Type. Having 
 sketched the main features of climate as determined by the 
 distribution of land and sea, we may notice some results of 
 the normal changes from one type of weather to another at 
 different seasons of the year ; reverting, in fact, to the con- 
 sideration of weather as a variable. As our purpose is 
 simply to illustrate a method, we confine ourselves to one or 
 two examples. 
 
 129. Interruptions of Temperature. By calculating the 
 mean temperature of every day in the year at a number of 
 representative stations in Scotland, Buchan 1 discovered the 
 occurrence of certain interruptions in the annual march of 
 temperature, which he arranges as follows : 
 
 Six cold periods 
 
 1. yth to 1 4th February. 
 
 2. nth to 1 4th April. 
 
 3. Qth to I4th May. 
 
 4. 29th June to 4th July. 
 
 5. 6th to nth August. 
 
 6. 6th to 1 3th November. 
 
 Three warm periods 
 
 1. 1 2th to 1 5th July. 
 
 2. 1 2th to 1 5th August. 
 
 3. 3rd to 1 4th December. 
 
 1 Jour. Scot. Met. Soc., vol. ii., 1869, p. 4. 
 
1 54 METEOROLOGY. 
 
 Some of these periods are familiar in weather lore. The 
 second cold spell, which according to the old style occurs 
 in the beginning of April, includes the " Borrowing Days." 
 
 " March borrows frae April 
 Three days, and they are ill. 
 The first o' them is wun' and weet ; 
 The second it is snaw and sleet ; 
 The third of them is peel a-bane, 
 And freezes the wee bird's neb tae stane." 
 
 Similar sayings are found in many parts of Europe, notably 
 in France and Andalusia. The third cold period is re- 
 marked all over Europe, and is associated with the " ice 
 saints," Mamertus, Pancratius, and Servatius, whose names 
 the Bohemians run into Pan Serboni. They say "Pan 
 Serboni withers the tree." So also in Russia " St. Isodor 
 (i4th May) is past ; the north winds are over." 
 
 Buchan shows that all these interruptions depend on 
 anomalies in the prevailing winds, fogs, or the rainfall, and 
 they can be connected with a tendency to periodic changes 
 in the type of weather recurring about the same time every 
 year. 
 
 130. Distribution of Rainfall. The changes of weather 
 type are perhaps most marked in their effect upon the rain- 
 fall. Since a change of type usually involves a redistribution 
 of pressure and temperature over a very large area, it is 
 easily understood that if the normal sequence is disturbed, 
 such disturbance is likely to be persistent for a considerable 
 time. If, for example, the Atlantic anti-cyclone does not 
 extend as usual in early summer, it is probably prevented 
 from doing so by a cause which will delay its extension for 
 a considerable time, as happened in 1888, when easterly 
 winds and cloudy skies prevailed almost throughout the 
 summer. Hence, when a change of type brings with it a 
 
THE ELEMENTS OF CLIMATE. 155 
 
 periodic fall of rain, that fall is usually either up to time, or 
 is considerably delayed. This can be abundantly illustrated 
 from folk-lore. 
 
 " If St. Swithin weeps, the proverb says, 
 The weather will be foul for forty days," 
 
 applies to the summer rains in the British Isles. In other 
 countries these rains are looked for at different dates, and 
 we find the saint altered to suit; in the Haute Marne 
 and Loire, Sf. Medard (June 8th) ; in the Province of Sarthe, 
 St. Calais (June ist) ; Visitation of B. V. M. (July 2nd), in 
 Belgium and Austria; St. Anne (July 26), in Northern Italy, 
 and so on. 
 
 We remark again that the method involved here is simply 
 that of probabilities ; and in making practical application of 
 it the value of the probability must always be carefully de- 
 termined. 
 
CHAPTER VIII. 
 
 APPLICATION OF METEOROLOGY TO AGRICULTURE. 
 
 131. Relation of Meteorology to other Sciences. The work of 
 forecasting weather, and in particular of giving timely warn- 
 ing of destructive gales, is the first and most obvious 
 practical application of meteorological science. From the 
 sketch we have given of the general methods employed it 
 appears that the problem is, in its larger aspects, ^Yholly 
 physical ; ultimately a question of motion of a given fluid 
 under given thermal conditions, and that the difficulties con- 
 sist in completely defining the properties of the fluid and 
 the conditions under which it moves. 
 
 But since no branch of science is independent of others, 
 it is necessary for the specialist in one department to meet 
 that in another half way, to arrange his data and results in a 
 form which facilitates comparison. Specially is this the case 
 with the group of subjects which has recently taken the family 
 name of Physiography. Meteorology involves geography. 
 Many geographical relations are dependent upon meteorology. 
 Biology must take account of the conditions of life deter- 
 mined by climate, and these are the geology of the present. 
 Narrowing the matter still further, the meteorologist finds 
 three classes of practical questions with which he must deal : 
 those of physical geography, or the influence of climate 
 directly upon mountains and rivers, land and sea, etc. ; those 
 of zoology, the distribution of animals, and, economically 
 the most important, of food fishes ; and those of botany, the 
 
 156 
 
APPLICATION TO AGRICULTURE. 157 
 
 distribution of plants, agricultural products most important 
 economically. 
 
 Of the first class it is unnecessary to speak further, inas- 
 much as meteorological results are so largely controlled and 
 modified by geographical conditions that they are perforce 
 expressed, as it were, in the first instance, in geographical 
 terms, and the geographer finds his meteorological facts, so 
 far as they are known, ready to his hand. 
 
 The relation of climate to the distribution of animals 
 can as yet scarcely be said to admit of special discussion by 
 the meteorologist except perhaps with regard to the human 
 species. The range of any given species depends directly 
 upon this element to only a very limited extent, according 
 to its power of withstanding heat or cold, dampness or dry- 
 ness, or of adapting itself to these varied conditions ; for 
 intermediate factors come in, such as the distribution of 
 animal food, of natural enemies and the like, which in turn 
 depend both directly and indirectly upon climate. We 
 should expect a priori \\~\2A. the fauna of any region would de- 
 depend ultimately upon its climate, and of course we have 
 rough classifications according to frigid, temperate, and torrid 
 zones, but a closer relation remains to be traced. It may 
 be said that in the case of food fishes a coincidence between 
 particular weather conditions and the presence or absence 
 of certain species has been clearly indicated, but the relation 
 is not direct ; whether it depends upon surface forms upon 
 which the fish feed or is connected with periodic reproduc- 
 tive changes is evidently not a meteorological question. 
 
 In the third case, that of plant life, the biological condi- 
 tions are simpler, and the relations of climate at the same 
 time more direct. For the individual plants cannot of 
 themselves move from place to place, and the elements of 
 climate are therefore so far constant and simpler than in the 
 case of, for example, migrating animals. Again, plants 
 derive their food partly from the air and partly from the 
 
158 METEOROLOGY, 
 
 soil, which is itself a chemical and mechanical product of 
 meteorological action on known types of rock-formation. 
 
 There is thus a sufficiently definite problem stated, the 
 final solution of which must of course rest with the botanist 
 and agriculturist, but which requires the application of 
 special methods to the meteorological data. 
 
 132. Relations of Soil to Climate: Different kinds of Soils. 
 The relations of soil to climate, in so far at least as the 
 United States are concerned, have been discussed by Pro- 
 fessor Hilgard 1 in a manner which affords excellent guidance 
 to the agricultural meteorologist. Soil is defined as the 
 residual product of the action of meteorological agencies 
 upon rocks. First, mechanical agencies (a) changes of 
 temperature affecting the several constituent minerals of 
 rocks differently, and thereby forming minute cracks which 
 give access to water and air ; (b) freezing water widening 
 these cracks and breaking the rock up into fragments ; (c) 
 ice or glacier action, reducing rocks to fine powder, remov- 
 ing the material from the parent rock and depositing it as 
 " till " or boulder clay ; (d) flowing water, the most active 
 agent at the present time, acting in a manner similar to 
 glaciation but with less intensity and over a wider area so 
 as to produce a deposit more variable in the size of its par- 
 ticles and more mixed in its ingredients. Second, chemical 
 agencies (a) solution by water alone, according to the solu- 
 bility of the rock-forming minerals; (b) carbonic acid, always 
 present in the air as the product of volcanic action, decay, 
 'fermentation, combustion, and the breathing of animals, 
 being absorbed by rain water, greatly increases its solvent 
 properties, decomposing the ingredients of the soil and ren- 
 dering them available for plant nutrition; (c) oxygen, acting 
 directly as well as in carbonic acid, chiefly in the formation 
 of " rust" from the lower oxide of iron common in most green 
 
 1 U.S. Department of Agriculture, Weather Bureau, Bulletin No. 3. 
 1892. 
 
APPLICATION TO AGRICULTURE. 159 
 
 and black minerals ; (d) water, entering directly into com- 
 bination with substances present or newly formed in the 
 rocks, and thereby increasing their bulk and breaking up the 
 original structure ; as in the formation of clay from felspars. 
 
 These forces produce different results according to the re- 
 lative parts played by each, and to the nature of the material 
 upon which they act. Soils may be classified according to 
 position into residual or sedentary, which rest upon the rock 
 from which they have been formed by weathering, and 
 transported soils, which are either colluvial moved a short 
 distance by water, gravity, or wind, without undergoing any 
 arrangement as to size of particles, or alluvial deposited 
 regularly in beds or strata from flowing water. Or, they may 
 be regarded as consisting of rock, clay, and vegetable matter, 
 each in different proportions and in different stages of 
 powdering and decomposition ; and as such described as 
 light or sandy, heavy or clayey, and humus. 
 
 133. Action of Weather in Formation of Soils. The 
 climatic factors involved in the formation of soil are there- 
 fore chiefly the temperature and the amount and seasonal 
 distribution of the rainfall. According to Hilgard, temper- 
 ature changes take effect through the greater chemical 
 activity at high temperatures and the more rapid breaking 
 up of material when the range of variation is large. Rainfall 
 acts for the most part by leaching or washing out the more 
 soluble ingredients from soils already formed by weathering, 
 to an extent depending on their permeability : light sandy 
 soils maybe washed to a great depth, while heavy clays allow 
 the water to run directly off their surface. The formation 
 of clay itself is an important instance of this action. When 
 a granitic rock has been crumbled down by mechanical 
 agencies variations of temperature, freezing of water in its 
 pores and so on the rain water containing carbonic acid 
 removes the potash from the felspar and mica in the form 
 of carbonate of potash, while the silicate of alumina and the 
 
160 METEOROLOGY. 
 
 quartz are separated by the action of the water ; the former 
 being much lighter than the latter, is washed away and re- 
 deposited as clay. This process is obviously impossible 
 where the rainfall is scanty, and loose sandy or dusty soils 
 accordingly predominate in arid climates. Hence also the 
 sharp distinction between surface and sub-soils, familiar in 
 humid regions, does not occur in dry. Hilgard quotes 
 an instance of tomatoes, water-melons, etc., growing better 
 on soil freshly dug from a depth of 7 to 10 feet than on sur- 
 face soil, in the dry climate of Nevada City, California. 
 
 The formation of humus or vegetable mould is also 
 largely dependent upon climate. Vegetable matter exposed 
 to the influence of hot rainless summers is " burnt up " 
 destroyed by a slow combustion which leaves little residue 
 beyond the mineral ash. On the other hand, when saturated 
 with or submerged in water it is transformed into a sub- 
 stance of which the extreme form is peat a sour soil, 
 partly soluble, as we know from the bitter coffee-coloured 
 water which drains from it. Such land must be reclaimed 
 by drainage and the use of lime, which neutralizes its acidity 
 and gradually changes the peaty matter into insoluble black 
 humus. This last, the true black mould desired of the 
 farmer, is formed by decomposition underground in pro- 
 perly drained land, the essential conditions being a certain 
 amount of moisture, sufficient to keep the decaying matter 
 moist, but without stagnation to allow of its becoming acid 
 or sour. Hence humus is much less plentiful in dry than 
 in humid climates. In the arid districts of America strawy 
 manure cannot be ploughed-in raw from the lack of moisture 
 to cause fermentation. It is necessary either to " cure " the 
 manure before laying it down, or to moisten by irrigation ; 
 otherwise it is simply reduced to ashes by slow combustion, 
 a process which may occupy two or three years. 
 
 In like manner the soils of humid and arid regions may be 
 contrasted with reference to other ingredients. Lime is in 
 
APPLICATION TO AGRICULTURE. l6l 
 
 the former dissolved out in the form of carbonate and trans- 
 ferred from the surface into the sub-soil, and from uplands 
 into lowlands and valleys ; while in the latter it remains more 
 uniformly distributed. Poor land is for this reason less 
 frequent in arid regions, which for the most part become, 
 when irrigated, amazingly fertile. Again, in the process of 
 clay-making, zeolites are usually formed ; and in these com- 
 plex hydrous silicates, potash wherever present takes pre- 
 cedence of soda. Hence when a potash solution comes in 
 contact with a zeolite containing soda, the soda is replaced 
 by potash and washed away. Soils retain potash tenaciously, 
 but soda only when the rainfall is insufficient to wash it out, 
 as in the case of alkali lands. 
 
 Many of these results, derived by Hilgard from observa- 
 tions of natural processes on a large scale, are simply 
 further confirmation of those arrived at by Sir J. B. Lawes 
 and Dr. Gilbert at the experimental station at Rothamsted, 
 from which we may extend our consideration of the same 
 modes of action to their application to the behaviour of 
 various manures, the development of roots, and so on. 
 Thus with dry weather between autumn and spring the soil 
 retains the nitrates left over from former dressings, which 
 would otherwise be drained off; a mild spring allows roots 
 to penetrate to a great depth in search of food ; and favouring 
 rains later in the season become concentrated solutions of 
 plant food as they soak into the soil. 
 
 It is, therefore, abundantly evident that a close examina- 
 tion of the effects of climatic agencies on soil alone is of the 
 first importance in agriculture. The points to be studied 
 are (i) temperature of the soil in relation to the tempera- 
 ture of the air and to radiation depending on (a) the nature 
 and colour of the surface, (b) the exposure, (c) the con- 
 ductivity of the soil, (d) drainage ; (2) moisture of the soil 
 in relation to rainfall depending on (a) its permeability 
 
 L 
 
1 62 METEOROLOGY. 
 
 (b) retentive power. Much remains to be done in in- 
 vestigating some of the heads just enumerated, and con- 
 sidering the endless variety of conditions involved it must 
 always be difficult to apply general rules to any particular 
 case. We may, however, state the results arrived at in some 
 instances. The temperature variations of the surface of the 
 soil and their relations to the nature of that surface, and to 
 the conductivity of the soil, have already been noticed ( 33). 
 When the rainfall is fairly distributed throughout the year, 
 and the winter snow-covering interrupted and irregular, the 
 mean temperature of the soil closely agrees with that of the 
 air ; but where there are prolonged seasons of drought, or 
 where the ground is long covered with snow, considerable 
 differences may occur. The greater diathermancy of the air 
 in the higher layers of the atmosphere causes a marked in- 
 crease in the power of the sun's rays in mountainous regions, 
 hence while the air is not heated to the same extent the soil 
 becomes relatively much warmer, a fact of great importance 
 in connection with the distribution of plants and animals j 1 
 for it follows that we cannot strictly regard a given change 
 of height as equivalent to a corresponding change of 
 latitude. 
 
 134. Temperature of the Soil. The propagation of surface 
 changes of temperature downwards into the ground takes 
 place in the form of a wave whose amplitude decreases steadily 
 till it ceases to be sensible ; the rate of propagation and of 
 diminution of amplitude depending on the nature of the soil 
 or rock and the extent of the surface fluctuations. It may 
 be assumed generally that daily variations of temperature 
 are insensible at depths greater than 3 to 5 feet, and annual 
 variations 30 to 50 feet. Observations at Edinburgh and 
 Greenwich have shown that at a depth of 25 feet the annual 
 maximum and minimum occur about six months later than 
 1 See Hann : Handbuch der Klimatoloyie, p. i46. 
 
APPLICATION TO AGRICULTURE. 163 
 
 at the surface. The averages of ten years' observations at 
 Eberswalde x at the surface, 6 inches, i foot, 2 feet, 3 feet, 
 and 4 feet, show the normal distribution of temperature in 
 the soil very clearly. In January, the temperature at the 
 surface remains near the freezing point ; at a depth of 4 feet 
 a minimum of 35'6 F. is reached about February 2oth. 
 The maximum at the surface (about 70 F.) occurs in the 
 middle of July, and at 4 feet (about 61 F.) early in August. 
 Temperature gradients are therefore very much steeper in 
 summer when the surface is warmest than in winter when it 
 is coldest. This series of observations, made in the open 
 field, is of special value in showing in detail the manner in 
 which the reversal of the gradients takes place, a subject 
 rather beyond the scope of this book, and also from its 
 association with a similar series made in a neighbouring 
 pine forest. Comparison shows that in winter the soil of 
 the forest is somewhat warmer than in the open field, the 
 excess increasing with the depth ; while in summer the 
 difference is reversed, the forest soil being some 5 F. colder 
 at the surface and 4 feet below it, and rather less at inter- 
 mediate depths. The dates of maxima and minima are de- 
 layed some five to ten days at all depths. Similar con- 
 clusions with regard to the soil in the open have been 
 reached by Buchan, 1 who shows from observations at 
 several stations in Scotland the marked liability of loose 
 sandy soils to excessive warming and intense frost, as com- 
 pared with heavy clayey soils, on account of the smaller 
 conducting power of the former. Buchan also obtains im- 
 portant results with regard to the effect of drainage on soil 
 temperature. The mean annual temperature of arable land 
 is, according to his observations, raised nearly i F. by 
 
 1 Schubert, Monats-u. Jahresmittel der Bodentemperaturen. Zeit. 
 fur Forst-u. 'Jagdivesen, January, 1888. 
 1 Jour. Scot. Met. Soc., ii., p. 273. 1869, 
 
1 64 METEOROLOGY. 
 
 drainage, and of hill pasture land nearly half that amount. 
 Cold penetrates more readily into undrained soil, and less 
 benefit is derived from high temperature through the loss of 
 heat by evaporation. We have already seen that the effect 
 of forest growth is to moderate the fluctuations of soil 
 temperature, and the Scottish observations show that the 
 same is true of all vegetable covering a black surface of 
 bare soil being a good radiator and absorber. These points 
 are of great importance with reference to the change of 
 temperature experienced by the roots of plants when they 
 attain a growth sufficient to form a " braird " or covering ; 
 as well as to their reaction on climate. 
 
 The modifying effect of exposure is,, of course, consider- 
 able. Observations in the Innthal and Gschnitzthal near 
 Innsbruck, discussed by F. von Kerner, 1 show that at a 
 depth of about i foot, the warmest soil in summer is that 
 having a S.E. exposure, but in winter S.W., and the 
 coldest that having a northerly exposure, except in January 
 and December, when the coldest exposure is east. The in- 
 fluence of exposure is most marked in spring and autumn, 
 less so in summer and winter ; because the sun in its up- 
 ward course attains nearly full power on a surface much in- 
 clined southward almost before it begins to be felt on one 
 equally inclined to the north, and retains that power longer 
 on its downward course. The following table, which gives 
 the radiation received by surfaces inclined to N. and S. at 
 an angle of 36, expressed in percentages of what would be 
 received if the rays fell perpendicular to them, will make 
 this clear. 
 
 Height of Sun. S. Exposure N. Exposure. Difference. 
 
 19 27' 58-6 o'o 58-6 
 
 42 55' 73*3 0-8 72-5 
 
 54 39' 7' I2 '4 62*6 
 
 60 23' 72-4 28*0 44*6 
 
 1 Siteb. der Wiener Akad., May, 1891. 
 
APPLICATION TO AGRICULTURE. 165 
 
 The fact that a S.E. exposure is the most favourable during 
 summer is probably due to the increase of cloudiness 
 towards afternoon. It is, of course, unsafe to draw general 
 conclusions from a single series of observations like that 
 just quoted, but it is important to note the shortening of the 
 warm season as well as the diminished temperature in the 
 case of a northern exposure. 
 
 135. Moisture in the Soil. According to Wollny, 1 the 
 effects of rainfall upon a soil apart from loss by evaporation 
 are determined by (i) the surface drainage, (2) the 
 vegetative covering, (3) the nature of the soil. In agri- 
 cultural work the great object is to secure a dry or 
 thoroughly porous surface soil, and at a proper depth, a 
 permanent supply of moisture : the surface soil allows a 
 supply of air to roots, and the soluble plant foods are not 
 washed out, while the roots are encouraged to extend to a 
 greater depth in search of moisture, increasing their supply 
 of food and lessening the risks of damage from surface frost 
 at the same time. The effect of a dry surface layer of soil 
 in maintaining a supply of moisture in the sub-soil is well 
 illustrated in the case of New South Wales, where in 
 summer the surface of the ground is often baked hard and 
 dry when the sub-soil is damp enough to keep weeds green 
 and growing, 2 and its loss by evaporation is found to be 
 almost nil. 
 
 The proportion of the rainfall which simply flows off the 
 surface of the soil depends upon the inclination, the ex- 
 posure most with N., next W., E., and S. the " closeness " 
 of the surface, and the covering most from bare land. 
 The interception of water by vegetation increases with the 
 
 1 Untersuch. uber das Verhalten der Niederschtage zur Pflanze und 
 zum Boden. Forsch. auf dem Gebiete der Ayrikulturphysik, xii., 13, 
 p. 316. 
 
 2 H. C. Russell, Rainfall Observations, Sydney, 1886, p. 14. 
 
1 66 METEOROLOGY. 
 
 density of the covering ; cultivated land may receive as 
 much as 31 per cent, less water when thickly covered than 
 when bare. 1 We may therefore conclude that even with 
 the same soil the quantity of water transmitted to any 
 depth must be extremely variable. Other things being 
 equal, it depends chiefly on the coarseness of the particles 
 the finer the grain of the soil, the greater difficulty the 
 water has in penetrating; but we have also to take into 
 account, apart from the rainfall, the water condensed 
 directly from the atmosphere on the earth particles, which 
 Giseler (Berggdst^ 1878) has shown may amount to as much 
 as 37 per cent, of the recorded rainfall ; and again we must 
 allow for the condition of the soil, for the same soil when 
 dry will not transmit the same proportion of the water falling 
 on it as when wet, and a frozen soil will not transmit water 
 at all. 
 
 The net result of all these conditions, varying as they do 
 from month to month, is well shown for the case of a 
 heavy loam soil by the following averages of eleven years' 
 observations at the experimental station at Rothamsted. 
 These give the percentage of rainfall which reached a 
 depth of 20 inches and 60 inches from a carefully weeded 
 surface. 
 
 20 inches 60 inches 
 per cent. per cent. 
 
 Jan. .... 85 92 
 
 Feb. .... 86 81 
 
 Mar. .... 49 55 
 
 April .... 36 39 
 
 May . 28 28 
 
 June .... 27 27 
 
 July .... 28 18 
 
 Aug. .... 32 28 
 
 Sept. . . 37 32 
 
 1 Wollny, loc. cit. 
 
APPLICATION TO AGRICULTURE. 1 67 
 
 20 inches 60 inches 
 per cent. per cent. 
 
 Oct. .... 60 54 
 Nov. . . . . 76 72 
 Dec. .... 80 79 
 
 Mar. June . . -35 37 
 July Sept. ... 32 26 
 Oct. Feb. ... 76 76 
 
 Year . . . . 51 50 
 
 Ebermayer's x observations enable us to extend these re- 
 sults to different soils. The following means are based 
 upon five years' observations of (i) coarse-grained quartz 
 sand, (2) fine-grained ditto, (3) loose loam, (4) fine cal- 
 careous sand, (5) black peat, contained in pits nearly 4 feet 
 deep. 
 
 PER CENT. OF RAINFALL PASSED. 
 
 Soil. 
 
 Year. 
 
 Winter. 
 
 Spring. 
 
 Summer. 
 
 Autumn. 
 
 i 
 
 86 
 
 100 
 
 74 
 
 84 
 
 85 
 
 2 
 
 107 
 
 129 
 
 96 
 
 104 
 
 104 
 
 3 
 
 94 
 
 125 
 
 98 
 
 77 
 
 98 
 
 4 
 
 48 
 
 66 
 
 30 
 
 39 
 
 5i 
 
 5 
 
 39 
 
 56 
 
 37 
 
 34 
 
 38 
 
 Here the condensation in quartz sand, and even in loam, is 
 very marked. Clay and humus allow little water to pass 
 through; finely-divided humus, mixed with mineral earth, 
 possesses greater absorbing power than pure peat, and there- 
 fore diminishes the amount of water transmitted with increase 
 of humus through cultivation, and with cultivation of strong 
 
 1 UntersucTi. uber flic Sickerwassermenyen in versch. Bodenarten 
 Wollny's Forsch. Bd. xiii. 
 
1 68 METEOROLOGY. 
 
 transpiring plants the proportion also diminishes, an impor- 
 tant fact in relation to the supply of springs. 
 
 136. Effects of Climate on Plants. We have seen that 
 the supply of water to the soil exercises an influence on 
 vegetation chiefly through the selection and arrangement 
 of plant food : and when that supply is provided by means 
 of irrigation, plants grow and thrive quite as well as when it 
 falls as rain. We are therefore justified in concluding that 
 rainfall acts upon plants chiefly through the roots, and does 
 not in general sensibly affect the parts above ground. On 
 the other hand, we find that the temperature of the soil 
 during the growing season is below that of the air, and are 
 familiar with cases like that of the vine, which is often 
 grown under glass at a high temperature, while the root is 
 placed in ground fully exposed to the external air ; whence 
 it appears that the temperature of the soil is not directly of 
 first importance to the plant. It is possible, as A. de Can- 
 dolle points out, that the soil may have some influence in 
 modifying extremes of temperature through the supply 
 of sap, especially in parts of the plant where, from the nature 
 of the surface, conduction of heat and evaporation of mois- 
 ture go on slowly. Thus, for example, the low temperature 
 in the heart of the coco-nut is probably almost the same as 
 that of the sap at the root. But it is obvious that for the 
 most part temperature affects the growth of the plant above 
 ground. 
 
 The inquiry, therefore, becomes much more complicated. 
 For the word temperature as employed by the meteorologist 
 means merely the temperature of the air in a shaded screen 
 at a height of 4 feet above the ground ; while a plant rises 
 from the ground, gradually increasing in height, and is 
 exposed to great diversities of light and shade. Other con- 
 siderations also come in, for we know that plants require 
 light as well as heat, and cannot assume that they do not 
 
APPLICATION TO AGRICULTURE 169 
 
 also require actinic rays. Plants require a supply of radiant 
 energy to enable them to decompose carbonic acid and 
 water in the leaf-cells, and it appears that the radiations 
 most available to them for the purpose are of a special wave- 
 length, or are arranged of special wave-lengths in certain 
 proportions, 
 
 137. Chemical Rays. To begin with rays of short wave- 
 length. Sir William Siemens x has found that many species 
 of plants grow freely and attain a full development of flower 
 and fruit under the electric light, provided a screen of clear 
 glass is interposed. On exposing the plants to the naked 
 arc light they soon withered and shrivelled, but when the 
 rays were filtered through glass a healthy growth was main- 
 tained. Now Stokes showed, in 1853, that rays from the 
 electric arc are particularly rich in ultra-violet chemical rays, 
 and that the greater part of these are intercepted by clear 
 glass, i.e. that glass is opaque to them, just as it is opaque 
 to the invisible heat rays from a source of lower temperature 
 like a coal fire. Siemens therefore concluded that rays of 
 short wave-length, when present in anything like excess, are 
 not favourable but fatal to plant-growth. This result is 
 again important in view of the absorption of blue rays by 
 the atmosphere (31), showing that the change of conditions 
 experienced by plants at different elevations is not equivalent 
 to that in different latitudes. 
 
 138. Rays most Useful to Vegetation. Since, then, dark 
 chemical rays are injurious to vegetation, and dark heat 
 rays alone insufficient for proper growth, it follows that 
 those most useful to plants generally are of intermediate 
 wave-length familiar to our eyes as yellow rays precisely 
 those in wh'.ch the solar radiations are especially rich. 
 While, therefore, we may artificially arrange such supplies 
 of radiant energy as are necessary to plants by properly 
 
 1 B. A. Report, 1881, p. 474. 
 
METEOROLOGY. 
 
 combining dark and luminous rays (by means of heating 
 apparatus, electric light, etc.) it appears that although the 
 sun provides them in rays of all wave-lengths, it does so 
 chiefly in those of a certain restricted period ; and hence 
 some measure of the total amount of energy required to 
 produce a given physiological effect on a particular plant 
 may be obtained from the intensity of these rays, taken as a 
 measure of the whole solar radiations, and the length of 
 time the plant is exposed to them. And since these 
 " yellow " rays are the richest both in light and in heat, we 
 may take either of these their effects as a measure of their 
 intensity, and therefore of the total energy of the solar 
 radiations under normal atmospheric conditions, i.e. so long 
 as there is not special selective absorption of rays of par- 
 ticular wave-lengths. We assume that to whatever con- 
 clusions the physiological botanist may be led with reference 
 to the relative action of heat and light upon vegetation for 
 the particular case of the sun's rays, the temperature, in so far 
 as it depends on the action of these rays, may be taken as at 
 least a rough measure of the total effect, and proceed to in- 
 quire how far the air-temperature, as observed by a pro- 
 tected thermometer, is a measure of that temperature this 
 being for the present the most reliable datum available. 
 
 139. Temperature of Plants. The direct determination of 
 the actual external temperature of a plant as a whole at any 
 given instant is obviously impossible. Every leaf, every 
 part of the stem, every bud and petal, presents a different 
 angle to the sun, and is protected to a different degree by 
 other parts, or by neighbouring plants ; consequently every 
 part of the surface receives different amounts of heat. Again, 
 every part of the structure has a different colour, so that each 
 reflects and absorbs a different proportion of the rays falling 
 upon it ; and every part has a different structure different 
 conductivity for heat, and different supply of moisture to 
 
APPLICATION TO AGRICULTURE. i;i 
 
 modify the temperature by evaporation from the surface. 
 The question is therefore forced into the form How far do 
 the variations of air temperature represent the variations 
 experienced by the plant ? At first sight, it would seem that 
 there should be a closer analogy between the experiences of 
 exposed thermometers and of plants ; but we have already 
 had to deal with the difficulties presented by radiation ther- 
 mometers, and in addition to these there is the necessity of 
 making some assumption that a particular kind of bulb re- 
 sembles the average of the varied surface of vegetation. We 
 have little reason to believe that a polished metal surface, or 
 one of lamp black, is sensible to radiant heat in any manner 
 resembling that of a leaf, even supposing that both were ex- 
 posed in exactly the same way. On the other hand, we 
 know that in most cases only a small proportion of the sur- 
 face of any plant is at any one time exposed to direct sun- 
 light, and that the parts most likely to have their temperature 
 greatly raised by such exposure are just those most abund- 
 antly supplied with moisture, which by evaporation will 
 reduce the temperature : considerations indicating a further 
 approximation to " shade " conditions, at least so far as heat 
 is concerned, although the same does not hold good for 
 light, as will be seen immediately. But, as A. de Candolle 1 
 points out, plants themselves afford evidence which is almost 
 conclusive ; for there is no observable difference between 
 the north and south sides of a tree ; we do not gather 
 flowers or fruit from the south side of a plant first and from 
 the north later. We might therefore expect to find that, 
 other things being equal, the average air temperature in 
 shade would give an approximate measure of the average 
 temperature of vegetation. 
 
 140. Relation of Air Temperature to Insolation. But 
 
 1 Geographic botanique raisonnee, p. 16. 
 
172 METEOROLOGY. 
 
 here a difficulty comes in which it must be admitted greatly 
 reduces the value of the method we are describing. Granting 
 that the temperature of the air in some degree follows or 
 represents that of vegetation as a whole, we cannot pretend 
 that it is wholly dependent on the amount or intensity of 
 insolation. While it is true that the temperature of the air 
 is in the first place determined by the effect of the sun's rays 
 on the surface of the ground, or on the vegetation covering 
 it, that temperature is greatly modified by the motion of the 
 air. When the prevailing winds at any place blow from a 
 warmer region, the temperature of the air is undoubtedly 
 higher than when they come from a colder, although the sun's 
 heat may be the same ; in the one case a certain amount of 
 transferred temperature must be added, and in the other 
 subtracted. Hence in the two cases the observed tempera- 
 ture of air and of vegetation bears a different and a varying 
 relation to that of the sun's rays, and it ceases to be 
 available as a measure of their intensity that is of their light 
 and heat. 
 
 The most important example of this which suggests itself 
 is that of oceanic and insular climates. The sea acts as a 
 kind of heating apparatus during winter and a refrigerator in 
 summer, controlling the temperature independently of the 
 direct insolation; and an ocean current may elevate or depress 
 the temperature during the whole year. During the summer 
 months the temperature is in general relatively lower, latitude 
 for latitude, in an oceanic climate than in a continental, and 
 while this is partly due to the direct decrease of the inten- 
 sity of insolation through greater cloudiness, it is also partly 
 on account of the coolness of the winds blowing in from the 
 sea. The fact that the floras of the two climates are usually 
 affected to such a degree by differences of rainfall that tem- 
 perature considerations scarcely come in, may allow us to 
 evade this difficulty in the more extreme cases, but, at the 
 
APPLICATION TO AGRICULTURE. 173 
 
 same time, makes it almost impossible to estimate its real 
 importance. 
 
 The approximate heating effect on the air of a given 
 amount of solar radiation has been calculated by Zenker, 1 
 who has found marked differences between observed and 
 computed effects due to the transference of heat by currents 
 and winds, but these results have not yet been applied to 
 the questions we are discussing. Attempts to estimate 
 solar intensity by means of air temperature must for the 
 present be limited to regions where that temperature is 
 chiefly regulated by the temperature of the earth's surface 
 that is, to continental climates. 
 
 141. The Plant as a Heat Engine. Under this provision 
 the various processes of germination, flowering, and matura- 
 tion are then to be regarded as forming together a definite 
 piece of work, requiring for its accomplishment the expendi- 
 ture of a certain amount of energy ; and that energy, as sup- 
 plied in the form of heat, to be transformed into what we 
 may call vital energy by the plant, which becomes, from 
 this point of view, a heat engine. Comparing the case of the 
 plant with that of other heat engines, some further points of 
 analogy suggest themselves, which may be taken as guides 
 in defining our ideas. As a concrete example, suppose a 
 steam-engine is employed to pump a given quantity of 
 water out of a tank ; we may burn tons of coals and con- 
 sume thousands of gallons of water in the boiler, but unless 
 we raise a certain minimum pressure of steam, the engine 
 refuses to move ; as soon as that minimum is reached, 
 pumping begins, and the higher the pressure the faster the 
 engine goes. The time taken to empty the tank is inversely 
 as the rate of pumping: if the engine goes slow, it will 
 take a long time if faster, shorter ; but observe that what- 
 ever is done does not require to be done a second time. If 
 1 Met. Zeit., 1892, p. 336. 
 
174 METEOROLOGY. 
 
 the pressure falls below the minimum again, the engine 
 stops, but the work done is not lost, the water does not flow 
 back into the tank. So with the plant; energy must be 
 supplied at a certain pressure or intensity, or (since we have 
 so agreed to measure it) at a certain temperature ; until 
 that minimum temperature is reached no growth takes 
 place, and the more it is exceeded, the faster growth goes 
 on ; and all progress made is so much to the good if the 
 temperature falls below the minimum, growth stops, but 
 there is no going back. 
 
 Now, suppose that (in the case of the steam-engine) the 
 tank were altogether inaccessible, and that we had no means 
 of ascertaining how much water it contained, or how far it 
 had to be pumped, suppose that the only data to be ob- 
 tained were hourly readings of the pressure gauge on the 
 boiler, obviously it would be a hard task to calculate how 
 much water the engine actually had to lift. But suppose 
 that we require to pump the tank out once a year, and that 
 each time all the machinery and gear are the same and in the 
 same state of efficiency, then it would be not so difficult to 
 compare, from the readings of the pressure gauge, the 
 amount of water in the tank one year with the amount in 
 another. We know that all the hours in which the pressure 
 is below a certain point may be struck out, because the 
 machinery must be at rest, and the rate of pumping at the 
 other hours depends on the excess of the pressure above 
 that point ; so by adding the excess for each hour to the 
 sum of all the previous hours since pumping began until the 
 tank is empty, we evidently obtain a result representing the 
 work done. Then if the machinery is the same every year, 
 the variations in the sum will give some measure of the 
 differences in the amount of water. But further, suppose 
 after repeating this calculation a number of times we find 
 that the pressure sum is the same every year, then we may 
 
APPLICATION TO AGRICULTURE. 175 
 
 fairly conclude that the quantity of water is always the same, 
 and in this way arrive at an estimate of the work to be done 
 each year ; knowing how much there is to do the tale to 
 be made up we can each year calculate at any hour how 
 far the work has progressed, and how much still remains to 
 be done. 
 
 So with plants. Adding the excess above the minimum 
 temperature required for growth for each hour from the time 
 germination begins, we set a sum which measures the work 
 done by the plant-engine, and we find year by year that be- 
 tween germination and maturation, whether the season be 
 early or late, the sum remains nearly the same ; justifying 
 our assumption that under favourable circumstances 
 temperature, and air temperature, is a measure of the energy 
 required for plant development, and that different progres- 
 sive stages of that development correspond to definite 
 amounts of work done. Hence at any point of the work 
 we can, by looking up the temperature sum, ascertain how 
 far the plant has progressed towards fruition, and how far it 
 has still to go, and comparing that sum with the average 
 of years for the same date, we get an accurate estimate of 
 the earliness or lateness of the season, and the chances 
 of full ripening before the supply of useful energy 
 fails. 
 
 142. Injurious Effects of Extremes of Temperature. Ob- 
 serve that for plant and steam-engine alike we have 
 assumed the amount of power lost through machinery to be 
 constant the familiar case of dealing with one variable at a 
 time. Such a condition corresponds most closely in the 
 case of the plant to a state of perfect health, with proper 
 supplies of normal food and moisture ; and hence we ob- 
 tain a method of judging of these matters. For if the 
 temperature account is satisfactory, and still a plant or crop 
 seems backward in development, we may look for disease, 
 
METEOROLOGY. 
 
 and failing that, conclude that food supplies have run short ; 
 and so, by top-dressing or the like, save a crop before it is 
 too late. 
 
 Again, we must carefully distinguish between the influ- 
 ence of temperature on the development of a healthy plant 
 and its direct action upon plants themselves. All tempera- 
 tures below a certain point are simply to be left out of con- 
 sideration so far as positive growth is concerned, but this 
 does not affect the fact that such low temperatures may kill 
 the plant altogether. It was long believed that the fatal 
 results of exposure to low temperatures were chiefly caused 
 by destruction of plant tissue through the freezing of the 
 contained water, but we know that while many plants are 
 unable to withstand temperatures far above freezing, others 
 are quite uninjured, although the sap in the tissues is frozen 
 solid, 1 and it is now generally admitted that the action is 
 physiological, arising from interference with the vital functions. 
 A sudden change through a wide range of temperature is 
 often almost as destructive as a severe frost. The following 
 extracts from observations by Mr. E. J. Lowe, F.R.S., 2 com- 
 paring the effects of the great frosts during the winters 
 1860-1 and 1878-9, afford interesting evidence that the 
 destruction is not caused by the direct freezing of the sap. 
 The greatest cold of 1860-1 exceeded that of 1878-9 by 
 10 F. ; the minima were 6 F. in 1860-1, and + 4 F. in 
 1878-9. See table opposite page. 
 
 In the same way, when the temperature rises above a cer- 
 tain point, the rate of growth stops increasing either through 
 destruction of tissue by excessive heat, or through simple 
 exhaustion; just as the statement, that the steam-engine 
 goes faster the greater the pressure, is not true when we 
 
 1 See A. de Candolle, loc. cit. p., 33. 
 
 2 B. A. Report, 1879, p. 377. 
 
APPLICATION TO AGRICULTURE. 
 
 177 
 
 
 1860- 1. 
 
 1878-9. 
 
 Elm. 
 
 Uninjured. 
 
 Three-fourths of the 
 boughs killed back at 
 least two feet. 
 
 Bay (sweet). 
 
 Killed to the ground. 
 
 The ends of the shoots 
 only killed, and all the 
 leaves. 
 
 Fennel. 
 
 Uninjured. 
 
 All killed. 
 
 Roses, standards. 
 
 All killed. 
 
 Many killed, nearly all 
 injured. 
 
 Laurel, Portugal. 
 
 Nearly all killed. 
 
 Uninjured. 
 
 Walnut. 
 
 Boughs killed, and 
 one tree. 
 
 Uninjured except a 
 want of vigour ; leaves 
 only half the usual size. 
 
 reach the point of bursting the boiler or straining all the 
 rods and connections. 
 
 143. Temperature at which Active Growth Begins. The 
 first practical difficulty encountered in applying the con- 
 siderations just explained to the growth of plants lies in 
 determining the temperature at which active vegetation be- 
 gins. The researches of Boussingault, to whom the method 
 is originally due, and of A. de Candolle, who first made 
 systematic application of it on a large scale, showed 
 that in the case of most plants growing in North- Western 
 Europe a temperature of 5 to 8 C. (41 to 46 F.) must be 
 reached before signs of active growth are manifested ; and 
 although it cannot be said that the exact starting-point is 
 accurately known for different species., subsequent investiga- 
 tions agree pretty unanimously in fixing it at 5'5 or 6 C. 
 The degree generally adopted in this country is 42 F. We 
 
 M 
 
1 78 METEOROLOGY. 
 
 therefore assume, provisionally, that whenever the tempera- 
 ture rises above 42, vegetation generally makes progress, and 
 whenever it falls below that point, progress stops. It must 
 certainly be admitted that in making this assumption we leave 
 out of consideration what is, undoubtedly, a most important 
 factor in practice. The warmth which suffices to make many 
 plants germinate, and even flower, may be quite inadequate to 
 make them produce fruit or to ripen that fruit. Buchan found 
 that in Scotland wheat only ripens where the mean tempera- 
 ture during the summer months is as high as 56 'o F., 1 and 
 that the chief difference between wheat and oat crops was 
 not so much that the latter required a smaller total of heat 
 to ripen it as that the former required greater concentration 
 between the times of flowering and ripening. In warmer 
 climates the excess above not only 42 but higher tempera- 
 tures is in summer so continuous that this point might easily 
 pass unnoticed, except for the inconsistencies introduced by 
 it in the final sums of temperature ; but when we are deal- 
 ing with a plant or crop anywhere near the limit of the area 
 within which it can be cultivated, it becomes obvious that 
 different stages of development require a different 
 minimum before progress in them begins. All tempera- 
 tures above 42 may supply sufficient " power " to 
 enable the plant to form leaves and flowers, but that done 
 the machinery may come to a stop unless a minimum of 50 
 or 56 is reached. Concerning this part of the subject little 
 is as yet accurately known, and we must content ourselves for 
 the present with the quite useful results derived from the 
 assumption of the 42 minimum. 
 
 144. Accumulated Temperature. The next difficulty is to 
 
 find a satisfactory method of expressing the accumulated 
 
 effect of a certain temperature, acting for a given time. 
 
 Perfect accuracy would require a knowledge of the tempera- 
 
 1 Jour. Scot. Met. Soc., 1862, No. 2. 
 
APPLICATION TO AGRICULTURE. \jg 
 
 ture at every instant during the whole period ; but, as we 
 have seen in a former chapter, it is necessary to employ an 
 indirect method which shall give, for any period, a close 
 approximation to the mean. Hourly observations would 
 give results very nearly correct, and we might express the 
 accumulated temperature above 42 in hour degrees ; sub- 
 tracting 42 from the reading at each hour and adding all 
 the positive differences together for the effective temperature, 
 and the negative for the non-effective. But for most pur- 
 poses it is sufficient, and it is always much more easy, to 
 take the day as the unit subtracting 42 from the mean of 
 the day (calculated by methods already described) and 
 adding the differences, which give the /required result ex- 
 pressed in dfoy-degrees. The only serious error introduced 
 by adopting so long an interval of time as twenty-four hours 
 as a unit occurs where during one part of the day the tem- 
 perature is not on the same side of 42 as the mean. This 
 error, which affects all the work of the earlier investigators, 
 is got rid of by the method of General Strachey already 
 described ( 94). 
 
 One other assumption involved in this manner of compar- 
 ing temperature with effect upon growth must be recognised. 
 We have taken for granted that the relation is of the most 
 simple character ; that the progress of the plant is simply 
 proportional to the effective temperature and to the time. 
 In this we are justified only by experience. Several investi- 
 gators, notably Quetelet and Babinet, have thought to obtain 
 better results by supposing the plant development to pro- 
 gress according to the square of the time or the effective 
 temperature, but in this they have not been successful. 
 
 145. Results obtained from Method of Accumulated Tem- 
 perature. Having so far examined the ground, we may 
 summarise our results in the proposition, that in order to 
 the development and maturation of a healthy plant or crop, 
 
1 80 METEOROLOGY. 
 
 besides proper supplies of food and moisture a certain 
 amount of heat and light is necessary, which may be in 
 continental climates measured in terms of accumulated 
 temperature expressed to a first approximation as day 
 degrees above a minimum of 42 F. That this rule, con- 
 fessedly tentative and imperfect as it is, fairly represents the 
 facts, appears from the following table extracted from a paper 
 by Prof. J. H. Gilbert, F.R.S. 1 
 
 Accumulated temperature in day degrees above 42 F., to 
 date of harvest at Rothamsted Experimental Agricultural 
 Station, calculated on the averages of observations in Eastern 
 and Central England. 
 
 Year. 
 
 1878 
 1879 
 1880 
 
 1881 
 
 1882 
 1883 
 
 1884 
 1885 
 
 Highest 2026 1976 
 
 Lowest 1772 1645 
 
 Mean 1900 1836 
 
 Calculating from Greenwich observations by the less ac- 
 curate method of multiplying the monthly means of tempera- 
 ture by the number of days from ist April to date of harvest, 
 Gilbert finds the average for the 27 years 1852 to 1878 to 
 
 1 Archives des Sciences 2ihysiqiies etnaturelles, Geneva, 1886, p. 421. 
 
 From date 
 
 From return 
 
 of frequent 
 Weekly Ex- 
 
 of Weekly 
 Excess after 
 
 cess. 
 Degree?. 
 
 interruption. 
 Degrees. 
 
 
 ICKO 
 
 IOQO 
 
 1061 
 
 IQ76 
 
 IQ76 
 
 1814 
 
 1804 
 
 / 'o' > 6 . . 
 
 1966 
 
 1772 
 
 164; 
 
 1847 
 
 ... 1686 
 
 1774 
 
 l6Q6 
 
APPLICATION TO AGRICULTURE. I Si 
 
 amount to 1980, the highest being 2183 and the lowest 
 1809. These figures are necessarily too large, but making 
 allowance for error, are in fair agreement with those in the 
 table j and we may take it that for wheat the accumulated 
 temperature required is about 1900 F. Herve-Mangon, 
 employing daily means, obtained a result corresponding to 
 1854 for north-eastern France from eight years observa- 
 tions. M. E. Risler gives 2192 for Nyon, in Switzerland. 
 For spring-sown barley, Boussingault gives 2005, De Can- 
 dolle about 2100, and Herve-Mangon, 1989; from which, 
 as Gilbert further points out, it would seem that the totals 
 required for wheat and for at least the finer kinds of barley 
 are nearly the same. 
 
 The figures quoted have been derived almost exclusively 
 from stations at which the climates are for the most part 
 continental in character. The data available for discussion 
 by this method are as yet too meagre to merit extensive 
 treatment, and it is impossible to trace with any accuracy 
 the relation of the solar energy received by plants to the 
 observed temperature in shade in different climates. A. de 
 Candolle found that in latitude 56 N., Alyssum required 
 50 to 70 more in Scotland than in Sweden, and Spindle- 
 tree about 200 more ; Columbine 900 more in Scotland 
 than in Eastern Russia. In latitude 57 30' to 58 the 
 Alder buckthorn required 350 less at Vialka than in 
 Scotland, and in latitude 59 the flax-seed (Radiola binoides] 
 45 less in Sweden than in Orkney. 1 All these facts may 
 be taken as showing that the accumulated temperature is 
 not the same measure of solar energy in the different 
 localities, but that Scotland is as it were artificially warmed 
 by the sea. 
 
 146. Conditions under which the Method is Applicable. 
 In view of all the evidence, we may fairly conclude that in 
 1 Geof/raj)hie botanique raisonnce, p. 402. 
 
,182 METEOROLOGY. 
 
 continental climates, where the temperature of the lower 
 strata of the atmosphere is chiefly controlled by terrestrial 
 radiation, air temperature fairly represents the intensity of the 
 sun's action ; but that in oceanic climates the balance of 
 light and heat is disturbed, and a certain proportion of the 
 latter must, when positive, be treated separately like the 
 warmth of a hot-house. This extra allowance of " dark 
 heat " is, of course, beneficial just as the warming of the hot- 
 house is, as is obvious from the fact that spring flowers 
 bloom about a month earlier in the west and south of 
 England and Ireland than in the midland counties and 
 Scotland, 1 but its influence is probably different from that 
 of sun heat which would alone produce the same average 
 temperature. When the effect is negative, as it generally is 
 during the warmer months of the year, the deficiency re- 
 duces the apparent power of the sun's rays as shown by air 
 temperature, and the relation between heat and light is 
 altered in the opposite direction. 
 
 147. Influence of Rainfall. Important as these considera- 
 tions undoubtedly are in any scientific treatment of the 
 relations of climate to plant-growth, it is probable that they 
 do not very seriously affect the value of the method of ac- 
 cumulated temperatures in its application to practical 
 agriculture, because in those climates where the air tempera- 
 ture is greatly modified by the influence of the sea, the ex- 
 cessive humidity and rainfall usually preclude the cultivation 
 of any but the coarser kinds of cereals. Buchan finds that 
 "where the mean summer temperature is as low as 56'o, 
 the cultivation of wheat is quite possible even though the 
 character of the springs be cold and backward, provided 
 always that the rainfall in summer and autumn be not in 
 excess," and again that temperature conditions on the west 
 
 1 See Hoffmann : Pflanzen-fhanoloyischen Beobachtungen in Europa. 
 Giessen, 1885. 
 
APPLICATION TO AGRICULTURE. 183 
 
 coast of Scotland are in many places more favourable than 
 on the east, but " on account of the moist atmosphere the 
 crops run rapidly into straw, which being weak from a de- 
 ficiency of silica are more easily laid by the torrents of rain 
 and violent winds." 1 Hence it appears that the difficulty 
 lies not in supplying power to the engine but in keeping it 
 in working order. 
 
 So long as the required total of sun heat and light can be 
 made up the more uniform a climate is the better, for vege- 
 tation is specially liable to injury from sudden large changes 
 of temperature during the period of active growth. In the 
 south of England, where the mean summer temperature is 
 4 p o K above the critical point required for wheat, the damage 
 and loss in a bad year is usually more complete than in the 
 more equable climate of the wheat-growing districts of 
 Scotland, where the usual margin is only 2 F., and even that 
 is eked out by the greater length of the days. 2 
 
 148. Effect of Vegetation on Climate: Forests. The 
 modifications of climate produced by the covering of vegeta- 
 tion on the surface of the ground have already been noticed, 
 and their nature will be sufficiently intelligible from the fore- 
 going details of the reverse problem. As an extreme case we 
 may take the differences observed between the climates of 
 open country and of forest-land under similar conditions. 
 Other things being equal, the range of temperature is smaller in 
 the former than in the latter. The daily minima of tempera- 
 ture are higher, and the daily maxima lower in the forest 
 than in the open, much as we find on comparing sea with 
 land. During still summer weather a movement of the air 
 exactly analogous to land and sea breezes is sometimes 
 observed on the outskirts of a forest, most frequently in the 
 evening, for the lowering of the maxima is in general much 
 
 l Jour. Scot. Met. Soc. t 1862. 
 
 2 Jour. Scot. Met. Soc. t 1863., *> P- I S 
 
1 84 METEOROLOGY. 
 
 more marked than the raising of the minima. The amount 
 by which the range is decreased depends greatly on the 
 thickness of the growth, and on the species of tree constitu- 
 ting the forest. It is in the case of evergreen trees least in 
 midwinter, increases steadily till August and then again 
 decreases. In the case of deciduous trees the difference is 
 least in April just before the leaves come out, increases 
 rapidly for two months, attains its maximum in July, and 
 after September diminishes rapidly on the falling of the 
 leaves. During summer, at least, the daily maximum is 
 somewhat later and the daily minimum somewhat earlier in 
 forest than in field. The observations of the German 
 Forestry department, 1 upon which these results are based, 
 afford the further important fact that the variations of tem- 
 perature among the upper branches of forest trees are inter- 
 mediate between those near the ground and in the open, but 
 tend to follow the former more closely. 
 
 According to Eckert, 2 the amount of vapour (absolute 
 humidity) decreases from the ground upwards amongst trees 
 if the soil be wet, but increases after a continuance of dry 
 weather, the supply of moisture to the leaves by transpira- 
 tion holding out longer than that to the surface of the soil. 
 Compared with open country the absolute humidity is 
 usually greater not only amongst the trees but above them ; 
 and the same is true of the relative humidity : but in these 
 cases it is impossible to lay down a general rule, as the condi- 
 tions are chiefly determined by the distribution of rainfall and 
 the nature of the soil and vegetation in the open. 
 
 That forests have, in certain cases, a marked influence 
 upon rainfall has been well established by Blanford in India 
 and by Miittrich and Ebermayer in Germany. By com- 
 paring a region in the Central Provinces, planted in 1875, 
 
 } See Muttrich, Met. Zeit., 1891, p. 55. 
 *Met. Zei't., 1890, p. 361. 
 
APPLICATION TO AGRICULTURE. 185 
 
 with surrounding districts, Blanford found that the rainfall 
 steadily increased year by year with the growth of the young 
 trees. 1 A similar method, although applied on a smaller 
 scale, gave identical results in the heart of the Luneburger 
 Haide. 2 
 
 . Asiatic Soc. offienyal, i., 1887. 
 sMiittrich, Met. Zeit, 1892, p. 307. 
 
INDEX. 
 
 PAGE 
 
 ABSORPTION, Atmospheric . . . 3 
 
 Air, Constitution of .... 23 
 
 Pressure .... 23 
 
 Specific Heat of ... 25 
 
 Temperature, Observations of - 104 
 
 Anemometers . .122 
 
 Anti-Cyclones ... 57 
 
 Intensity of . -73 
 
 ,, Moving . . ' . . -75 
 
 ,, Permanent ... .74 
 
 ,, Relation to Cyclones . .75 
 
 ,, Temperature Gradient in . 79 
 
 ,, Weather in . . 76 
 
 Atlantic Anti-Cyclone .... .74 
 
 ,, Ocean, Winds and Currents of . . . 140 
 
 Atmosphere, Height of Homogeneous 53 
 
 Vertical Circulation in . . . 52 
 
 BAROMETER . . . . . .24 
 
 Errors of . . . . .113 
 
 ,, Non-mercurial ..... 120 
 
 Reduction to Sea Level . . ~ . . . 118 
 
 Boen ........ 8^. 
 
 186 
 
Borrowing Days 
 Boyle's Law . 
 Buys Ballot's Law 
 
 INDEX. 
 
 187 
 
 PAGE 
 
 *54 
 24 
 
 5 
 
 CAPACITY Correction . 
 Capillarity ,, . 
 
 Charles' Law .. 
 Charts, Synoptic . 
 
 Climates, Oceanic . 
 Cloud 
 
 ,, Amount .. 
 
 Clouds, Classification of 
 
 ,, Heights of . 
 
 Condensation Nuclei . 
 
 Continental Axis . 
 
 ,, Climate . 
 
 Currents, Ascending . 
 
 ,, Descending . 
 
 Cyclone, Clouds in 
 
 ,, Condensation in 
 ,, Form of 
 ,, Front . 
 
 Intensity of . 
 ,, Motion of Centre 
 ,, Origin of 
 
 Rainfall in . 
 Rear . 
 
 ,, Temperature in 
 Tracks of . 
 Trough 
 
 ,, Weather in . 
 
 Wind in . 
 
 Cyclonic Areas . 
 
 IJ 3 
 
 2 4 
 
 I2 
 
 138 
 
 I 3 I 
 
 128 
 
 '3 2 
 
 28 
 
 75 
 143 
 
 54 
 
 57 
 58 
 
 5^ 
 60 
 
 49 
 21, 48 
 
 I7> 7 1 
 
 5^ 
 66 
 
 49 
 66 
 68 
 
 49 
 18 
 
 17 
 72 
 
1 88 INDEX. 
 
 D 
 
 PAGE 
 
 DERECHOS ....... 87 
 
 Dew 35 
 
 Dewpoint ....... 27 
 
 Disturbances ....... 40 
 
 Dust Storms . . . . . . .91 
 
 B 
 
 EQUAL Areas, Principle of . . . .52 
 
 Equinoctial Gales ...... 5 
 
 Evaporation . . . . . . 135 
 
 F 
 
 FERREL'S Law .... ... 42 
 
 Fog . .... 29 
 
 FohnWind ....... 81 
 
 Forecasting for Isolated Observers .... 98 
 
 ,, Methods of ... . . 94 
 
 ,, by Synoptic Charts .... 95 
 
 Forests ....... 183 
 
 Fortin's Barometer . . . . . .114 
 
 Friction ....... 42 
 
 G 
 
 GRADIENTS ....... 44 
 
 ,, Influence of Circulation on ... 52 
 
 ,, and Winds . . . . . 45, 62 
 
 Gravity . . . .42 
 
INDEX. 189 
 
 PAGE 
 
 HEAT, Latent ...... 28 
 
 Heights, Measurement of 54 
 
 Helm Wind ..... .148 
 
 Humidity, Absolute ... . . 127 
 
 ,, Relative ...... 128 
 
 Hygrometry . . . . . .126 
 
 ICELAND Depression ...... 75 
 
 Insolation, Relation of Temperature to . .171 
 
 Instruments, Function of .... 5 
 
 Isobars . . 44 
 
 ,, in Cyclones . 4 6 
 
 ,, Straight Line ...... 82 
 
 LAND and Sea Breezes . . J 44 
 
 Line Squalls . . . . . 8 5 
 
 M 
 
 METEOROLOGY, Definition of , . . I 
 
 }> Relation to Other Sciences . . .15^ 
 Meteors ..... 
 
 Mist . 2 9 
 
 Moisture . 
 
 Monsoons . . r 44 
 
 Moon, Influence of ...... 6 
 
 Motion Absolute * . ... . ,40 
 
INDEX. 
 
 I'AGE 
 
 Motion, Relative ...... 42 
 
 Mountains, Effect on Climate of . . . .144 
 
 ,, as Obstacles . . . . .146 
 
 Effect on Rainfall of . . . .147 
 
 N 
 
 NOTATION, Beaufort . . . . . 13 
 
 OBSERVATION, Errors of ..... 4 
 
 Observations, Instrumental ..... 103 
 
 ,, must be Comparable .... 103 
 
 PAN Serboni , . . . . .154 
 
 Plant as Heat Engine . . . . . 173 
 
 Plants. Effect of Climate on . . . . .168 
 
 ,, Temperature of . . . . .170 
 
 Precipitation . . . . . . .132 
 
 Pressure, Calculation of Mean . . . . .118 
 
 ,, Diurnal Variation of .... 38 
 
 ,, at High Levels . . 65 
 
 Prognostics, Evidence contained in . . . 2 
 
 R 
 
 RADIATION ....... 29 
 
 Terrestrial . . . . . . 35, 112 
 
 Rain 8 
 
Rainfall, Distribution of 
 Rain Gauge 
 
 Range, Corrections for . 
 Reasoning, Inaccurate . 
 
 k 
 INDEX. 
 
 PAGE 
 154 
 133 
 
 39 
 5 
 
 SAINT SWITHIN 
 
 Seasonal Variations 
 
 Secondaries 
 
 Senses, Deception by . 
 
 Snow Covering, Effect of 
 
 Soils, Formation of 
 
 Soil, Moisture in 
 
 ,, Relation to Climate of 
 
 ,, Temperature of 
 Squalls . . 
 
 ,, Prediction of 
 Storms 
 
 Storm Warnings 
 Sunshine Recorders 
 Siphon Barometer 
 
 155 
 
 39 
 
 85 
 
 4 
 
 151 
 
 159 
 
 165 
 
 158 
 
 162 
 
 87 
 
 93 
 
 70 
 
 97 
 
 112 
 114 
 
 TEMPERATURE 
 
 Accumulated 
 Calculation of Mean 
 Diurnal Variation of 
 Effects of Extremes of 
 at which growth begins 
 Interruptions of 
 Radiation 
 Range of 
 
 8 
 
 108, 178 
 
 108 
 
 36 
 
 175 
 
 177 
 
 153 
 ill 
 149 
 
192 INDEX. 
 
 PAGE 
 
 Temperature, Variability of . . . . .151 
 
 Thermometers . . . . . . .105 
 
 Thermometer Exposure . . . . .106 
 
 Tornadoes ....... 89 
 
 Trade Winds ....... 74 
 
 Transmission of Rays . . . . . -33 
 
 V 
 
 V-SHAPED Depressions ..... 85 
 
 Vapour, Aqueous ...... 26 
 
 ,, Supersaturated , . . .28 
 
 Vegetation, Effect on Climate of . . . .183 
 
 ,, Influence of Rainfall on .... 182 
 
 ,, Rays Useful to .... 169 
 
 Vernier . . . . . . .116 
 
 w 
 
 WATERSPOUTS . . . . . . .-91 
 
 Weather, Analysis of . . . . . 7 
 
 ,, Definition of . . . . . . i 
 
 ,, Successive Changes of .... 8 
 
 ,, "Radiation" ..... 76 
 
 Types of 69 
 
 ,, Types, Recurrence of .... 153 
 
 Wedges .... ... 83 
 
 Wind .... ... 7, 23 
 
 ,, Direction, Observation of . . . .121 
 
 ,, Diurnal Variation of . . . .125 
 
 ,, Force, Estimation of . . . .124 
 
 Winds, Inclination at High Levels of , , , .67 
 
A LIST OF NEW BOOKS 
 
 AND ANNOUNCEMENTS OF 
 
 METHUEN AND COMPANY 
 
 PUBLISHERS : LONDON 
 
 1 8 BURY STREET 
 
 W.C. 
 
 CONTENTS 
 
 FORTHCOMING BOOKS, . * 
 
 POETRY, . ..... 
 
 HISTORY AND BIOGRAPHY, .... 
 
 GENERAL LITERATURE, .... 
 
 WORKS BY S. BARING GOULD, . . . . 
 
 FICTION, ...... 17 
 
 NOVEL SERIES, .... . l8 
 
 BOOKS FOR BOYS AND GIRLS, . . , 2O 
 
 LEADERS OF RELIGION, .... 22 
 
 UNIVERSITY EXTENSION SERIES, ... 22 
 
 SOCIAL QUESTIONS OF TO-DAY, aj 
 
 OCTOBER 1893 
 
OCTOBER 1893. 
 
 MESSRS. METHUEN'S 
 
 ANNOUNCEMENTS 
 
 Gladstone. THE SPEECHES AND PUBLIC ADDRESSES 
 OF THE RT. HON. W. E. GLADSTONE, M.P. With Notes. 
 Edited by A. W. HUTTON, M.A. (Librarian of the Gladstone 
 Library), and H. J. COHEN, M.A. With Portraits. 8vo. Vol. IX. 
 12s. 6d. 
 
 Messrs. METHUEN beg to announce that they are about to issue, in ten volumes 
 8vo, an authorised collection of Mr. Gladstone's Speeches, the work being under- 
 taken with his sanction and under his superintendence. Notes and Introductions 
 will be added. 
 
 In view of the interest in the Home Rule Question, it is proposed to issue Vols. IX, 
 and X.) "which will include the speeches of the last seven or eight years, im- 
 mediately, and then to proceed "with the earlier volumes. Volume X. is already 
 published. 
 
 Henley & Whibley. A BOOK OF ENGLISH PROSE. 
 Collected by W. E. HENLEY and CHARLES WHIBLEY. Crown 
 Svo. 
 
 Also small limited editions on Dutch and Japanese paper. 2is. 
 and 42^. net. 
 
 A companion book to Mr. Henley's well-known Lyra Heroica. It is believed that 
 no such collection of splendid prose has ever been brought within the compass of 
 one volume. Each piece, whether containing a character-sketch or incident, is 
 complete in itself. The book will be finely printed and bound. 
 
 Henley. ENGLISH LYRICS. Selected and Edited by 
 W. E. HENLEY. In Two Editions : 
 
 A limited issue on hand-made paper. Large crown 8vo. 
 A small issue on finest large Japanese paper. Demy Svo. 
 The announcement of this important collection of English Lyrics will excite wide 
 interest. It will be finely printed by Messrs. Constable & Co., and issued at 
 first in limited editions. 
 
 Dixon. ENGLISH POETRY FROM BLAKE TO BROWN- 
 ING. By W. M. DIXON, M.A. Crown 8vo. $s. 
 A Popular Account of the Poetry of the Century. 
 
 Prior. CAMBRIDGE SERMONS. Edited by C. H. PRIOR, 
 
 M.A., Fellow and Tutor of Pembroke College. Crown Svo. 6s. 
 A volume of sermons preached before the University of Cambridge by various 
 preachers, including the Archbishop of Canterbury and Bishop Westcott. 
 
MESSRS. METHUEN'S LIST 3 
 
 Oscar Browning. GUELPHS AND GHIBELLINES: A Short 
 
 History of Mediaeval Italy, A.D. 1250-1409. By OSCAR BROWNING, 
 Fellow and Tutor of King's College, Cambridge. Crown Svo. $s. 
 O'Grady. THE STORY OF IRELAND. By STANDISH 
 O'GRADY, Author of 'Finn and His Companions.' Small crown 
 Svo. 
 A short sketch of Irish History, simply and picturesquely told, for young people. 
 
 Scott. THE MAGIC HOUSE AND OTHER VERSES. By 
 DUNCAN C. SCOTT. Extra Post Svo, bound in buckram. 5^. 
 
 Lock. THE LIFE OF JOHN KEBLE. By WALTER LOCK, 
 M.A. With Portrait from a painting by GEORGE RICHMOND, R.A. 
 Crown Svo., buckram^ 5*. Fifth Edition just ready. 
 
 1 A fine portrait ot one of the most saintly characters of our age, and a valuable con- 
 tribution to the history of that Oxford Movement.' Times. 
 
 Classical Translations 
 
 Irwin. LUCIAN Six Dialogues (Nigrinus, Icaro-Menippus, 
 Cock, Ship, Parasite, Law of Falsehood). Translated into English 
 by S. T. IRWIN, M.A., Assistant Master at Clifton; late Scholar of 
 Lincoln College, Oxford. Crown Svo. 
 
 Morshead. SOPHOCLES Electra and Ajax. Translated into 
 English by E. D. A. MORSHEAD, M. A., late Scholar of New College, 
 Oxford ; Assistant Master at Winchester. Crown Svo. 
 Two new volumes of the ' Classical Translations ' series. 
 
 Fiction 
 
 Corelli. BARABBAS : A DREAM OF THE WORLD'S 
 TRAGEDY. By MARIE CORELLI, Author of ' A Romance of Two 
 Worlds,' * Vendetta,' etc. 3 vols. Crown Svo. 31*. 6d. 
 
 Baring Gould. CHEAP JACK ZITA. By S. BARING GOULD, 
 
 Author of ' Mehalah,' ' In the Roar of the Sea,' etc. 3 vols., Crown 
 Svo. 31*. 6d. 
 A Romance of the Ely Fen District in 1815. 
 
 Fenn. THE STAR GAZERS. By G. MANVILLE FENN, 
 Author ol ' Eli's Children,' etc. 3 vols. Crown Svo. 31 s. 6d. 
 
 Esm Stuart. A WOMAN OF FORTY. By ESM STUART, 
 Author of 'Muriel's Marriage,' 'Virginia's Husband,' etc. 2 vols. 
 Crown Svo. 21 s. 
 
4 MESSRS. METHUEN'S LIST 
 
 Parker. THE TRANSLATION OF A SAVAGE. By 
 GILBERT PARKER, Author of 'Pierre and His People,' 'Mrs, 
 Falchion,' etc. Crown Svo. 55. 
 
 A picturesque story with a pathetic and original motive, by an author whose rise in 
 the estimation of the critics and the public has been rapid. 
 
 Gilchrist. THE STONE DRAGON. By MURRAY GILCHRIST. 
 
 Crown Svo. Buckram, 6s, 
 A volume of stories of power so weird and original as to ensure them a ready welcome. 
 
 Benson, DODO : A DETAIL OF THE DAY. By E. F. 
 
 BENSON. Crown Svo. Seventh Edition. 2 vols. 2is. 
 
 A story of society by a new writer, full of interest and power, which has already 
 passed through six editions, and has attracted by its brilliance universal atten- 
 
 feast of epigram and paradox \ the ' Athenaeum ' spoke of the author as a writer 
 of quite exceptional ability ; the 'Academy' praised his amazing cleverness \ the 
 'World ' said the book was brilliantly -written ; and half-a-dozen papers declared 
 there was not a dull page in the two volumes. 
 
 FOR BOYS AND GIRLS 
 
 Baring Gould. THE ICELANDER'S SWORD. By S. 
 BARING GOULD, Author of 'Mehalah,' etc. With twenty-nine 
 Illustrations by J. MOYR SMITH. Crown Svo. 6s. 
 A stirring story of Iceland, written for boys by the author of ' In the Roar of the Sea. ' 
 
 Outhell. TWO LITTLE CHILDREN AND CHING. By 
 
 EDITH E. CUTHELL. Profusely Illustrated. Crown Svo. Cloth, 
 gilt edges, 6s. 
 
 Another story, with a dog hero, by the author of the very popular ' Only a Guard- 
 Room Dog.' 
 
 Blake. TODDLEBEN'S HERO. By M. M. BLAKE, Author of 
 'The Siege of Norwich Castle.' With 36 Illustrations. Crown 
 Svo. $j. 
 A story of military life for children. 
 
 NEW AND CHEAPER EDITIONS 
 
 Baring Gould. MRS. CURGENVEN OF CURGENVEN. 
 ByS. BARING GOULD, Author of 'Mehalah,' 'Old Country Life,' 
 etc. Crown Svo. Third Edition. 6s. 
 
 hich 
 it as a 
 
 ly JNews says 
 
 that the swing of the narrative is splendid ', and the 'Speaker' mentions its 
 bright imaginative power. 
 
MESSRS. METHUEN'S LIST 5 
 
 Parker. MRS. FALCHION. By GILBERT PARKER, Author of 
 
 ' Pierre and His People.' New Edition in one volume. 6s. 
 Mr. Parker's second book has received a warm welcome. The ' Athenaeum ' called 
 it a splendid study of character ; the ' Pall Mall Gazette ' spoke of the writing as 
 but little behind anything that has been done by any writer of our time ; the 
 'St. James' 1 called it a. very striking and admirable novel \ and the 'West- 
 minster Gazette ' applied to it the epithet of distinguished. 
 
 Norris. HIS GRACE. By W. E. NORRIS, Author of 
 'Mademoiselle de Mersac,' 'The Rogue,' etc. Third and Cheaper 
 Edition. Crown 8vo. 6s. 
 
 An edition in one volume of a novel which in its two volume form quickly ran through 
 two editions. 
 
 Pearce. JACO TRELOAR. By J. H. PEARCE, Author of 
 
 'Esther Pentreath.' New Edition. Crown 8vo. $s.6d. 
 A tragic story of Cornish life by a writer of remarkable power, whose first novel has 
 
 been highly praised by Mr. Gladstone. 
 The ' Spectator' speaks of Mr. Pearce as a writer of exceptional power \ the ' Daily 
 
 Telegraph ' calls it powerful and picturesque \ the ' Birmingham Post ' asserts 
 
 that it is a novel of high quality. 
 
 Pryce. TIME AND THE WOMAN. By RICHARD PRYCE, 
 
 Author of ' Miss Maxwell's Affections,' ' The Quiet Mrs. Fleming,' 
 etc. New and Cheaper Edition. Crown Svo. 6s. 
 
 1 Mr. Pryce's work recalls the style of Octave Feuillet, by its clearness, conciseness, 
 its literary reserve.' Atheneeum. 
 
 1 It is impossible to read the book without interest and admiration. 1 Scotsman. 
 
 ' He has, in fact, written a book of some distinction, and the more his readers have 
 thought and observed for themselves the more are they likely to appreciate it.' 
 Pall Mall Gazette. 
 
 ' Quite peculiar fascination is exercised by this novel. The story is told with un- 
 usual cleverness. ' Time and the Woman ' has genuine literary distinction, and 
 the rarity of this quality in the ordinary novel needs no expression.' Vanity Fair. 
 
 Dickenson. A VICAR'S WIFE. By EVELYN DICKENSON. 
 Cheap Edition. Crown Svo. 3*. 6d. 
 
 Prowse. THE POISON OF ASPS. By R. ORTON PROWSE. 
 Cheap Edition. Crown Svo. $s. 6d. 
 
 UNIVERSITY EXTENSION SERIES 
 
 NEW VOL UMES. Crown Svo. 
 
 A MANUAL OF ELECTRICAL SCIENCE. By GEORGE 
 
 J. BURCH, M.A. With numerous Illustrations. 3^. 
 A practical, popular, and full handbook. 
 
6 MESSRS. METHUEN'S LIST 
 
 THE CHEMISTRY OF FIRE. By M. M. PATTISON Mum, 
 
 M.A. Illustrated. 2s. 6d. 
 An exposition of the Elementary Principles of Chemistry. 
 
 A TEXT-BOOK OF AGRICULTURAL BOTANY. ByM.C. 
 POTTER, M.A., F.L.S. Illustrated. 3*. >d. 
 
 THE VAULT OF HEAVEN. A Popular Introduction to 
 Astronomy. By R. A. GREGORY. With numerous Illustrations. 
 Crown 8vo. 2s. 6d. 
 
 METEOROLOGY. The Elements of Weather and Climate. 
 By H. N. DICKSON, F.R.S.E., F.R. Met. Soc. Illustrated. 2s. 6d. 
 
 SOCIAL QUESTIONS OF TO-DAY 
 
 NEW VOLUMES. 
 
 Crown 8vo, 2s. 6d. 
 
 WOMEN'S WORK. By LADY DILKE, Miss BULLEY, and 
 
 Miss ABRAHAM. 
 
 TRUSTS, POOLS AND CORNERS. As affecting Commerce 
 and Industry. By J. STEPHEN JEANS, M.R.I., F.S.S. 
 
 (Etwcatfonal 3Boofc0 
 
 Davis. TACITI GERMANIA. Edited with Notes and In- 
 troduction. By R. F. DAVIS, M.A., Editor of the ' Agricola. 
 Small crown 8vo. 
 
 Stedman. GREEK TESTAMENT SELECTIONS. Edited by 
 A. M. M. STEDMAN, M. A. Third and Revised Edition. Fcap. Svo. 
 2s. 6d. 
 
 Stedman. A SHORTER GREEK PRIMER OF ACCI- 
 DENCE AND SYNTAX. By A. M. M. STEDMAN, M.A. 
 Crown 8vo. 
 
 Stedman. STEPS TO FRENCH. By A. M. M. STEDMAN, 
 
 M.A. i8mo. 
 An attempt to supply a very easy and very short book of French Lessons. 
 
 Stedman. THE HELVETIAN WAR. Edited with Notes 
 and Vocabulary by A. M. M. STEDMAN, M.A. iSnto. is. 
 
MESSRS. METHUEN'S LIST 7 
 
 Methuen's Commercial Series 
 
 Crown 8vo. Cloth. 
 
 Gibbins. BRITISH COMMERCE AND COLONIES FROM 
 ELIZABETH TO VICTORIA. By H. DE B. GIBBINS, M.A,, 
 Author of 'The Industrial History of England,' etc., etc. 2s. 
 
 Bally. A MANUAL OF FRENCH COMMERCIAL COR- 
 RESPONDENCE. By S. E. BALLY, Modern Language Master 
 at the Manchester Grammar School. 
 
 Lyde. COMMERCIAL GEOGRAPHY, with special reference 
 to Trade Routes, New Markets, and Manufacturing Districts. By 
 L. D. LYDE, M.A., of The Academy, Glasgow. 2s. 
 
 Simplified Classics 
 
 A series of Classical Readers, Edited for Lower Forms with Introduc- 
 tions, Notes, Maps, and Illustrations. 
 
 Herodotus. THE PERSIAN WARS. Edited by A. G. LIDDELL, 
 M.A., Assistant Master at Nottingham High School. 
 
 Plautus. THE CAPTIVI. Edited by J. H.-FREESE, M.A., 
 late Fellow of St. John's College, Cambridge. 
 
 Livy. THE KINGS OF ROME. Edited by A. M. M. STED- 
 
 MAN, M.A. 
 
 Methuen's Novel Series 
 
 3/6 
 
 A Series of copyright Novels, by well-known Authors, 
 bound in red buckram, at the price of three shillings and 
 sixpence. The first volumes will be : 
 
 1. JACQUETTA. By S. BARING GOULD, Author of * Mehalah,' 
 
 etc. 
 
 2. ARMINELL: A Social Romance. By S. BARING GOULD, 
 
 Author of ' Mehalah,' etc. 
 
 3. MARGERY OF QUETHER. By S. BARING GOULD. 
 
 4. URITH. By S. BARING GOULD. 
 
 5. IN THE ROAR OF THE SEA. By S. BARING GOULD. 
 
8 MESSRS. METHUEN'S LIST 
 
 6. DERRICK VAUGHAN, NOVELIST. With Portrait of 
 
 Author. By EDNA LYALL, Author of * Donovan,' etc. 
 
 7. JACK'S FATHER. By W. E. NORRIS. 
 
 8. MY DANISH SWEETHEART. By W. CLARK RUSSELL. 
 
 HALF-CROWN NOVELS. 
 
 A Series of Novels by popular Authors, tastefully 
 bound in cloth. 
 
 2/6 
 
 1. THE PLAN OF CAMPAIGN. By F. MABEL ROBINSON. 
 
 2. DISENCHANTMENT. By F. MABEL ROBINSON. 
 
 3. MR. BUTLER'S WARD. By MABEL ROBINSON. 
 
 4. HOVENDEN, V.C. By F. MABEL ROBINSON. 
 
 5. ELI'S CHILDREN. By G. MANVILLE FENN. 
 
 6. A DOUBLE KNOT. By G. MANVILLE FENN. 
 
 7. DISARMED. By M. BETHAM EDWARDS. 
 
 8. A LOST ILLUSION. By LESLIE KEITH. 
 
 9. A MARRIAGE AT SEA. By W. CLARK RUSSELL. 
 
 10. IN TENT AND BUNGALOW. By the Author of ' Indian 
 
 Idylls.' 
 
 11. MY STEWARDSHIP. By E. M'QUEEN GRAY. 
 
 12. A REVEREND GENTLEMAN. By J. M. COBBAN. 
 
 13. THE STORY OF CHRIS. By ROWLAND GREY. 
 
 Other Volumes will be announced in due course. 
 
 Books for Girls 
 
 A Series of Books by well-known Authors, bound uniformly. 
 
 Walford. A PINCH OF EXPERIENCE. By L. B. WAL- 
 FORD, Author of 'Mr. Smith.' With Illustrations by GORDON 
 BROWNE. Crown 8vo. 3^. 6d. 
 
 1 The clever authoress steers clear of namby-pamby, and invests her moral with a 
 fresh and striking dress. There is terseness and vivacity of style, and the illustra- 
 tions are admirable. 'A nti- Jacobin. 
 
MESSRS. METHUEN'S LIST 9 
 
 Molesworth. THE RED GRANGE. By Mrs. MOLESWORTH, 
 Author of * Carrots. ' With Illustrations *by GORDON BROWNE. 
 Crown 8vo. 3*. 6d. 
 
 'A volume in which girls will delight, and beautifully illustrated.' Pall Mall 
 Gazette. 
 
 Author of Mdle. Mori. 3 THE SECRET OF MADAME DE 
 Monluc. By the Author of 'The Atelier du Lys,' 'Mdle. Mori.' 
 Crown 8vo. $s. 6d. 
 'An exquisite literary cameo.' World. 
 
 Parr. DUMPS. By Mrs. PARR, Author of ' Adam and Eve,' 
 ' Dorothy Fox,' etc. Illustrated by W. PARKINSON. Crown 8v0. 
 3s. 6d. 
 
 ' One of the prettiest stories which even this clever writer has given the world for a 
 long time.' World. 
 
 Meade. OUT OF THE FASHION. By L. T. MEADE, Author 
 of 'A Girl of the People,' etc. With 6 Illustrations by W. PAGET. 
 Crown 8v0. 3$. 6d. 
 
 ' One of those charmingly-written social tales, which this writer knows so well how to 
 write. It is delightful reading, and is well illustrated by W. Paget.' Glasgow 
 Herald. 
 
 Meade. A GIRL OF THE PEOPLE. By L. T. MEADE, 
 Author of ' Scamp and I, 'etc. Illustrated by R. BARNES. Crown 
 8v0. 3.?. 6d. 
 
 An excellent story. Vivid portraiture of character, and broad and wholesome 
 
 lessons about life.' Spectator. 
 1 One of Mrs. Meade's most fascinating books.' Daily News. 
 
 Meade. HEPSY GIPSY. By L. T. MEADE. Illustrated by 
 
 EVERARD HOPKINS. Crown 8vo. 2s. 6d. 
 'Mrs. Meade has not often done better work than this.' Spectator. 
 
 Meade. THE HONOURABLE MISS : A Tale of a Country 
 Town. By L. T. MEADE, Author of ' Scamp and I,' ' A Girl of the 
 People,' etc. With Illustrations by EVERARD HOPKINS. Crown 
 8vo. %s. 6d. 
 
 Adams. MY LAND OF BEULAH. By MRS. LEITH ADAMS. 
 With a Frontispiece by GORDON BROWNE. Crown 8v0. 3^. 6d. 
 
 A 2 
 
to MESSRS. METHUEN'S LIST 
 
 ant) decent 
 
 Poetry 
 
 Rudyard Kipling. BARRACK-ROOM BALLADS; And 
 Other Verses. By RUDYARD KIPLING. Sixth Edition. Crown 
 8v0. 6s. 
 
 A Special Presentation Edition, bound in white buckram, with 
 extra gilt ornament. JS. 6d. 
 
 ' Mr. Kipling's verse is strong, vivid, full of character. . . . Unmistakable genius 
 rings in every line.' Times. 
 
 'The disreputable lingo of Cockayne is henceforth justified before the world ; for a 
 man of genius has taken it in hand, and has shown, beyond all cavilling, that in 
 its way it also is a medium for literature. You are grateful, and you say to 
 yourself, half in envy and half in admiration : " Here is a book ; here, or one is a 
 Dutchman, is one of the books of the year." ' National Observer. 
 
 ' " Barrack-Room Ballads " contains some of the best work that Mr. Kipling has 
 ever done, which is saying a good deal. " Fuzzy-Wuzzy," " Gunga Din," and 
 " Tommy," are, in our opinion, altogether superior to anything of the kind that 
 English literature has hitherto produced.' Athen&um. 
 
 1 These ballads are as wonderful in their descriptive power as they are vigorous in 
 their dramatic force. There are few ballads in the English language more 
 stirring than "The Ballad of East and West, " worthy to stand by the Border 
 ballads of Scott.' Spectator. 
 
 The ballads teem with imagination, they palpitate with emotion. We read them 
 with laughter and tears ; the metres throb in our pulses, the cunningly ordered 
 words tingle with life ; and if this be not poetry, what is?' Pall Mall Gazette. 
 
 Henley. LYRA HEROICA: An Anthology selected from the 
 best English Verse of the i6th, I7th, i8th, and iQth Centuries. By 
 WILLIAM ERNEST HENLEY, Author of 'A Book of Verse,' 'Views 
 and Reviews,' etc. Croivn 8vo. Stamped gilt buckram^ gilt top, 
 edges uncut. 6s. 
 
 Mr. Henley has brought to the task of selection an instinct alike for poetry and for 
 chivalry which seems to us quite wonderfully, and even unerringly, right.' 
 Guardian. 
 
 Tomson. A SUMMER NIGHT, AND OTHER POEMS. By 
 GRAHAM R. TOMSON. With Frontispiece by A. TOMSON. Fcap. 
 Svo. 2 s - *>d. 
 Also an edition on hand-made paper, limited to 50 copies. Large crown 
 
 Svo. los. 6d. net. 
 
 ' Mrs. Tomson holds perhaps the very highest rank among poetesses of English birth. 
 This selection will help her reputation.' Black and White. 
 
MESSRS. METHUEN'S LIST 11 
 
 Ibsen. BRAND. A Drama by HENRIK IBSEN. Translated b> 
 WILLIAM WILSON. Crown Svo. $s. 
 
 'The greatest world-poem of the nineteenth century next to "Faust." "Brand* 
 will have an astonishing interest for Englishmen. It is in the same set with 
 "Agamemnon," with " Lear," with the literature that we now instinctively regard 
 as high and holy.' Daily Chronicle. 
 
 ' Q." GREEN BAYS : Verses and Parodies. By " Q.," Author 
 
 of ' Dead Man's Rock ' etc. Second Edition. Fcap. Svo. %s. 6d. 
 'The verses display a rare and versatile gift of parody, great command of metre, and 
 a very pretty turn of humour.' Times. 
 
 "A. G." VERSES TO ORDER. By "A. G." Crown 8^, 
 
 cloth extra, gilt top. 2s. 6d. net. 
 
 A small volume of verse by a writer whose initials are well known to Oxford men. 
 ' A capital specimen of light academic poetry. These verses are very bright and 
 engaging, easy and sufficiently witty.' St. James's Gazette. 
 
 Hosken. VERSES BY THE WAY. BY J. D. HOSKEN 
 Printed on laid paper, and bound in buckram, gilt top. 5^. 
 Also a small edition on large Dutch hand-made paper. Price 
 I2s. 6d. net, 
 
 A Volume of Lyrics and Sonnets by J. D. Hosken, the Postman Poet, of Helston, 
 Cornwall, whose interesting career is now more or less well known to the literary 
 public. Q, the Author of 'The Splendid Spur,' etc., writes a critical and 
 biographical introduction. 
 
 Langbridge. A CRACKED FIDDLE. Being Selections from 
 the Poems of FREDERIC LANGBRIDGE. With Portrait. Crown Svo. $s. 
 
 Langbridge. BALLADS OF THE BRAVE : Poems of Chivalry, 
 Enterprise, Courage, and Constancy, from the Earliest Times to the 
 Present Day. Edited, with Notes, by Rev. F. LANGBRIDGE. 
 Crown Svo. Buckram 3*. 6d. School Edition, 2s. 6d. 
 ' A very happy conception happily carried out. These "Ballads of the Brave" are 
 intended to suit the real tastes of boys, and will suit the taste of the great majority.' 
 Spectator. ' The book is full of splendid things. 1 World. 
 
 History and Biography 
 
 Collingwood. JOHN RUSKIN : His Life and Work. By 
 \V. G. COLLINGWOOD, M.A., late Scholar of University College, 
 Oxford, Author of the 'Art Teaching of John Ruskin,' Editor of 
 Mr. Ruskin's Poems. 2 vols. Svo. 32*. Second Edition. 
 This important work is written by Mr. Collingwood, who has been for some years 
 Mr. Ruskin's private secretary, and who has had unique advantages in obtaining 
 
12 MESSRS. METHUEN'S LIST 
 
 materials for this book from Mr. Ruskin himself and from his friends. It contains 
 a large amount of new matter, and of letters which have never been published, 
 and is, in fact, a full and authoritative biography of Mr. Ruskin. The book 
 contains numerous portraits of Mr. Ruskin, including a coloured one from a 
 water-colour portrait by himself, and also 13 sketches, never before published, by 
 Mr. Ruskin and Mr. Arthur Severn. A bibliography is added. 
 
 ' No more magnificent volumes have been published for a long time than " The Life 
 and Work of John Ruskin." . . .' Times. 
 
 ' This most lovingly written and most profoundly interesting book.' Daily News. 
 
 ' It is long since we have had a biography with such varied delights of substance 
 and of form. Such a book is a pleasure for the day, and a joy for ever.' Daily 
 Chronicle. 
 
 ' Mr. Ruskin could not well have been more fortunate in his biographer.' Globe. 
 
 'A noble monument of a noble subject. One of the most beautiful books about one 
 of the noblest lives of our century.' Glasgow Herald. 
 
 Gladstone. THE SPEECHES AND PUBLIC ADDRESSES 
 OF THE RT. HON. W. E. GLADSTONE, M.P. With Notes 
 and Introductions. Edited by A. W. HUTTON, M.A. (Librarian of 
 the Gladstone Library), and H. J. COHEN, M.A. With Portraits. 
 Svo. Vol. X. i2s. 6d. 
 
 Russell. THE LIFE OF ADMIRAL LORD COLLING- 
 WOOD. By W. CLARK RUSSELL, Author of ' The Wreck of the 
 Grosvenor.' With Illustrations by F. BRANGWYN. 8vo. i$s. 
 
 ' A really good book.' Saturday Review. 
 
 ' A most excellent and wholesome book, which we should like to see in the hands of 
 every boy in the country.' St. James's Gazette. 
 
 Clark. THE COLLEGES OF OXFORD : Their History and 
 their Traditions. By Members of the University. Edited by A. 
 CLARK, M.A., Fellow and Tutor of Lincoln College. 8vo. izs. 6d. 
 
 'Whether the reader approaches the book as a patriotic member of a college, as an 
 antiquary, or as a student of the organic growth of college foundation, it will amply 
 reward his attention.' Times, 
 
 'A delightful book, learned and lively." Academy. 
 
 1 A work which will certainly be appealed to for many years as the standard book on 
 the Colleges of Oxford." Athenceum. 
 
 Hulton. RIXAE OXONIENSES : An Account of the Battles 
 of the Nations, The Struggle between Town and Gown, etc. By 
 S. F. HULTON, M.A. Crown 8vo. 5*. 
 
 James. CURIOSITIES OF CHRISTIAN HISTORY PRIOR 
 TO THE REFORMATION. By CROAKE JAMES, Author of 
 * Curiosities of Law and Lawyers.' Crown 8vo. 7s. 6d. 
 
MESSRS. METHUEN'S LIST 13 
 
 Perrens. THE HISTORY OF FLORENCE FROM THE 
 TIME OF THE MEDICIS TO THE FALL OF THE 
 REPUBLIC. By F. T. PERRENS. Translated by HANNAH 
 LYNCH. In three volumes. Vol. I. 8vo. \is. 6d. 
 
 This is a translation from the French of the best history of Florence in existence. 
 This volume covers a period of profound interest political and literary and 
 is written with great vivacity. 
 
 ' This is a standard book by an honest and intelligent historian, who has deserved 
 well of his countrymen, and of all who are interested in Italian history.' Man- 
 chester Guardian. 
 
 Kaufmann. CHARLES KINGSLEY. By M. KAUFMANN, 
 
 M.A. Crown 8vo. $s. 
 
 A biography of Kingsley, especially dealing with his achievements in social reform. 
 1 The author has certainly gone about his work with conscientiousness and industry.' 
 Sheffield Daily Telegraph. 
 
 Oliphant. THOMAS CHALMERS : A Biography. By Mrs. 
 
 OLIPHANT. With Portrait. Crown &vo. Buckram, $s. 
 ' A well-executed biography, worthy of its author and of the remarkable man who is 
 its subject. Mrs. Oliphant relates lucidly and dramatically the important part 
 which Chalmers played in the memorable secession.' Times. 
 
 'Written with all the facile literary grace that marks this indefatigable authoress' 
 work, it presents a very complete picture of Chalmers as he lived and worked. . . . 
 The salient points in his many-sided life are seized with unerring judgment.' 
 North British Daily Mail. 
 
 Wells. THE TEACHING OF HISTORY IN SCHOOLS. A 
 
 Lecture delivered at the University Extension Meeting in Oxford, 
 Aug. 6th, 1892. By J. WELLS, M.A., Fellow and Tutor of Wadham 
 College, and Editor of ' Oxford and Oxford Life.' Crown 8vo> 6d. 
 
 Pollard. THE JESUITS IN POLAND. By A. F. POLLARD, 
 B.A. Oxford Prize Essays The Lothian Prize Essay 1892. Crown 
 8v0. 2s. 6d. net. 
 
 Clifford. THE DESCENT OF CHARLOTTE COMPTON 
 (BARONESS FERRERS DE CHARTLEY). By her Great-Granddaughter 
 ISABELLA G. C. CLIFFORD. Small tfo. IDS. 6d. net. 
 
 General Literature 
 
 Bowden. THE IMITATION OF BUDDHA: Being Quota- 
 tions from Buddhist Literature for each Day in the Year. Compiled 
 by E. M. BOWDEN. With Preface by Sir EDWIN ARNOLD. Third 
 Edition. i6mo. 2s. 6d. 
 
14 MESSRS. METHUEN'S LIST 
 
 Ditchfield. OUR ENGLISH VILLAGES : Their Story and 
 their Antiquities. By P. H. DITCHFIELD, M.A., F.R.H.S., Rector 
 of Barkham, Berks. Post Svo. 2s. 6d. Illustrated. 
 
 1 An extremely amusing and interesting little book, which should find a place in 
 every parochial library. 1 Guardian. 
 
 Ditchfield. OLD ENGLISH SPORTS. By P. H. DITCH- 
 
 FIELD, M.A. Crown 8vo. 2s. 6d. Illustrated. 
 'A charming account of old English Sports.' Morning Post. 
 
 Burne. PARSON AND PEASANT: Chapters of their 
 Natural History.- By J. B. BURNE, M.A., Rector of Wasing. 
 Crown 8vo. $s. 
 
 ' " Parson and Peasant " is a book not only to be interested in, but to learn something 
 from a book which may prove a help to many a clergyman, and broaden th 
 hearts and ripen the charity of laymen." Derby Mercury. 
 
 Massee. A MONOGRAPH OF THE MYXOGASTRES. By 
 GEORGE MASSEE. With 12 Coloured Plates. Royal 8vo. i8s. net. 
 
 This is the only work in English on this important group. It contains 12 Coloured 
 Plates, produced in the finest style of chromo-lithography. 
 
 1 Supplies a want acutely felt Its merits are of a high order, and it is one of the 
 most important contributions to systematic natural science which have lately 
 appeared. ' Westminster Review. 
 
 'A work much in advance of any book in the language treating of this group ol 
 organisms. It is indispensable to every student of the Mxyogastres. The 
 coloured plates deserve high praise for their accuracy and execution.' Nature. 
 
 Cunningham. THE PATH TOWARDS KNOWLEDGE: 
 Essays on Questions of the Day. By W. CUNNINGHAM, D.D., 
 Fellow of Trinity College, Cambridge, Professor of Economics at 
 King's College, London. Crown 8vo. qs. 6d. 
 Essays on Marriage and Population, Socialism, Money, Education, Positivism, etc. 
 
 Bushill. PROFIT SHARING AND THE LABOUR QUES- 
 TION. By T. W. BUSHILL, a Profit Sharing Employer. With an 
 Introduction by SEDLEY TAYLOR, Author of ' Profit Sharing between 
 Capital and Labour.' Crown 8vo. 2s. 6d. 
 
 John Beever. PRACTICAL FLY-FISHING, Founded on 
 Nature, by JOHN BEEVER, late of the Thwaite House, Coniston. A 
 New Edition, with a Memoir of the Author by W. G. COLLINGWOOD, 
 M.A., Author of 'The Life and Work of John Ruskin/etc. Also 
 additional Notes and a chapter on Char-Fishing, by A. and A. R. 
 SEVERN. With a specially designed title-page. Crown 8vo. 3^. 6d. 
 little book on Fly-Fishing by an old friend of Mr. Ruskin. It has been out of 
 print for some time, and being still much in request, is now issued with a Memoir 
 of the Author by W. G. Collingwood. 
 
MESSRS. METHUEN'S LIST 15 
 
 Anderson Graham. NATURE IN BOOKS : Studies in Literary 
 Biography. By P. ANDERSON GRAHAM. Crown Svo. 6s. 
 
 The chapters are entitled : I. ' The Magic of the Fields ' (Jefferies). II. ' Art and 
 Nature' (Tennyson). III. 'The Doctrine of Idleness' (Thoreau). IV. 'The 
 Romance of Life ' (Scott). V. ' The Poetry of Toil ' (Burns). VI. 'The Divinity 
 of Nature ' (Wordsworth). 
 
 Wells. OXFORD AND OXFORD LIFE. By Members of 
 the University. Edited by J. WELLS, M.A., Fellow and Tutor of 
 Wadham College. Crown 8vo. 3-r. 6d. 
 
 This work contains an account of life at Oxford intellectual, social, and religious 
 a careful estimate of necessary expenses, a review of recent changes, a statement 
 of the present position of the University, and chapters on Women's Education, 
 aids to study, and University Extension. 
 
 'We congratulate Mr. Wells on the production of a readable and intelligent account 
 of Oxford as it is at the present time, written by persons who are, with hardly an 
 exception, possessed of a close acquaintance with the system and life of the 
 University.' A theneeum. 
 
 Driver. SERMONS ON SUBJECTS CONNECTED WITH 
 THE OLD TESTAMENT. By S. R. DRIVER, D.D., Canon of 
 Christ Church, Regius Professor of Hebrew in the University of 
 Oxford. Crown 8vo. 6s. 
 
 'A welcome volume to the author's famous ' Introduction.' No man can read these 
 discourses without feeling that Dr. Driver is fully alive to the deeper teaching of 
 the Old Testament.' Guardian, 
 
 Cheyne. FOUNDERS OF OLD TESTAMENT CRITICISM: 
 Biographical, Descriptive, and Critical Studies. By T. K. CHEYNE, 
 D.D., Oriel Professor of the Interpretation of Holy Scripture at 
 Oxford. Large crown 8z'0. Js. 6d. [Ready, 
 
 This important book is a historical sketch of O.T. Criticism in the form of biographi- 
 cal studies from the days of Eichhorn to those of Driver and Robertson Smith. 
 It is the only book of its kind in English. 
 
 ' The volume is one of great interest and value. It displays all the author's well- 
 known ability and learning, and its opportune publication has laid all students of 
 theology, and specially of Bible criticism, under weighty obligation.' Scotsman. 
 
 1 A very learned and instructive work.' Times. 
 
 WORKS BY 
 
 S. Baring Gould, Author of ' Mehalah,' etc. 
 OLD COUNTRY LIFE. With Sixty-seven Illustrations by 
 W. PARKINSON, F. D. BEDFORD, and F. MASEY. Large Crown 
 8vo, doth super extra, top edge gilt, los. 6d. Fourth and Cheaper 
 Edition. 6s. 
 
 1 " Old Country Life," as healthy wholesome reading, full of breezy life and move- 
 ment, full of quaint stories vigorously told, will not be excelled by any book 
 to be published throughout the year. Sound, hearty, and English to the core.' 
 World. 
 
16 MESSRS. METHUEN'S LIST 
 
 HISTORIC ODDITIES AND STRANGE EVENTS. Third 
 
 Edition, Crown Svo. 6s. 
 
 1 A collection of exciting and entertaining chapters. The whole volume is delightful 
 reading. ' Times. 
 
 FREAKS OF FANATICISM. Third Edition. Crown*. 6s. 
 
 1 Mr. Baring Gould has a keen eye for colour and effect, and the subjects he has 
 chosen give ample scope to his descriptive and analytic faculties. A perfectly 
 fascinating book. 1 Scottish Leader, 
 
 SONGS OF THE WEST : Traditional Ballads and Songs of 
 the West of England, with their Traditional Melodies. Collected 
 by S. BARING GOULD, M.A., and H. FLEETWOOD SHEPPARD, 
 M.A. Arranged for Voice and Piano. In 4 Parts (containing 25 
 Songs each), Parts /., //., ///., 3*. each. Part IV., 55. In one 
 Vol., roan, i$s. 
 
 'A rich and varied collection of humour, pathos, grace, and poetic fancy.' Saturday 
 Review. 
 
 YORKSHIRE ODDITIES AND STRANGE EVENTS. 
 Fourth Edition. Crown 8vo. 6s. 
 
 STRANGE SURVIVALS AND SUPERSTITIONS. With 
 Illustrations. By S. BARING GOULD. Croivn Svo. 'js. 6d. 
 
 A book on such subjects as Foundations, Gables, Holes, Gallows, Raising the Hat, Old 
 Ballads, etc. etc. It traces in a most interesting manner their origin and history. 
 
 ' We have read Mr. Baring Gould's book from beginning to end. It is full of quaint 
 and various information, and there is not a dull page in it. ' Notes and Queries. 
 
 THE TRAGEDY OF THE CAESARS: The 
 
 Emperors of the Julian and Claudian Lines. With numerous Illus- 
 trations from Busts, Gems, Cameos, etc. By S. BARING GOULD, 
 Author of 'Mehalah,' etc. Second Edition. 2 vols. Royal 8v0. 30^. 
 This book is the only one in English which deals with the personal history of the 
 Caesars, and Mr. Baring Gould has found a subject which, for picturesque detail 
 and sombre interest, is not rivalled by any work of fiction. The volumes are 
 copiously illustrated. 
 
 ' A most splendid and fascinating book on a subject of undying interest. The great 
 feature of the book is the use the author has made of the existing portraits of the 
 Caesars, and the admirable critical subtlety he has exhibited in dealing with this 
 line of research. It is brilliantly written, and the illustrations are supplied on a 
 scale of profuse magnificence. ' Daily Chronicle. 
 
 ' The volumes will in no sense disappoint the general reader. Indeed, in their way, 
 there is nothing in any sense so good in English. . . . Mr. Baring Gould has 
 presented his narrative in such a way as not to make one dull page.' Athenaum. 
 
 JACQUETTA, and other Stories. Crown %vo. $s. 6d. 
 
MESSRS. METHUEN'S LIST 17 
 
 ARM I NELL: A Social Romance. New Edition. Crown &vo. 
 35. 6d. 
 
 1 To say that a book is by the author of " Mehalah" is to imply that it contains a 
 story cast on strong lines, containing dramatic possibilities, vivid and sympathetic 
 descriptions of Nature, and a wealth of ingenious imagery. All these expecta- 
 tions are justified by " Arminell." ' Speaker. 
 
 URITH: A Story of Dartmoor. Third Edition. CrownKvo. $s.6d. 
 
 ' The author is at his best.' Times. 
 
 ' He has nearly reached the high water-mark of " Mehalah." ' National Observer. 
 
 MARGERY OF QUETHER, and other Stories. Crown Svo. 
 
 3s. 6d. 
 IN THE ROAR OF THE SEA : A Tale of the Cornish Coast. 
 
 New Edition. 3*. 6d. 
 MRS. CURGENVEN OF CURGENVEN. Third Edition. 6s. 
 
 Fiction 
 
 Pryce. TIME AND THE WOMAN. By RICHARD PRYCE, 
 
 Author of Miss Maxwell's Affections,' 'The Quiet Mrs. Fleming,' 
 etc. New and Cheaper Edition. Crown Svo. 6s. 
 
 ' Mr.\Pryce's work recalls the style of Octave Feuillet, by its clearness, conciseness, 
 its literary reserve." Athenaunt, 
 
 Gray. ELSA. A Novel. By E. M'QUEEN GRAY. CroivnZvo. 6s. 
 
 'A charming novel. The characters are not only powerful sketches, but minutely 
 and carefully finished portraits.' Guardian. 
 
 Anthony Hope. A CHANGE OF AIR : A Hovel. By 
 ANTHONY HOPE, Author of 'Mr. Witt's Widow,' etc. I vol. 
 Crown &vo. 6s. 
 A bright story by Mr. Hope, who has, the Athencrum says, 'a decided outlook and 
 
 individuality of his own. 1 
 
 'A graceful, vivacious comedy, true to human nature. The characters are traced 
 with a masterly hand.' Times. 
 
 Edna Lyall. DERRICK VAUGHAN, NOVELIST. By 
 EDNA LYALL, Author of 'Donovan.' Crown 8vo. $ist Thousand. 
 3J. 6d. ; paper, is. 
 
 Lynn Linton. THE TRUE HISTORY OF JOSHUA DAVID- 
 SON, Christian and Communist. By E. LYNN LINTON. Eleventh 
 Edition. Post Svo. is. 
 
 Dicker. A CAVALIER'S LADYE. By CONSTANCE DICKER. 
 With Illustrations. Crown Svo. &. 6d. 
 
i8 MESSRS. METHUEN'S LIST 
 
 Author of 'Vera.' THE DANCE OF THE HOURS. By 
 
 the Author of * Vera,' Blue Roses,' etc. Crown Svo. 6s. 
 'A musician's dream, pathetically broken off at the hour of its realisation, is vividly 
 represented in this book. . . . Well written and possessing many elements of 
 interest. The success of " The Dance of the Hours" may be safely predicted.'' 
 Morning Post, 
 
 Norris. A Deplorable Affair. By W. E. NORRIS, Author of 
 
 'His Grace.' Crown Svo. $s. 6d. 
 
 'What with its interesting story, its graceful manner, and its perpetual good 
 humour, the book is as enjoyable as any that has come from its author's pen.' 
 Scotsman. 
 
 Dickinson. A VICAR'S WIFE. By EVELYN DICKINSON. 
 
 Crown Svo. $s. 6d. 
 
 Prowse. THE POISON OF ASPS. By R. ORTON PROWSE. 
 Crown Svo. 35. 6d. 
 
 Parker. PIERRE AND HIS PEOPLE. By GILBERT 
 PARKER. Crown Svo. Buckram. 6s. 
 
 ' Stories happily conceived and finely executed. There is strength and genius in Mr. 
 Parker's style.' Daily Telegraph. 
 
 Marriott Watson. DIOGENES OF LONDON and other 
 Sketches. By H. B. MARRIOTT WATSON, Author of The Web 
 of the Spider.' Crown Svo. Buckram. 6s. 
 
 ' By all those who delight in the uses of words, who rate the exercise of prose above 
 the exercise of verse, who rejoice in all proofs of its delicacy and its strength, who 
 believe "that English prose is chief among the moulds of thought, by these 
 Mr. Marriott Watson's book will be welcomed/ National Observer, 
 
 3/6 
 
 Methuen's Novel Series 
 
 A series of copyright Novels, by well-known Authors, 
 bound in red buckram, at the price of three shillings and 
 sixpence. The first volumes (ready) are : 
 
 1. JACQUETTA. By S. BARING GOULD, Author of ' Mehalah,' 
 
 etc. 
 
 2. ARM I NELL : A Social Romance. By S. BARING GOULD, 
 
 Author of * Mehalah,' etc. 
 
 3. MARGERY OF QUETHER. By S. BARING GOULD. 
 
 4. URITH. By S. BARING GOULD. 
 
MESSRS. METHUEN'S LIST 19 
 
 5. IN THE ROAR OF THE SEA. By S. BARING GOULD. 
 
 6. DERRICK VAUGHAN, NOVELIST. With Portrait of 
 
 Author. By EDNA LYALL, Author of 'Donovan,' etc. Also 
 
 paper, is. 
 
 7- JACK'S FATHER. By W. E. NORRIS. 
 8. MY DANISH SWEETHEART. By W. CLARK RUSSELL. 
 Other Volumes will be announced in due course. 
 
 HALF-CROWN NOVELS 
 
 A Series of Novels by popular Authors, tastefully 
 bound in cloth. 
 
 2/6 
 
 1. THE, PLAN OF CAMPAIGN. By F. MABEL ROBINSON. 
 
 2. DISENCHANTMENT. By F. MABEL ROBINSON. 
 
 3. MR. BUTLER'S WARD. By MABEL ROBINSON. 
 
 4. HOVENDEN,'V.C. By F. MABEL ROBINSON. 
 
 5. ELI'S CHILDREN. By G. MANVILLE FENN. 
 
 6. A DOUBLE KNOT. By G. MANVILLE FENN. 
 
 7. DISARMED. By M. BETHAM EDWARDS. 
 
 8. A LOST ILLUSION. By LESLIE KEITH. 
 
 9. A MARRIAGE AT SEA. By W. CLARK RUSSELL. 
 
 10. IN TENT AND BUNGALOW. By the Author qf < Indian 
 
 Idylls.' 
 
 11. MY STEWARDSHIP. By E. M'QUEEN GRAY. 
 
 12. A REVEREND GENTLEMAN. By J. M. COBBAN. 
 
 13. THE STORY OF CHRIS. By ROLAND GREY. 
 
 Other volumes will be announced in due course. 
 
 NEW TWO-SHILLING EDITIONS ^ 
 
 Crown Svo, Ornamental Boards. ~ 
 
 ELI'S CHILDREN. By G. MANVILLE FENN. 
 
 DISENCHANTMENT. By F. MABEL ROBINSON. 
 
 THE PLAN OF CAMPAIGN. By F. MABEL ROBINSON. 
 
20 MESSRS. METHUEN'S LIST 
 
 Crown Svo. Picture Boards. 
 
 A REVEREND GENTLEMAN. By J. MACLAREN COBBAN. 
 MR. BUTLER'S WARD. By MABEL ROBINSON. 
 JACK'S FATHER. By W. E. NORRIS. 
 THE QUIET MRS. FLEMING. By RICHARD PRYCE. 
 
 Books for Boys and Girls 
 
 Outhell. ONLY A GUARD-ROOM DOG. By Mrs. CUTHELL. 
 With 1 6 Illustrations by W. PARKINSON. Square Crown Svo. 6s. 
 
 1 This is a charming story. Tangle was but a little mongrel Skye terrier, but he had a 
 big heart in his little body, and played a hero's part more than once. The book 
 can be warmly recommended.' Standard, 
 
 Collingwood. THE DOCTOR OF THE JULIET. By HARRY 
 COLLINGWOOD, Author of 'The Pirate Island,' etc. Illustrated by 
 GORDON BROWNE. Crown Svo. 6s. 
 
 1 "The Doctor of the Juliet," well illustrated by Gordon Browne, is one of Harry 
 Collingwood's best efforts. 1 Morning Post. 
 
 Clark Russell. MASTER ROCKAFELLAR'S VOYAGE. By 
 W. CLARK RUSSELL, Author of The Wreck of the Grosvenor,' etc. 
 Illustrated by GORDON BROWNE. Crown Svo. $s. 6d. 
 
 'Mr. Clark Russell's story of "Master Rockafellar's Voyage" will be among the 
 favourites of the Christmas books. There is a rattle and '' go " all through it, and 
 its illustrations are charming in themselves, and very much above the average in 
 the way in which they are produced.' Guardian. 
 
 Manville Fenn. SYD B ELTON : Or, The Boy who would not 
 go to Sea. By G. MANVILLE FENN, Author of ' In the King's 
 Name,' etc. Illustrated by GORDON BROWNE. Crown Svo. 3*. 6d. 
 
 1 Who among the young story-reading public will not rejoice at the sight of the old 
 combination, so often proved admirable a story by Manville Fenn, illustrated 
 by Gordon Browne ? The story, too, is one of the good old sort, full of life and 
 vigour, breeziness and fun.' Journal of Education. 
 
 Walford. A PINCH OF EXPERIENCE. By. L. B. WAL- 
 FORD, Author of 'Mr. Smith.' With Illustrations by GORDON 
 BROWNE. Crown Svo. 3*. 6d. 
 
 ' The clever authoress steers clear of namby-pamby, and invests her moral with a 
 fresh and striking dress. There is terseness and vivacity of style and the illustra- 
 tions re admirable.' Anti- Jacobin. 
 
MESSRS. METHUEN'S LIST 21 
 
 Molesworth. THE RED GRANGE. By Mrs. MOLESWORTH, 
 Author of ' Carrots.' With Illustrations by GORDON BROWNE. 
 Crown Svo. 3*. 6d. 
 
 'A volume in which girls will delight, and beautifully illustrated. 'Pall Mall 
 Gazette. 
 
 Author of ' Mdle. Mori.' THE SECRET OF MADAME DE 
 Monluc. By the Author of The Atelier du Lys,' 'Mdle. Mori.' 
 Crown Svo. 3s. 6d. 
 1 An exquisite literary cameo.' World. 
 
 Parr. DUMPS. By Mrs. PARR, Author of 'Adam and Eve,' 
 ' Dorothy Fox,' etc. Illustrated by W. PARKINSON. Crown Svo. 
 3s. 6d. 
 
 ' One of the prettiest stories which even this clever writer has given the world for a 
 long time.' World. 
 
 Meade. OUT OF THE FASHION. By L. T. MEADE, Author 
 of ' A Girl of the People,' etc. With 6 illustrations by W. PAGET. 
 Crown Svo, 35. 6d. 
 
 ' One of those charmingly-written social tales, which this writer knows so well how to 
 write. It is delightful reading, and is well illustrated by W. Paget." Glasgow 
 Herald. 
 
 Meade. A GIRL OF THE PEOPLE. By L. T. MEADE, 
 Author of Scamp and I,' etc. Illustrated by R. BARNES. Crown 
 Svo. 3*. 6d. 
 
 'An excellent story. Vivid portraiture of character, and broad and wholesome 
 
 lessons about life. ' Spectator. 
 1 One of Mrs. Meade's most fascinating books.' Daily News. 
 
 Meade. HEPSY GIPSY. By L. T. MEADE. Illustrated by 
 
 EVERARD HOPKINS. Crown Svo. 2s. 6d. 
 1 Mrs. Meade has not often done better work than this.' Spectator. 
 
 Meade. THE HONOURABLE MISS: A Tale of a Country 
 Town. By L. T. MEADE, Author of ' Scamp and I,' ' A Girl of the 
 People,' etc. With Illustrations by EVERARD HOPKINS, Crown 
 Svo. 3s. 6d. 
 
 Adams. MY LAND OF BEULAH. By Mrs. LEITH ADAMS. 
 With a Frontispiece by GORDON BROWNE. Crown Svo. 35. 6d. 
 
22 MESSRS. METHUEN'S LIST 
 
 Leaders of Religion 
 
 Edited by H. C. BEECHING, M.A. With Portrait, crown Svo. 2s. 6d. 
 
 A series of short biographies of the most prominent leaders 
 of religious life and thought. 
 
 The following are ready 
 CARDINAL NEWMAN. By R. H. HUTTON. 
 
 ' Few who read this book will fail to be struck by the wonderful insight it displays 
 into the nature of the Cardinal's genius and the spirit of his life." WILFRID 
 WARD, in the Tablet. 
 
 'Full of knowledge, excellent in method, and intelligent in criticism. We regard it 
 as wholly admirable.' Academy. 
 
 JOHN WESLEY. By J. H. OVERTON, M.A. 
 
 ' It is well done : the story is clearly told, proportion is duly observed, and there is 
 no lack either of discrimination or of sympathy." Manchester Guardian. 
 
 BISHOP WILBERFORCE. By G. W. DANIEL, M.A. 
 CHARLES SIMEON. By H. C. G. MOULE, M.A. 
 Other volumes will be announced in due course. 
 
 University Extension Series 
 
 A series of books on historical, literary, and scientific subjects, suitable 
 for extension students and home reading circles. Each volume is com- 
 plete in itself, and the subjects are treated by competent writers in a 
 broad and philosophic spirit. 
 
 Edited by J. E. SYMES, M.A., 
 
 Principal of University College, Nottingham. 
 
 Crown 8v0. Price (with some exceptions) 2s. 6d. 
 
 The following volumes are ready : 
 
 THE INDUSTRIAL HISTORY OF ENGLAND. By H. DE 
 B. GIBBINS, M.A., late Scholar of Wadham College, Oxon., Cobden 
 Prizeman. Third Edition. With Maps and Plans. 3*. 
 A compact and clear story of our industrial development. A study of this concise 
 but luminous book cannot fail to give the reader a clear insight into the principal 
 phenomena of our industrial history. The editor and publishers are to be congrat- 
 ulated on this first volume of their venture, and we shall look with expectant 
 interest for the succeeding volumes of the series.' University Extension Journal, 
 
MESSRS. METHUEN'S LIST 23 
 
 A HISTORY OF ENGLISH POLITICAL ECONOMY. By 
 L. L. PRICE, M.A., Fellow of Oriel College, Oxon. 
 
 PROBLEMS OF POVERTY : An Inquiry into the Industrial 
 Conditions of the Poor. By J. A. HOBSON, M.A. 
 
 VICTORIAN POETS. By A. SHARP. 
 
 THE FRENCH REVOLUTION. By J. E. SYMES, M.A. 
 
 PSYCHOLOGY. By F. S. GRANGER, M.A., Lecturer in Philo- 
 sophy at University College, Nottingham. 
 
 THE EVOLUTION OF PLANT LIFE : Lower Forms. By 
 G. MASSEE, Kew Gardens. With Illustrations. 
 
 AIR AND WATER. Professor V. B. LEWES, M.A. Illustrated. 
 
 THE CHEMISTRY OF LIFE AND HEALTH. By C. W. 
 
 KIMMINS, M.A. Camb. Illustrated. 
 
 THE MECHANICS OF DAILY LIFE. By V. P. SELLS, M.A. 
 Illustrated. 
 
 ENGLISH SOCIAL REFORMERS. H. DE B. GIBBINS, M.A. 
 
 ENGLISH TRADE AND FINANCE IN THE SEVEN- 
 TEENTH CENTURY. By W. A. S. HEWINS, B.A. 
 
 Social Questions of To-day 
 
 Edited by H. DE B. GIBBINS, M.A. 
 
 Crown 8v0. 2s. 6d. f+ \i\ 
 
 A series of volumes upon those topics of social, economic, I 
 
 and industrial interest that are at the present moment fore- 
 most in the public mind. Each volume of the series is written by an 
 author who is an acknowledged authority upon the subject with which 
 he deals. 
 
 The following' Volumes of the Series are ready : 
 
 TRADE UNIONISM NEW AND OLD. By G. HOWELL, 
 M.P., Author of * The Conflicts of Capital and Labour.' 
 
 THE CO-OPERATIVE MOVEMENT TO-DAY. By G. J. 
 HOLYOAKE, Author of The History pf Co-operation.' 
 
24 MESSRS. METHUEN'S LIST 
 
 MUTUAL THRIFT. By Rev. J. FROME WILKINSON, M.A., 
 
 Author of ' The Friendly Society Movement.' 
 
 PROBLEMS OF POVERTY : An Inquiry into the Industrial 
 Conditions of the Poor. By J. A. HOBSON, M.A. 
 
 THE COMMERCE OF NATIONS. By C. F. BASTABLE, 
 M.A., Professor of Economics at Trinity College, Dublin. 
 
 THE ALIEN INVASION. By W. H. WILKINS, B.A., Secretary 
 to the Society for Preventing the Immigration of Destitute Aliens. 
 
 THE RURAL EXODUS. By P. ANDERSON GRAHAM. 
 LAND NATIONALIZATION. By HAROLD Cox, B.A. 
 
 A SHORTER WORKING DAY. By H. DE B. GIBBINS 
 and R. A. HADFIELD, of the Hecla Works, Sheffield. 
 
 BACK TO THE LAND : An Inquiry into the Cure for Rural 
 Depopulation. By H. E. MOORE, 
 
 Edinburgh: T. &> A. Constable, Printers to Her Majesty.