,,cloud ,,studies ,illustration3 ,a ,,sunset ,,sky4 ,frontispiece4 ,,cloud ,,studies ,by ,,arthur ,w4 ,,clayden1 ,m4,a4 ,,principal ,,of ,,the ,,royal ,,albert ,,memorial ,,college1 ,,exeter ,,london ,,john ,,murray1 ,,albemarle ,,street #aije ,,printed ,,by ,,william ,,clowes ,,and ,,sons1 ,,limited1 ,,london ,,and ,,beccles4 ,,preface ,to the meteorologist ,i hope the following pages may prove not only of some interest1 but of practical value as a small step towards that greater exactness of language which is essential before we can attempt to explain all the details of cloud structure1 or even interchange our ideas and observations with adequate precision4 ,the varieties depicted and described have been selected from many hundreds1 as those which seem to me to show such differences of form as to imply distinct differences in the conditions to which they are due4 ,i have not attempted to deal with the physical causes of condensation except in a general way1 being unwilling to introduce diagrams of isothermals and adiabatics and such purely scientific methods into a work also intended for a wider public4 ,for those who wish to pursue this part of the subject ,i have appended a list of papers from the ,quarterly ,journal of the ,royal ,meteorological ,society and other sources1 which may serve as references4 ,i also hope that some more votaries of the science may be induced to realize that meteorology does not consist solely of the tabulation of long columns of records1 but includes subjects for investigation as much more beautiful as they are more difficult4 ,to the artist ,i trust they may also be of some use1 by calling attention to the variety and exquisite beauty of the sky4 ,nothing is more extraordinary in art than the general negligence of cloud-forms4 ,many of them are quite as worthy of careful drawing as the leaves of a tree1 the flowers of a field1 the ripples on a stream1 or the texture of a carpet1 or a marble pavement4 ,yet it is the common rule to find pictures1 which are otherwise marvellous examples of skill and care1 disfigured by impossible skies with vague1 shapeless clouds1 as untrue to nature as it would be possible to make them4 ,grace of outline1 delicacy of detail and texture1 richness of contrast1 beauty of form and light and colour1 all are present in the skies1 and combine to make a whole well worthy of the best that art can give4 ,the illustrations ,i offer are not selected for pictorial effect2 they are chosen from a purely scientific point of view2 but they are enough to indicate what could be done if the facts of nature were treated with high artistic skill4 ,in addition to the meteorologist and the artist1 there are a much larger number who follow neither profession1 but who love ,nature in all her moods2 and to them also ,i hope these pages may be of interest4 ,indeed1 if only a few of them should be stimulated to take up a branch of nature study which has given me many an hour of quiet enjoyment1 the labour of bringing these notes together will not have been in vain4 ,,arthur ,w4 ,,clayden4 ,st4 ,john's1 ,exeter4 ,,contents ,,chapter ,,page ,i4 ,introductory #a ,,ii4 ,cirrus #ba ,,iii4 ,cirro-stratus and ,cirro-cumulus #de ,,iv4 ,alto ,clouds #ei ,v4 ,lower ,clouds #ga ,,vi4 ,cumulus #hd ,,vii4 ,cumulo-nimbus #aje ,,viii4 ,wave ,clouds #aai ,,ix4 ,cloud ,altitudes #acg ,x4 ,cloud ,nomenclature #aed ,,xi4 ,cloud ,photography #afe ,references #aha ,index #ahc ,,list ,,of ,,illustrations ,,plate ,,page ,a ,sunset ,sky ,frontispiece #a4 ,part of a ,great ,halo #bb #b4 ,part of a ,solar ,halo #bc #c4 ,cirro-nebula changing to ,cirro-stratus #bd #d4 ,cirro-nebula changing to ,cirro-cumulus #bg #e4 ,high ,cirrus4 7,cirrus ,excelsus7 #ca #f4 ,windy ,cirrus4 7,cirrus ,ventosus7 #cb #g4 ,thread ,cirrus4 7,cirro-filum7 #cd #h4 ,tailed ,cirrus4 7,cirrus ,caudatus7 #ce #i4 ,hazy ,cirrus4 7,cirrus ,nebulosus7 #cf #aj4 ,change ,cirrus4 7,cirrus ,inconstans7 #cg #aa4 ,common ,cirrus4 7,cirrus ,communis7 #dj #ab4 ,band ,cirrus4 7,cirrus ,vittatus7 #da #ac4 ,band ,cirrus4 7,cirrus ,vittatus7 #db #ad4 ,hazy ,cirro-stratus4 7,cirro-stratus ,nebulosus7 #df #ae4 ,cirro-stratus #dg #af4 ,cirro-stratus4 7,cirro-stratus ,communis7 #dh #ag4 ,flocculent ,cirro-stratus4 7,cirro-stratus ,cumulosus7 #di #ah4 ,cirro-stratus and ,cirro-cumulus #ej #ai4 ,cirro-cumulus #ej #bj4 ,hazy ,cirro-cumulus4 7,cirro-cumulus ,nebulosus7 #ea #ba4 ,hazy ,cirro-cumulus4 7,cirro-cumulus ,nebulosus7 #ea #bb4 ,a ,sunset ,sky #eb #bc4 ,speckle ,cloud 7,ley74 7,cirro-macula7 #ec #bd4 ,cirrus ,caudatus and ,cirro-macula #ee #be4 ,alto-cumulus ,informis #fd #bf4 ,hazy ,alto-cumulus4 7,alto-cumulus ,nebulosus7 #fe #bg4 ,flat ,alto-cumulus4 7,alto-cumulus ,stratiformis7 #fe #bh4 ,high ,turreted ,cloud4 7,alto-cumulus ,castellatus7 #ff #bi4 ,high ,ball ,cumulus4 7,alto-cumulus ,glomeratus7 #fg #cj4 ,mackerel ,sky4 7,alto-stratus ,maculosus7 #fh #ca4 ,mackerel ,sky4 7,alto-stratus ,maculosus7 #fi #cb4 ,alto-strato-cumulus #gj #cc4 ,sunset4 7,alto-cumulus ,castellatus ,fractus7 #gj #cd4 ,three ,layers of ,stratiform ,cloud after ,rain #gc #ce4 ,rain-,cloud4 7,nimbus7 #ge #cf4 ,rain-,cloud4 7,nimbus7 #ge #cg4 ,stratus ,communis #gg #ch4 ,strato-cumulus #gg #ci4 ,strato-cumulus #gh #dj4 ,stratus ,maculosus #gh #da4 ,common ,stratus4 7,stratus ,communis7 #gi #db4 ,roller ,cloud4 7,stratus ,radius7 #hj #dc4 ,small ,cumulus4 7,cumulus ,minor7 #id #dd4 ,cumulus #ie #de4 ,large ,cumulus4 7,cumulus ,major7 #if #df4 ,fracto-cumulus #ig #dg4 ,fall ,cloud4 7,stratus ,lenticularis7 #ih #dh4 ,thunder-clouds forming #aji #di4 ,thunder-clouds4 7,cumulo-nimbus7 #aaj #ej4 ,thunder-clouds4 7,cumulo-nimbus7 #aaa #ea4 ,thunder-cloud4 7,cumulo-nimbus7 #aaa #eb4 ,thunder-cloud4 7,cumulo-nimbus7 #aaa #ec4 ,the ,flank of a ,great ,storm #aab #ed4 ,crested ,alto ,waves4 7,alto-cumulus ,undatus7 #abj #ee4 ,alto ,waves4 7,alto-stratus ,undatus7 #aba #ef4 ,cirro ,ripples4 7,cirro-cumulus ,undatus7 #abb #eg4 ,waved ,cirro-stratus4 7,cirro-stratus ,undatus7 #acf #eh4 ,camera for measuring ,altitudes #ada #ei4 ,print from a ,negative used for measuring ,altitude #add #fj4 ,pair of ,prints showing the ,displacement of the ,cloud #ade #fa4 ,cloud ,camera for ,studies #aga ,,cloud ,,studies ,,chapter ,i ,,introductory ,,all who have the faculties proper to man must have been to some extent students of cloud form4 ,go where we will1 do what we will1 we cannot easily escape from the sky1 or avoid noticing some of its features and coupling them with the varying conditions of weather4 ,we all sometimes want to know if it is likely to rain1 or whether some other change is probable2 and experience soon shows us that the clouds give the simplest and most obvious indication of what we may expect4 ,it is almost impossible to avoid noticing that certain types of cloud1 or the simultaneous appearance of certain types1 is the usual accompaniment of definite kinds of weather or of particular changes4 ,thus it is that most people acquire some small measure of weather wisdom before their schooldays are over4 ,generation after generation1 through all human history1 the same causes must have led to the same conclusions2 and the study of clouds must1 therefore1 be one of the oldest of all branches of scientific inquiry4 ,yet1 old as it is1 it is still in its infancy1 having made very little advance indeed towards the precision of an exact science4 ,there are many reasons for this want of growth1 and so far as the theoretical aspects of the subject are concerned it is easy enough to understand4 ,clouds are among the most inaccessible of terrestrial objects4 ,except by balloon ascents1 by sending up kites bearing recording instruments1 or by making observations among the mountain-tops1 we have no means of getting at them to study the conditions under which they exist4 ,temperature1 pressure1 humidity1 have generally to be guessed at1 those guesses being based on the scanty data which have been laboriously obtained by one or another of these cumbrous methods4 ,moreover1 many clouds have such vast dimensions that it is very difficult to grasp all that goes on in such a space4 ,besides the difficulty of attacking the problems presented by cloud formation1 it is probable that even if we could have got among the clouds at will1 we should have understood little more than we do1 from a want of sufficient certainty on many of the purely physical questions involved4 ,it is not many years since ,mr4 ,j4 ,aitken discovered the necessity for material nuclei as a first step in the formation of cloud particles1 and not many months have elapsed since ,mr4 ,c4 ,t4 ,r4 ,wilson showed that those particles can be formed by the action of radiation on the air itself4 ,there is nothing surprising1 therefore1 in the fact that our theoretical knowledge of the why and wherefore of the facts revealed by a study of clouds is limited to general principles1 and quite fails to say exactly why each special form should be assumed4 ,the matter for surprise is quite different4 ,theoretical explanations are not the first step in the working out of a branch of science4 ,it begins with the acquisition1 by diligent and painstaking observation1 of a great mass of facts4 ,this may go on for centuries1 the accumulation growing greater and greater1 until at last some one comes who examines the records1 classifies them carefully1 and finally makes a summary in the form of a number of generalizations1 which are announced under the name of ,laws4 ,two examples of such 8,laws0 will suffice4 ,astronomers for centuries had observed the movements of the planets1 always with increasing accuracy1 until ,tycho ,brahe made his famous series of observations on the planet ,mars4 ,these materials fell into the hands of ,kepler1 and the result of his work was the announcement of ,kepler's ,laws1 which state the rules which govern the movements of the planets in their orbits4 ,he found that the records could not be accounted for unless the planets moved in a certain way1 but he knew nothing of the reasons for a method and order which clearly existed4 ,kepler's ,laws1 in fact1 rest upon another set1 namely1 ,newton's ,laws of ,gravitation1 and these are themselves a second example4 ,they are the summary of accumulated experience1 and even at the present day we know nothing certain as to why two bodies should attract each other1 and nothing as to why that mutual attraction should act as it was found to act by ,newton4 ,the observational part of cloud study1 however1 is still in its infancy1 in spite of the fact that it has been going on for such countless ages4 ,we are still in the condition of the humble observers engaged in the comparatively humdrum task of gathering facts for future arrangement and interpretation4 ,cloud observers1 in all ages1 have suffered from a peculiar difficulty4 ,they have had no common language1 no code of signs by which they could benefit from the work of those who had gone before them1 no means of transmitting their own experience to each other1 or to those who would come after them4 ,no progress would be possible in any study under such conditions1 for each person would begin where the previous generation began1 instead of taking up the task where others had left it4 ,in all languages there is an extraordinary scarcity of cloud names1 and such as do exist are frequently applied to quite different forms by different people4 ,so pronounced is this lack of terms1 that any one who tries to describe a sky without using any of the modern scientific names1 finds himself obliged to rely on long detailed descriptions1 backed with references to well-known objects1 whose outlines or structures resemble the clouds more or less vaguely2 and even then he has to be a word-painter of singular skill if his description calls up in the mind of the reader a picture much like the original4 ,it was to meet this want of a common tongue that ,luke ,howard1 in #ahjc1 proposed his scheme of cloud names4 ,he recognized three main types of cloud architecture1 which he named ,cirrus1 ,stratus1 and ,cumulus4 ,cirrus included all forms which are built up of delicate threads1 like the fibres in a fragment of wool2 ,stratus was applied to all clouds which lie in level sheets2 and ,cumulus was the lumpy form4 ,by combinations of these terms other clouds were described4 ,thus1 a quantity of cirrus arranged in a sheet was called cirro-stratus1 while high1 thin clouds like cirrus1 but made up of detached rounded balls1 was cirro-cumulus4 ,many cumulus clouds1 arranged in a sheet with little space between them1 became cumulo-stratus1 while the great clouds from which our heavy rains descend partake1 to some extent1 of all three types1 and were therefore distinguished by a special name--,nimbus4 ,this system had much to recommend it4 ,the three fundamental types were obvious to all4 ,their names were descriptive1 and were derived from a dead language1 so that no living international jealousies were raised4 ,it was sufficiently detailed to serve the purposes of the time1 when accurate observation was in its infancy4 ,hence it was universally adopted1 and will pretty certainly hold its own as the broad basis upon which any more detailed system must necessarily rest4 ,it has done excellent service2 but although observation of clouds in a general way is far from complete1 attention is now being given to much smaller details and much more minute differences of form1 and our vocabulary must be amplified4 ,precision of description is the first essential of a satisfactory system1 and the question is1 what sort of edifice can we build on ,luke ,howard's foundation4 ,the great difficulty is the infinite variety of clouds4 ,certain forms may be arbitrarily selected as types1 and names may be given to them2 but however well they are chosen1 a very short period of observation will show that there are all manner of intermediate forms1 which make a perfect gradation from one type to another4 ,this fact should never be forgotten4 ,there is always a danger that the use of any system of names based on types shall lead to the neglect of everything not typical4 ,a curious illustration is afforded by the well-known fact1 that in arranging collections of fossil shells1 it is frequently found that some specimens do not exactly match the type examples to which names have been assigned4 ,in former days it was the custom to throw aside such 8bad specimens10 as they did not show plainly the specific characters4 ,it is now realized that they have a value of their own1 in that they are the links in the evolutionary chain1 once supposed to be missing4 ,indeed1 it is not unfrequent nowadays to see carefully selected series1 showing the gradual change whereby one species passed into another1 displayed in the place of honour1 while the type specimens are relegated to humbler places in the general collection4 ,types there must be1 no doubt1 and where the series is continuous1 some one must make the selection4 ,with clouds the series is absolutely continuous4 ,the task is like choosing typical links from a long chain in which each link is almost exactly like its neighbours1 yet no two are alike1 and the greater the distance between them the less their likeness4 ,clearly any system put forward must be accompanied by illustrations1 so that all may know exactly which links have been chosen4 ,many attempts have been made to meet the want2 some of the systems proposed being based on the forms assumed by the clouds1 some on their supposed mode of origin1 and some on their altitudes4 ,those which were not founded on ,luke ,howard's types had no chance of being accepted1 while knowledge was not yet sufficiently far advanced to make classifications based on origin of form at all possible4 ,but the great reason why none of the proposed schemes could come into general use was that they were put forward without adequate illustration1 so that none but their authors knew exactly what they meant4#a ,matters came to a head in #ahia1 when an ,international ,meteorological ,conference met at ,munich4 ,one object of this gathering was to promote inquiries into the forms and motions of clouds1 by means of concerted observations at the various institutes and observatories of the globe4 ,luke ,howard's system was not enough for the purpose in view1 and the addition of more detailed terms had to be settled before work could be begun4 ,professor ,hildebrandsson1 of ,upsala1 and the ,hon4 ,ralph ,abercromby jointly submitted a revised scheme1 the main feature of which was the introduction of a new class of clouds1 to be distinguished by the prefix alto-before the other name4 ,such alto clouds were less lofty and denser than cirrus4 ,this scheme was the best before the ,conference1 and without waiting to discuss1 and possibly improve it1 it was formally adopted1 and a committee appointed to arrange and publish an atlas showing pictures of the type-forms4 ,this atlas did not appear until #ahif1 and in the mean time the ,rev4 ,w4 ,clement ,ley had published proposals of his own1 some of which had much to recommend them4 ,but he was too late4 ,the ,international ,committee had come to a decision1 and1 although it may be far from ideal1 the system backed by such an authority must be regarded as the standard until some similar gathering has given worldwide sanction to a change1 and even then it would be better to modify by addition rather than by substitution4 ,the subjects of the following pages are named in general accordance with this ,international ,code1 but they are by no means restricted to types4 ,their object is not to attempt any repetition of the work which has already been well done by the ,atlas ,committee1 but rather to show the chief varieties within a type4 ,it will1 however1 become abundantly evident that the standard system is far from complete1 and that any minute and detailed study of cloud-form must take note of the precise variety4 ,this at once raises the question whether many of these varieties are not sufficiently distinct to be given definite names4 ,if a meteorologist is told that cirrus clouds were seen on a particular occasion1 he instinctively asks--,what sort of cirrus8 and is utterly unable to form any mental picture of the clouds until the question has been answered by a detailed description4 ,a glance at a few of the plates further on will show the difficulty plainly1 and it occurs with other forms of cloud as well as cirrus4 ,is it not time that the ,international names were regarded as those of the cloud genera1 and to add specific names for those varieties which seem to imply some difference in kind in the conditions which have led to their formation8 ,this has been here attempted by translating into ,latin the ordinary ,english term by which the variety would naturally be described4 ,more extended observation will probably show that other species should be introduced1 and possibly some of those suggested in these pages may have to be subdivided4 ,whatever the names may be1 specific distinction of some sort is an essential preliminary to detailed study of the why and wherefore of the particular forms4 ,the ,international system is as follows3-- ,a4 ,upper clouds4 7a7 ,cirrus4 7b7 ,cirro-stratus4 ,b4 ,intermediate clouds4 7a7 ,cirro-cumulus and alto-cumulus4 7b7 ,alto-stratus4 ,c4 ,lower clouds4 7a7 ,strato-cumulus4 7b7 ,nimbus4 ,d4 ,clouds of diurnal ascending currents4 7a7 ,cumulus and cumulo-nimbus4 ,e4 ,high fogs4 7b7 ,stratus4 ,in this tabulation the forms marked 7a7 are detached and occur in dry weather1 while those marked 7b7 are widely extended4 ,the original scheme also gives the mean heights of the various types1 but these values have been omitted here because they are extremely variable1 and impossible to ascertain with any approach to accuracy by mere eye estimates4 ,they vary also with the season1 and probably also with the locality4 ,moreover1 the altitude is no guide to the name1 except that on the whole the types occur in the order given1 taking group ,a as the highest and group ,e as the lowest4 ,in the chapter on cloud altitudes this subject will be further considered1 and under the descriptions of cloud-forms their average height or actual measurements for the particular specimen figured will be given whenever possible4 ,before coming to the description of individual forms1 it may not be out of place to give brief consideration to the best means of observing them in nature4 ,for eye observation1 of course1 no directions are needed when we are dealing with the lower and denser varieties2 but when we come to the highest groups it sometimes becomes necessary to protect the eye from the brilliant glare which may make it impossible to detect the real structure4 ,smoked glass1 neutral-tinted spectacles1 or yellow glass all have something to recommend them2 but by far the most convenient means is to look1 not at the clouds themselves1 but at their images formed in a black mirror4 ,a lantern cover glass1 or a thin piece of plate-glass1 blacked on the back with some black paint1 serves admirably4 ,but all black paints are not equally good4 ,the best are oil paints which dry with a glossy surface1 the so-called enamels4 ,they have the advantage that the varnish with which they are mixed has an index of refraction not very different from that of the glass4 ,the consequence is that so little light is reflected from the blackened back1 compared with that which is reflected from the front surface of the glass1 that the second image can only be detected with difficulty4 ,if the mirror is a piece of black or deeply coloured glass all trace of the second image is lost4 ,with this simple appliance it is easy to study the details of the thinnest clouds right up to the sun1 and even the image of the sun itself may be glanced at without serious discomfort4 ,nor is the general diminution of brightness the only gain4 ,if the glass is so held that the light from the cloud makes an angle of about #cc degrees with the surface1 some of the blue light from the sky is suppressed altogether1 while that from the cloud is practically unaffected4 ,the exact fraction suppressed depends upon the part of the sky relative to the sun1 and also on the position of the mirror1 but a few minutes' trial will show when the maximum effect has been reached4 ,it is astonishing to see for the first time how the delicate filaments of cirrus or the beautiful structures of cirro-cumulus stand out shining white on the deep blue background2 and the use of the black mirror is a revelation to most4 ,it also has one indirect advantage1 which is really more important than it seems4 ,by gazing down into a mirror long-continued observations can be made1 and one form of cloud may be watched changing into another1 and possibly back again into its original shape1 without any danger of incurring that unpleasant result of much looking upwards which is sometimes known as exhibition headache4 ,such a mirror may be quite small1 so that it can be carried in a pocket-book1 a point of some moment1 as many of the forms of cirrus are exceedingly transient1 coming and going in a few minutes1 while others are in a state of continuous change4 ,this is particularly often the case with the exquisite ripple clouds1 and the delicate lacework of the higher kinds of cirrus4 ,still another advantage possessed by the mirror is that it makes it easy to see the solar halos formed on the verge of a cyclone1 and to detect their iridescent colouring in a way which is quite beyond the reach of the naked eye or any protective spectacles4 ,every one is familiar with the faint halos formed round the moon1 but the corresponding solar phenomenon is comparatively little known1 though it is far commoner1 much more brilliant1 and often glows with colour4 ,its very brightness1 and that of the background on which it is projected1 hides it from the eye1 except on those rare occasions when the sun is conveniently hidden by some thicker cloud4 ,if some permanent record is desired1 much can be done with a few light strokes of a pencil1 but more ambitious pictures are best secured by the use of soft pastels1 aided by a liberal use of the finger or leather stump4 ,ordinary paints1 whether oil or water-colour1 are of little use for actual study of cloud detail1 except in the hands of a highly skilled artist who knows how to get the effect he wants in the minimum of time4 ,but no sketching or drawing can make records of cirrus or alto clouds with the speed and accuracy necessary for careful study4 ,photography is really the only way in which the amazing wealth of detail can be truthfully portrayed4 ,yet even the camera has its limitations4 ,it does not record colour1 and completely fails to delineate the forms of alto-stratus1 stratus1 or nimbus1 if they are present in the most typical condition1 that is to say1 when they cover the whole sky with a uniform tint4 ,it is only when these forms are more or less broken up that a photograph1 or anything other than a carefully coloured picture1 will represent them at all4 ,cloud photography1 even of the most delicate and brilliant varieties1 is easy enough when the right methods are followed2 but these are not the same as those which are right for portraiture or landscape work of the usual kind4 ,the background of blue sky produces almost the same effect on the plate as the image of the cloud itself1 and the whole art consists in an adequate exaggeration of the minute difference so as to reveal the details of form and structure4 ,a slow plate--the accompanying illustrations have all been taken on ,mawson and ,swan's photo-mechanical plates--extremely cautious development1 and sometimes intensification of the image1 are all that is necessary2 but the process becomes easier if1 instead of pointing the camera to the cloud1 it is directed to the image formed in a properly constructed black mirror4 ,many of the following studies have been taken by this method1 and details of the camera and processes employed will be found in a later chapter1 for the convenience of any one who may be inspired to take up a fascinating branch of photography4 ,it has been said that reference will be made to the average altitudes of the different types of cloud1 and to the actual altitude of some of the varieties shown4 ,the question will1 no doubt1 have occurred to some as to how those altitudes have been measured4 ,the methods are all more or less complicated1 involving rather laborious calculation4 ,they generally depend upon simultaneous observations made from two stations at opposite ends of a measured base line4 ,sometimes the observations are made directly by pointing an instrument at each station to some agreed point of the cloud4 ,it is obvious that the two directions must converge to this point4 ,if the convergence is measured1 the exact distance from either station can be calculated1 and if the angle between the cloud-point and the horizon beneath it is noted1 it is a simple matter to deduce the actual altitude of the cloud4 ,at other places the observers have relied upon the comparison of photographs simultaneously taken from the two stations4 ,in this method it is necessary to know the exact direction in which the camera is pointed1 and the position of the image upon the plate then gives the direction of the cloud as seen from that particular station1 and the subsequent calculations are the same4 ,measurements by one or the other of the above methods have been made at several places1 the most extensive series being those which have been compiled at ,upsala1 and at the ,blue ,hill ,observatory in ,massachusetts4 ,the method employed by the writer at ,exeter has been rather different1 and a description will be found later on in the chapter on ,cloud ,altitudes1 the fuller consideration of which comes naturally after the different forms have been described and compared4 ,,footnote3 #a ,see reference ,no4 #b on p4 #aha4 ,,chapter ,,ii ,,cirrus ,a ,,cloud is sometimes defined as any visible mass composed of small particles of ice or water suspended in the air1 and formed by condensation from the state of vapour4 ,as a general rule this is exact enough1 but under certain circumstances it is possible to have the particles so small1 and so thinly scattered1 that it is not fully satisfied4 ,the resulting mass may not be actually visible4 ,the presence of the condensed particles may be indicated by nothing more than a slight whitening of the blue sky1 or by the formation about the sun or moon of bright circles of light known as halos4 ,if such a halo appears1 it is generally a phenomenon of brief duration4 ,sometimes the circle breaks and becomes incomplete by the passing away of the thin patch of cloud1 sometimes the cloud increases in density until the rings are destroyed4 ,the thinnest variety of this halo-producing structure is quite invisible to the eye4 ,it is so thin as to have no distinctly noticeable effect upon the colour of the sky1 but the optical results of its presence may be very remarkable4 ,highly complicated systems of rings are sometimes produced1 the rings1 as a rule1 falling into two groups4 ,the commonest form has the sun 7or moon7 in the centre1 and a circle of pale light at a distance of about #bb degrees4 ,larger rings are seen less frequently1 which have an angular radius of about #df degrees1 and as a rule have the sun situated on the ring itself4 ,in ,plate #a we have a part of such a great halo4 ,the camera was directed towards the east1 and tilted upwards at an angle of about #dj degrees4 ,the sun was behind the camera1 in the south-west1 and the ring could be traced right up to it on either side4 ,at the same time the sun was surrounded partially by a halo of the more ordinary type1 which was brightly coloured1 making an effective contrast to the dull white of the greater ring4 ,the phenomenon did not last more than half an hour1 and the changes in its appearance coincided with a growing density of cloud4 ,when first noticed the great ring was alone1 and the sky was of a full blue1 but a silvery film came gradually up from the south-west1 and the smaller and brighter halo flashed out as the delicate curtain came near the sun4 ,slowly the cloud spread to the north-east1 gathering density from the opposite point of the compass2 and by the time the ordinary halo was at its best1 the great white ring had completely vanished4 ,illustration3 ,plate #a4 ,,part ,,of ,a ,,great ,,halo4 ,illustration3 ,plate #b4 ,,part ,,of ,a ,,solar ,,halo4 ,these circles1 and the bright spots called mock-suns or mock-moons which often accompany them1 can all be explained on the assumption that their cause is the passage of light through a veil composed of hexagonal crystals of ice4 ,the simple halo of #bb degrees radius is common in most parts of the world1 being very generally formed by the film of high cloud which marks the advancing edge of a cyclonic cloud system4 ,a portion of one is shown in ,plate #b1 in which the rudimentary fibrous structure of the sheet of cloud is distinctly seen4 ,halos of this sort are frequently coloured1 often most brilliantly so2 but the tints are seldom noticed unless a black mirror is used4 ,they are sometimes quite as bright as those of an ordinary rainbow1 but instead of being projected upon a background of dark rain-clouds1 they are seen against a part of the sky which is near the sun1 and therefore exceptionally bright4 ,the red is always on the inside of the ring1 the violet outside1 thereby distinguishing them at once from the so-called coron1 which are formed around the sun or moon when shining through a sheet of alto or other lower cloud made up of liquid particles4 ,in these the radius of the rings is much less1 and the red is on the outside1 the violet actually touching the central luminary4 ,the cloud which produces halos is called cirro-nebula4 ,it is much thinner1 and on an average higher than cirro-stratus4 ,mr4 ,ley named it cirro-velum 7or cirro-veil71 but cirro-nebula has now got to be fairly well understood4 ,it sometimes appears and disappears in a curious manner1 showing that it occurs in patches1 which drift about or which keep forming and melting away1 only to repeat the process4 ,if1 however1 it forms part of an advancing cyclone fringe1 then the sky gets whiter and whiter1 until it is covered with a sheet of undoubted cirro-stratus4 ,this process of growing density is shown in progress in ,plate #c4 ,illustration3 ,plate #c4 ,,cirro-,,nebula ,,changing ,,to ,,cirro-,,stratus4 ,cirro-nebula1 as we shall call it1 floats at very great altitudes in temperate regions2 but in polar latitudes1 where the optical phenomena peculiar to it are most brilliant and diversified1 it seems probable that the ice dust is much lower down1 even in actual proximity to the ground4 ,in ,england its height varies greatly with the time of year1 and other circumstances1 but mounts up in summer to such altitudes as nine miles or more2 the greatest height yet recorded being #if miles1 or about #ae1ejj metres1 at ,exeter4 ,the change from cirro-nebula to cirro-stratus is generally accompanied by the formation of a distinct fibrous structure1 easily observable by the black mirror4 ,this is not really a new feature1 but only a further development of a structure already existing1 but too minute to be easily seen4 ,true halo-producing cirro-nebula may usually be shown to possess more or less of a fibrous texture in an indirect way1 which is worth a brief description4 ,in order to observe the spots on the sun and other features of the solar surface1 it is a common practice to hold a white screen1 say1 about a foot from the eyepiece of a telescope1 while the instrument is pointed to the sun4 ,an image1 considerably magnified1 is thus projected on to the screen1 and the solar details can be studied with ease and safety4 ,if thin clouds drift before the sun1 their images are similarly projected as they pass across its disc1 and it is possible thus to detect not only the fibrous texture but also the movement of cirro-nebula4#b ,the change into cirro-stratus is also attended by a marked fall in altitude1 but whether this is due to an actual descent of the cloud particles1 or to a downward spread of the conditions which give rise to them1 cannot at present be definitely settled4 ,the balance of probability points very strongly towards the downward spread of the conditions4 ,it is likely that the clouds1 particularly the cyclonic specimens1 are wedge-shaped1 and that as they pass overhead we see first the thin edge1 and later on the thicker parts1 which project much lower down4 ,this is just one of those many minor problems in cloud mechanics which we are not able to solve from the scanty data on record4 ,occasionally cirro-nebula breaks up into little detached semi-transparent cloudlets1 all of them exceedingly thin1 and showing a complicated mottling1 resembling1 on a minute scale1 the ripple clouds of much lower altitudes4 ,such a sky is depicted in ,plate #d1 but no reproduction can possibly do justice to the minute and delicate features of the real thing4 ,the arrangement of the faint markings was in a state of continual flux1 curiously similar to the ever-changing aspect of the sun's photosphere when seen under adequate power4 ,some parts of the cloud stratum would at one moment break up into distinct granules arranged in complicated patterns1 other parts would assume a fibrous texture1 and yet other places would show a continuous smooth sheet4 ,in a minute or two all would be changed--the smooth part granulated1 the fibres vanished1 and the granules fused together1 and so on1 no two of a series of photographs representing the same details4 ,illustration3 ,plate #d4 ,,cirro-,,nebula ,,changing ,,to ,,cirro-,,cumulus4 ,these changes of form continued until the whole was hidden from view by a veil of much lower stratiform cloud1 one advance portion of which is shown4 ,plate #d does not represent a type or a distinct variety of cloud4 ,it is an intermediate form1 or a temporary condition1 showing cirro-nebula in the act of changing into cirro-cumulus1 or possibly cirro-stratus4 ,cirro-nebula itself1 in its simpler form1 is1 however1 a distinct type4 ,it is true that it never persists over one locality for more than an hour or two without passing into some denser form1 but while it lasts its features are so distinctive1 and the optical phenomena to which it gives rise are so striking and significant1 that it is a matter for surprise that it should in the ,international system have been relegated to the position of a subordinate variety of cirrus4 ,it is more nearly related to cirro-stratus1 but is sufficiently distinct from that to deserve at least specific rank4 ,true typical cirrus must have a plainly shown fibrous structure4 ,the fibres may cross and interlace1 they may radiate in fan-like manner1 or they may curl and twist like a well-trimmed ostrich feather4 ,the clouds so formed must not be arranged in a continuous level sheet1 or they at once become cirro-stratus1 and it is impossible to invent a definition which will mark the exact limits of either type4 ,typical cirrus consists of detached clouds4 ,they cast no shadows on the landscape1 for the simple reasons that they are semi-transparent and their component parts too narrow4 ,if the sun is shining down obliquely through the naked boughs of a tall tree1 it will be seen that the lowest twigs cast fairly sharp shadows on the ground1 but that even these are bordered by a fading rim2 the twigs further up cast no sharp shadows1 but broader faint bands of shade2 while the topmost boughs cast no shadows which can clearly be identified4 ,in other words1 the more distant the narrow twig is from the ground the narrower the real shadow or umbra1 and the broader the penumbra becomes1 until when the distance is sufficient the shadow is all penumbra4 ,cirrus filaments throw nothing but a faint penumbra4 ,indeed1 it is only when they lie in the earth's shadow1 and stand against the background of a faintly lighted sky1 that they show any sign of shadow even on themselves4 ,there is no doubt that they are composed of particles of ice4 ,they are formed at altitudes where the thermometer must be many degrees below freezing-point1 and not a few of the thinner examples show fragmentary halos like those of cirro-nebula4 ,their actual altitudes are very variable1 being greater in summer than in winter1 and reaching a maximum for any given station after a long spell of hot weather4 ,exact measurements have not yet been made in tropical latitudes or in polar regions1 but there is every reason to expect that the upper limit of cirrus for equatorial districts will be found to be much higher than in the temperate zones where actual observations have been made4 ,in places nearer to the ,arctic ,circle it is also almost certain that the altitudes will be less4 ,in the ,new ,england states1 as shown by the ,blue ,hill observations1 the maximum altitude for summer was found to be little under #ae1jjj metres4 ,at ,upsala1 in ,sweden1 it was #ac1cjj metres4 ,the average altitudes at the same observatories were1 respectively1 about #iijj and #hhjj metres4 ,at ,exeter the writer's own measurements give an average for the summer months of #aj1bjj metres1 with a minimum rather lower than was the case in ,america or ,sweden1 and with a maximum far above the foreign values4 ,in winter cirrus certainly comes much lower down1 but the number of observations is fewer4 ,illustration3 ,plate #e4 ,,high ,,cirrus4 7,cirrus ,excelsus47 ,the loftiest variety of cirrus appears in the afternoon in very hot weather1 sometimes quite late in the evening2 and in autumn it is by no means a rare event for it to suddenly form just when the sunset colours are fading1 or even after they have paled into twilight4 ,under such circumstances it stands out of a shining silvery grey colour against the background of the darkening sky4 ,a specimen of it is shown in ,plate #e1 which shows its extreme slightness of form and delicacy of texture4 ,sometimes it remains visible so long after the stars have begun to show as to give the idea that it is self luminous1 and the illusion is certainly very strong4 ,the writer has noted several instances in which it was plainly visible1 like a silvery curtain1 though the sky as a whole was so dark that stars like the five brightest points of the ,great ,bear could be seen through the cloud1 and much smaller stars down to the third and fourth magnitude were plainly visible in the clear intervals4 ,it has sometimes been called luminous cloud1 and ,mr4 ,ley estimated its altitude at upwards of #ij1jjj metres2 but if we think of it as reflecting the light of the distant colourless twilight there is no need to regard it as anything fundamentally different from other clouds1 or to assume a greater altitude than we know to have been the case4 ,the specimen figured occurred in the early afternoon on ,june #ab1 #ahii1 at ,exeter1 and careful measurements of its altitude were made4 ,this worked out as #agjb miles1 or more than #bg1jjj metres1 a value so much greater than all other measurements of the kind that it was only after most careful verification and reference to duplicate records that it could be accepted4 ,it differs in several ways from the lower varieties1 being thinner1 more glistening1 and in every way more delicate4 ,a suitable distinctive name would be high cirrus1 or cirrus excelsus4 ,illustration3 ,plate #f4 ,,windy ,,cirrus4 7,cirrus ,ventosus47 ,lower down by thousands of metres come the feathery masses of typical windy cirrus1 such as are shown in ,plate #f4 ,indeed1 in cold winter weather they occur within three or four thousand metres of the ground4 ,in the instance figured the wind was blowing from left to right1 and the clouds were travelling swiftly4 ,the upper filaments appeared to be repeatedly torn away from the main masses1 while the long faint streaks which trail below and behind are evidently due to streams of fine particles falling from the main centres of condensation into a less rapidly moving stratum below4 ,there is no room for doubt that these clouds1 like others of a similar order1 are formed by a direct passage from the vapour to the solid1 or that the fibres are made of minute snowflakes4 ,the condensation is evidently attended by rapid movements1 which draw out the cloud1 as fast as it is formed1 into long curving lines which mark lines of motion4 ,the variety is always1 therefore1 an indication of strong winds and rapid eddying movements in the region in which it occurs4 ,such strong disturbances overhead almost always accompany similar but less intense movements at the ground-level1 and when they do not accompany them they precede them4 ,the cloud is well named windy cirrus1 which may be converted into a specific name1 cirrus ventosus4 ,the next variety we come to 7,plate #g7 is in some ways rather similar4 ,it is1 however1 thinner1 more delicate1 and is entirely composed of fine threads1 which are more systematically arranged4 ,generally there is a bundle1 or several bundles1 of long parallel fibres1 which form1 so to say1 the quill of the feather1 with numbers of shorter threads branching out from them at various angles4 ,cirrus ventosus was indicative of irregular movements in various directions2 this variety points also to complicated movements1 but executed in accordance with some sort of system1 strangely complex and wonderfully ordered4 ,the specimen figured is the type of what ,mr4 ,ley called cirro-filum1 or thread cirrus1 and his name can hardly be bettered4 ,it is a cloud of summer1 and occurs rather high up in the cirrus zone1 but no actual measurements can be quoted4 ,it is fairly common1 but not nearly so frequent as the last4 ,illustration3 ,plate #g4 ,,thread ,,cirrus4 7,cirro-filum47 ,a somewhat more familiar variety is shown in ,plate #h4 ,little irregular feathers of cirrus1 from which long tapering streamers point downwards in graceful curves1 or else lag behind in the direction from which the clouds have travelled4 ,if clouds of this type are carefully watched1 it will soon be seen that each feathery head is a centre of condensation1 and that the tails or streamers are nothing else than falling particles1 which dwindle slowly away by evaporation1 and which gradually sink below the level of the heads4 ,it is usual1 in dealing with cloud-forms like these1 to speak of air-currents of different velocities almost as if the winds at different levels were as clearly separated as oil and water1 or even air and water4 ,this can hardly be the case1 for if such a thing should occur as an air-current of one velocity flowing over another of less speed1 or of a current in one direction over another moving in a different course1 the two must inevitably mix at their junction1 and in a very short time the passage from the lower current to the upper one would be quite gradual4 ,no doubt we can often observe two1 three1 or more layers of cloud moving in different directions2 but if we were to send up a balloon1 it would be rare indeed to find its direction of horizontal movement changed in a few metres of ascent4 ,different and distinct air-currents are often invoked to explain cloud-forms quite unnecessarily4 ,it is far more likely that the differential movements involved in the explanation of the features of these cirrus varieties are due to increased velocity with greater altitude1 to progressive change of direction1 to irregular eddies1 or to the interaction of ascending and descending convection currents4 ,indeed1 it is probable that careful study of the growth and decay of these clouds will1 in time1 lead to a clearer understanding of atmospheric movements1 and so enable us to say more precisely why they are as we see them to be4 ,the variety shown in ,plate #h is rare except in combination with other forms4 ,it might well be termed tailed cirrus or cirrus caudatus4 ,illustration3 ,plate #h4 ,,tailed ,,cirrus4 7,cirrus ,caudatus47 ,the form of cirrus shown in ,plate #i is far more frequently seen than either of those which have been described4 ,in this the fibrous texture is very imperfect1 and the cloudlets show a tendency to arrange themselves in a kind of ribbed structure in two directions almost at right angles to each other4 ,but this last is an accidental feature of the particular example1 and not in any way a specific character of the cloud4 ,the reason for regarding it as a distinct variety is the total absence of sharply defined lines1 not only the heads of condensation1 but even the long streamers attached to them being uniformly hazy and ill-defined4 ,it is a form of cirrus which comes at all seasons1 but most frequently in summer2 it moves always with great slowness1 indicating a quiet atmosphere free from disturbance of any kind4 ,the conditions necessary for its appearance are a nearly uniform distribution of pressure over a considerable area1 chequered by little shallow depressions of some trifling fraction of an inch4 ,in hot weather these are the conditions under which thunder-storms develop1 and this hazy cirrus1 or cirrus nebulosus1 may be taken as a certain sign of such an atmospheric state4 ,illustration3 ,plate #i4 ,,hazy ,,cirrus4 7,cirrus ,nebulosus47 ,so far as permanency of form is considered1 hazy cirrus is one of the most persistent1 and affords a marked contrast to the species shown in ,plate #aj1 which represents the most fugitive4 ,five minutes before the photograph was taken the same part of the sky was a deep1 clear blue1 without any trace of cloud4 ,suddenly a few short curling wisps made their appearance4 ,these rapidly increased in number1 until a delicate filmy network extended over the greater part of the field of view4 ,but while the camera was being adjusted for an exposure1 part of the net had broken up into the granular structure shown in the lower part of the photograph4 ,the granulation rapidly spread through the net1 almost as if the fibres had been curdled1 and five minutes later the whole had been converted into a patch of cirro-cumulus which soon fused into a uniform sheet4 ,meanwhile the same series of phenomena were taking place in other parts of the sky4 ,illustration3 ,plate #aj4 ,,change ,,cirrus4 7,cirrus ,inconstans47 ,on other occasions exactly the same set of events have been seen to follow each other in the inverse order4 ,beginning with a fairly even sheet1 this broke up into granules1 and they in turn seemed to be frayed out into short hazy and wavy fibres which slowly melted away4 ,clearly we have here to do1 not with a distinct type of cloud1 but rather with the first step towards the formation of one1 or the last stage in the life of one which is drying up4 ,but sometimes the life of the cloud is so short that it never passes beyond this first stage2 and it is by no means a universal rule for a growing sheet of cirrus to pass through this stage at all4 ,it therefore represents a peculiar state of instability1 and requires a name of its own4 ,sometimes patches of it will come and go in an apparently capricious manner for an hour or more before permanent condensation is effected or before the sky finally clears4 ,but this is a rare event1 since the slow change of conditions which has brought the stratum of air to the unstable condition is generally progressive1 and instead of stopping at the critical point1 goes beyond it1 with the result that the condensation grows or the cloud disappears entirely4 ,change cirrus1 or cirrus inconstans1 would be an appropriate name for a kind of cloud which is so plainly indicative of instability4 ,the critical condition referred to is1 of course1 that in which a particular stratum of air is just saturated1 or is just on the point of forming visible cloud4 ,if any cause is brought to bear on such a stratum which brings about even slight cooling1 cloud must be produced2 and1 conversely1 anything which results in the slightest heating must cause it to disappear4 ,the shortness and haziness of the fibres1 and the fact that they gather themselves into granules1 shows that the cloud is formed in a stratum of air which is either still1 or is moving as a whole1 without any of those differential movements which seem to be necessary for the longer fibrous details4 ,the causes which may bring about the local cooling and heating are easy to understand when we remember how the air will be affected by the uneven contours of the ground4 ,as it passes over hill and valley the up-and-down movements of the lower layers1 or even the disturbances caused by passing over a wood or clump of trees1 all must be propagated upwards4 ,each disturbance must slowly spread laterally and diminish vertically1 so that it will reach the cirrus zone as a broad and gentle dome-like oscillation4 ,suppose now a series of such slight upheavals to reach the critical level4 ,the passage of the waves will mean alternate expansion and compression4 ,expansion means cooling1 and therefore cloud-production2 compression means heating1 and therefore the destruction of cloud4 ,from the most transient form of cirrus we pass1 in ,plate #aa1 to the most persistent and probably the most frequent4 ,it occurs in detached masses which have very variable forms but are wholly fibrous1 with the details arranged in a very irregular manner4 ,the example figured was taken in the evening during a long spell of fine weather4 ,if such a cloud is watched1 its permanence of detail is very striking1 and must be due to a persistence of slow eddying movements and to a continual renewal and waste of the component particles of each wisp4 ,this is the kind of cirrus selected generally as the type of cirrus1 and the selection is a good one4 ,common cirrus1 or cirrus communis1 it should be called4 ,settled conditions and fine weather are its usual attendants4 ,illustration3 ,plate #aa4 ,,common ,,cirrus4 7,cirrus ,communis47 ,we next come to a variety which is anything but a harbinger of good1 namely1 the long stripes or bands of cirrus which stretch outwards from the margin of the cloud canopy of a cyclonic storm4 ,in some ways these appendages to the great nimbus resemble the strips of cirriform cloud which fringe the summit of a thunder-cloud4 ,they look as if they must have been formed by the blowing away1 by a rapid wind1 of the top of an uprising column of vapour-charged air4 ,their main outline may thus be easily accounted for1 but we have only to study their detailed structure for a few minutes to feel that they really present a problem of a very high order4 ,plate #ab shows a fairly simple example1 but ,plate #ac represents a cloud of very great complexity4 ,to take this last the camera was tilted upwards at an angle of #de degrees1 so that the top of the picture is not far from the zenith4 ,the wonderful plume of cloud rose from the southern horizon1 and ended in a great sheaf of fibres and films spread out like a partly opened fan whose edge was only about #ej degrees above the northern horizon4 ,its length as it passed overhead lay between a point a little east of south to a little west of north2 and the broad band moved as a whole1 without any marked internal changes1 from the south-west towards the north-east4 ,the weather was very unsettled4 ,a long procession of cyclones had been sweeping along our western shores1 and the barometer was just beginning a fresh and rapid fall4 ,during the ensuing night a heavy gale burst over the south of ,england4 ,illustration3 ,plate #ab4 ,,band ,,cirrus4 7,cirrus ,vittatus47 ,illustration3 ,plate #ac4 ,,band ,,cirrus4 7,cirrus ,vittatus47 ,the whole phenomenon was highly characteristic4 ,these great bands with the divergent striation might well be known as storm bands1 from their almost invariable connection with the violent atmospheric movements to which they are most probably due4 ,plate #ab shows a much less dangerous variety of the same species1 which is distinguished from it by the comparative absence of internal detail and by the curled ends4 ,clouds of this character have sometimes been called cirro-filum1 but a comparison of the plates with the typical cirro-filum of ,plate #g will show that there is little resemblance2 and the attendant weather is also in marked contrast1 both of which facts imply a fundamental difference in the causes to which their features are due4 ,banded or ribboned cirrus is the name which they immediately suggest1 and this may be rendered cirrus vittatus4 ,this ends our survey of cirrus clouds4 ,any one who compares the plates so far given will see that they represent forms so diverse that it is impossible to avoid the conclusion that the conditions under which they are produced must differ not only in degree but also in kind4 ,what those conditions are we have attempted here and there to suggest1 but in no case can we feel that the explanation has been at all complete4 ,in some cases1 notably the last1 we are face to face with such complicated details that it is hopeless to attempt to explain them in the present state of our knowledge4 ,fact upon fact must be accumulated until we can give their history from their earliest beginnings2 and far more accurate and detailed knowledge of the attendant atmospheric conditions must be acquired before we can hope to rob such elaborate structures of their present mystery4 ,this requires the co-operation of many eyes and many minds1 and exact specific names must be an essential preliminary4 ,those which have been suggested in these pages are-- #a4 ,cirro-nebula1 or ,cirrus haze4 #b4 ,cirrus excelsus1 or ,high cirrus4 #c4 ,cirrus ventosus1 or ,windy cirrus4 #d4 ,cirro-filum1 or ,thread cirrus4 #e4 ,cirrus nebulosus1 or ,hazy cirrus4 #f4 ,cirrus caudatus1 or ,tailed cirrus4 #g4 ,cirrus vittatus1 or ,band cirrus4 #h4 ,cirrus inconstans1 or ,change cirrus4 #i4 ,cirrus communis1 or ,common cirrus4 ,,footnote3 #b ,the telescope with which these observations have been made is a #fh-;inch refractor equatorially mounted4 ,,chapter ,,iii ,,cirro-,,stratus ,,and ,,cirro-,,cumulus ,,several of the varieties of cirrus already discussed may gather so abundantly at some given level in the atmosphere1 that the most obvious feature comes to be this arrangement in a sheet4 ,the cloud then becomes cirro-stratus1 and should be so named4 ,we have described how cirro-nebula frequently grows in density until it fails to produce halo phenomena1 and may even reduce the sun to a hazy patch of light showing no outline4 ,this is the most typical of all forms of cirro-stratus4 ,it has always a distinctly fibrous or streaky appearance1 whereby it is at once distinguished from a similar but lower cloud which will be described later on4 ,a similar sheet may be formed from the fusion together of the streaks of cirrus-nebulosus1 the bands of cirrus vittatus1 or the development of cirrus inconstans4 ,but the general rule is that the cirro-stratus retains more or less of the specific characters of the parent form4 ,plate #ad shows a hazy form of cirro-stratus developed from the nebulous cirrus4 ,its altitude was great1 being about #aj1jjj metres4 ,the processes of growth and change could be studied easily4 ,first would appear some faint spots and streaks2 these quickly fused together into larger patches1 which again joined to their neighbours4 ,in a few minutes the cloud so formed would return to the mottled or streaky appearance1 and either disappear entirely or become very thin1 only to recommence the process4 ,this went on for more than an hour1 the cloudy patches getting larger and larger1 until the critical condition was passed1 and the sky was covered with a general veil of typical cirro-stratus4 ,illustration3 ,plate #ad4 ,,hazy ,,cirro-,,stratus4 7,cirro-stratus ,nebulosus47 ,in ,plate #ae we have an example of a cloud which is clearly cirro-stratus1 but the sheet is broken up into long bands1 and each of these is made up of common cirrus4 ,in the upper part of the picture the sprays of cirrus are forming1 and as they come into being they are arranged in rows4 ,we have here to do with a phenomenon of a very different order from the one presented by the true banded cirrus4 ,the arrangement into belts must here be due to some kind of wave-movement in the air1 breaking up the critical plane into long ribs transverse to the direction of the wave-movement4 ,the specimen shown was moving in a direction nearly at right angles to the bands1 though the surface wind was nearly parallel to their length4 ,the question at once presented itself as to whether the movement of the bands was really a drift of the cloud1 or whether it was not a case of the propagation of cloud production with the advancing wave4 ,this was easily answered by watching the details of a band4 ,the advancing side was always feathery1 and careful observation showed that the edge advanced by throwing out new threads and curls4 ,a given thread or curl1 one moment at the edge1 would in a few minutes be well in the band4 ,quite opposite events were taking place in the rear of each band4 ,the cloud was there obviously melting away4 ,indeed1 to sum the matter up1 the cloud bands flowed past their details just as the waves on the sea flow past the floating foam4 ,evidently we have in ,plate #ae the result of a plane of commencing cirrus formation broken into a series of troughs and waves by an undulatory movement of the air4 ,but1 as we have already said when speaking of cirrus inconstans1 the condition in which trifling up-and-down movements can determine whether condensation shall1 or shall not1 take place seldom lasts long4 ,it is usually only a stage in a continuous change1 and in this particular instance the banded structure was soon replaced by a fairly continuous sheet of typical cirro-stratus4 ,illustration3 ,plate #ae4 ,,cirro-,,stratus4 ,the next plate1 ,no4 #af1 shows a similar process4 ,in the upper part we have cirrus inconstans forming in patches out of a deep clear-blue sky4 ,its hazy fibres grow closer and closer1 betraying a slight tendency to gather in narrow ripple-like bands1 but the structure is soon lost in the uniform white sheet of interlacing fibres1 which differ from common cirrus in little else than their number and closeness4 ,nevertheless1 the stratiform arrangement is quite obvious enough to warrant the use of the term 8cirro-stratus40 ,illustration3 ,plate #af4 ,,cirro-,,stratus4 7,cirro-stratus ,communis47 ,the change of cirro-nebula into cirro-stratus is shown in ,plate #d1 to which reference has already been made4 ,the structures are remarkably delicate1 showing in the middle a distinct irregular mottling2 and rather further towards the top right-hand corner a ripple structure appears1 and in the top left-hand corner the sheet is denser and whiter4 ,the altitude of this cloud was evidently great1 and actual measurement showed it to be #gf miles4 ,it did not last long1 and after its change into broken patches of denser cirro-stratus1 still higher clouds were revealed through the gaps4 ,cirro-stratus often forms almost simultaneously at more than one level1 and when that happens the full stratiform appearance is generally reached first by the lower layer4 ,in ,plate #ag we have two layers4 ,the fluffy bits of cirrus nebulosus1 in the lower part of the picture1 are really the higher clouds4 ,below them1 probably by many thousands of feet1 floats the denser cloud shown in the upper part of the picture4 ,this is an interesting link between the fibrous and the granular forms of cirrus1 and is probably best described as spotted cirro-stratus1 or cirro-stratus maculosus4 ,it is a form very frequently met with1 but seldom showing any persistence4 ,it is indicative of condensation in a calm atmosphere1 and not unfrequently marks either the small irregularities of pressure which form the conditions for thunderstorms1 or the beginning of the break up of an anticyclone4 ,illustration3 ,plate #ag4 ,,flocculent ,,cirro-,,stratus4 7,cirro-stratus ,cumulosus47 ,a coarser texture and greater density are shown in ,plate #ah1 where we have cirro-stratus in the lower part of the picture1 and cirro-cumulus in the upper4 ,the altitude of this cloud was only about #djjj metres1 one of the least values recorded at ,exeter for cirro-stratus of any kind4 ,the intimate admixture of the fibrous and granular forms is very clearly shown4 ,illustration3 ,plate #ah4 ,,cirro-,,stratus ,,and ,,cirro-,,cumulus4 ,this close relation is equally obvious in ,plate #ai1 where the cloudlets are arranged in loosely marshalled rows1 dimly resembling the banded structure of ,plate #ae4 ,but in this case the direction of movement was with the long lines1 and the propagation of cloud production followed the same course4 ,some of the little cloudlets have an opacity1 and therefore brilliancy1 quite unusual for cirro-cumulus1 but their intimate association evidently in the same plane with undoubted cirrus shows that they must fall under that general description4 ,it is a cloud indicative of unsettled weather1 and the exceptional brilliancy is doubtless due to an unusual quantity of vapour at the cloud plane1 which must mean that the change from the dry stratum above to the damp one below must be much more sudden than is ordinarily the case4 ,clouds of this kind might well be called cirro-stratus cumulosus4 ,illustration3 ,plate #ai4 ,,cirro-,,cumulus4 ,we now come to two companion pictures1 ,plates #bj and #ba1 which were taken within half a minute of each other4 ,in the first the camera was directed towards the west1 and in the second it was facing the north-west4 ,the sun was nearing the horizon1 and was only just outside the field of view in each case1 so that the two photographs form a panorama of the western sky4 ,a solar halo had disappeared about half an hour previously1 and the cirro-nebula had changed into the remarkable forms of cloud depicted4 ,plate #ba shows cirrus ripples in the upper part1 and cirro-cumulus in soft1 ill-defined balls in the lower part2 but they were at the same level1 and are only different parts of the same cloud plane4 ,in ,plate #bb we see similar ball-like cloudlets ranged in long lines which run almost at right angles to the ripples of the companion picture4 ,clouds like these are rare4 ,they are almost unknown during the early part of the day1 and1 so far as the writer's experience goes1 they are only to be found in the afternoon towards sunset4 ,some of our most gorgeous sunset skies are due to them2 for their altitude is considerable1 and they do not light up with the sunset colours until the lower clouds have become dark shadows against the glowing background4 ,the hottest months of the year1 the still air and great evaporation which are the contributing causes of thunderstorms1 are also the conditions under which such skies may be seen4 ,indeed1 while these photographs were being taken1 heavy thunderstorms were in progress within less than a hundred miles4 ,cirro-cumulus nebulosus1 or hazy cirro-cumulus1 describes the form correctly4 ,illustration3 ,plate #bj4 ,,hazy ,,cirro-,,cumulus4 7,cirro-cumulus ,nebulosus47 ,illustration3 ,plate #ba4 ,,hazy ,,cirro-,,cumulus4 7,cirro-cumulus ,nebulosus47 ,the next plate1 ,no4 #bb1 gives a view of an evening sky about half an hour after sunset4 ,the lower clouds1 cirro-cumulus and cirro-stratus1 of a deep purple brown1 standing out dark against a gold-coloured sheet of higher cirro-stratus1 which comes out white in the photograph1 while the purple-tinted sky comes dark4 ,we have here three distinct layers1 all cirrus4 ,first1 the hazy cirro-cumulus1 forming two bars across the lower part of the picture2 then long bands of cirrus or cirro-stratus1 best seen in the bottom right-hand corner2 and1 far above both1 the cirro-stratus which was reflecting the yellow sunlight4 ,such a sky might be an indication of thunder conditions1 or it might be due to an unusual quantity of vapour in the atmosphere produced by some other cause4 ,the actual conditions were the gentle flow over ,england of vapour-laden air from the western ocean1 heralding the change from a long spell of fine hot weather1 due to a ,july anticyclone1 to a month of heavy rains and western gales1 accompanying the passage of a long procession of cyclones along our western shores4 ,illustration3 ,plate #bb4 ,a ,,sunset ,,sky4 ,again1 a marked contrast is shown in ,plate #bc4 ,here we have the highest and thinnest form of cirro-cumulus1 the one named cirro-macula by ,mr4 ,ley4 ,it is rarely1 if ever1 seen before eleven o'clock in the morning1 and is far commoner in the afternoon4 ,the example shown was photographed at sunset at the close of a day which had been almost cloudless4 ,cirro-macula forms here and there in a clear sky4 ,a hazy1 whitish patch appears1 which at first shows no definite structure1 but looks almost like a little bit of cirro-nebula4 ,this suddenly splits up by clear blue lanes running through it1 and cutting the patch up into irregular segments1 which quickly round themselves off into minute bits usually whiter on their edges and semi-transparent in the centre4 ,the process can be strikingly imitated by scattering on water some fine powder which will float4 ,if left without disturbance1 the particles draw together into numerous small groups1 leaving lanes of clear water between them4 ,illustration3 ,plate #bc4 ,,speckle ,,cloud 7,ley74 7,cirro-macula47 ,cirro-macula frequently gives rise to the fibrous form of cirrus we have called cirrus caudatus4 ,the granules of the cirro-macula grow denser1 and begin to drop their frozen particles as soon as they become large enough4 ,indeed1 a cloudlet of cirro-macula may sometimes be seen to turn bodily into a fine line of falling crystals1 which will be a curving line of cirrus4 ,on the other hand1 it will sometimes remain visible for an hour or more without any trace of descending streaks or floating fibres4 ,pure cirro-macula such as ,plate #bc is not often seen2 it is far more frequently mixed with more solid-looking cloudlets and descending fibres1 such as are shown in ,plate #bd1 which gives the same point of view as #bc1 but a quarter of an hour later1 and photographed with a longer focus lens4 ,these two photographs1 together with #bj and #ba1 give excellent examples of the use of the black mirror4 ,in none of the four could the naked eye detect all of the cloud structures4 ,the whole sky was a blaze of dazzling light1 but by adjusting exposure and development the details are fully brought out without the least difficulty4 ,illustration3 ,plate #bd4 ,,cirrus ,,caudatus ,,and ,,cirro-,,macula4 ,cirro-stratus1 we see from the examples which have been considered1 hardly deserves to be treated as a distinct genus of cloud4 ,its formation is identical with that of many species of cirrus1 or in some cases with that of the speckle cloud1 cirro-macula1 or even the coarser kinds of cirro-cumulus4 ,the different varieties which it shows are best rendered by reference to the specific names of the detached forms which have similar features4 ,cirro-cumulus1 on the other hand1 does present clearly marked varieties4 ,cirro-macula is so distinct that it might well be given the name awarded to it by ,mr4 ,ley1 while the term 8cirro-cumulus0 is reserved for the coarser and rounder forms4 ,the hazy1 ripple-like structures of ,plate #d and ,plate #bj should also have some distinctive appellation1 as will be suggested later on when dealing with wave clouds as a whole4 ,it is difficult to find any short way of expressing the various ideas which should be summed up in the name of a cloud4 ,there seems no alternative to the use of additional words1 unless it be to follow the example of chemists1 and compound appalling names similar to those which terrify the uninitiated who think they would like to read something about1 let us say1 the coal-tar dyes4 ,if a cloud belongs to the order cirrus1 is in a level sheet1 and that sheet is composed of interlacing or curling fibres1 like those of common cirrus1 we can hardly express the facts more briefly than by calling it cirro-stratus communis1 or common cirro-stratus4 ,if it consists of cirrus bands fused together1 but still showing the banded structure1 it is cirro-stratus vittatus4 ,again1 if it is finely speckled1 like cirro-macula1 it may be described as cirro-stratus maculosus1 and if the structure is coarser it may be called cirro-stratus cumulosus4 ,as a general average1 cirro-stratus lies somewhat lower in the atmosphere than the detached forms1 probably because the conditions which give rise to the latter reach to greater altitudes in patches than it is possible for them to reach in a continuous manner4 ,vapour becomes rarer with increased height and with diminished temperature1 so that it must1 on the whole1 be less frequently present in cloud-producing quantity as the height increases4 ,at great altitudes it will be seldom that the quantity is great enough to produce a stratiform cloud1 though it may well be enough for cirro-macula1 or the detached forms of cirrus1 like cirrus excelsus4 ,the production of cirro-cumulus and cirro-stratus sometimes spreads across the sky with astonishing speed1 and this rapid advance of the edge of the cloud may lead to quite mistaken ideas as to the velocity of the wind at that altitude4 ,in the case of cirro-cumulus1 or cirro-macula1 it is easy to fix attention on a single cloudlet4 ,if this has the usual ball-like form1 it can only be regarded as floating in the air and moving with it4 ,meanwhile new cloudlets may be forming and growing denser1 so that the cloud patch as a whole may be apparently advancing at a much greater rate4 ,careless observation would then lead to the idea that the cloud was moving much faster than it really is1 but if the attention is rigidly fixed on a particular cloudlet the mistake is impossible4 ,if the cloud is a variety of cirro-stratus1 it is not always easy1 or even possible1 to distinguish between the advance of condensation and the movement of the whole1 but it can nearly always be done if the cloud shows any definite features upon which attention can be fastened4 ,sometimes none sufficiently marked can be seen1 and when that happens it is still possible in most cases1 by watching the edge of the cloud-mass1 to see whether new cloud is being added to that edge4 ,the wave-like forms present a special case1 which will be dealt with in a later chapter1 after the general principles of cloud formation have been discussed in connection with the great clouds of the lower air1 whose causes and conditions are far better understood4 ,,chapter ,,iv ,,alto ,,clouds ,,from cirro-cumulus and cirro-stratus we pass through almost insensible gradations to the denser forms classed together in the alto group4 ,these clouds are fundamentally different1 in that they are always composed of liquid particles1 though there is no doubt1 from their great altitude1 that their temperature must often be many degrees below the ordinary freezing-point of water4 ,when this is the case1 they are not unfrequently more or less mixed with streaks and filaments exactly like those described under the name of cirrus1 which have been explained as due to slowly falling snowflakes4 ,it is not immediately obvious how such apparently contradictory statements can be reconciled4 ,the explanation is that minute droplets of water may be cooled many degrees below freezing-point without changing into ice1 and that such super-cooled droplets congeal instantly if a few of them join together to form a larger drop4 ,practically the same process may be watched any day when there is a sharp frost and dense fog drifting slowly along4 ,the fog-particles are liquid1 and produce optical effects in the neighbourhood of any brilliant light1 like an arc lamp1 absolutely the same as those which would be produced if the temperature were above freezing-point1 while there are none of the different phenomena which might be expected if the particles were crystalline ice-dust4 ,as these liquid particles drift along they come in contact with branches of trees and other obstacles1 the surface stratum which surrounds them and binds them into spheres is broken1 and the drop instantly solidifies4 ,it is to be noted1 moreover1 that the drop does not freeze as such1 but merely adds some more particles to the branching crystals of hoar frost1 which grow outwards always towards the direction from which the fog is drifting4 ,most liquids1 when freed from contact with solid bodies1 or when surrounded by a smooth envelope of uniform character1 can be cooled below their normal freezing-point without solidification taking place2 but the introduction of a particle of the solid1 or sometimes of any foreign body1 instantly brings about a rapid freezing of the whole4 ,these phenomena of surfusion1 as it is called1 have long been known1 and many of them are very interesting and difficult to understand4 ,indeed1 it is probable that we shall have to add largely to our knowledge of the forces which bind the molecules of a body together to form a solid1 and which direct the processes of crystallization before we shall be able to interpret with any certainty a series of facts depending on the attributes of those very forces4 ,water is no exception4 ,if finely divided1 as by placing it in fine capillary tubes1 in the pores of wood1 or in the narrow spaces of a wick1 it may be cooled several degrees below normal freezing-point4 ,in a cloud1 or fog1 all the conditions necessary for surfusion to take place are undoubtedly present4 ,the water is pure1 the envelope is uniform1 the subdivision is exceedingly minute1 and the drops are free from most of the mechanical disturbances which bring about the solidification of larger masses in the laboratory4 ,thus we see there is nothing at all surprising in the fact that clouds composed of liquid particles may exist at temperatures below the ordinary freezing-point4 ,on the contrary1 we should expect that the solidification of the cloud particles would not take place until the temperature was many degrees below freezing1 as is certainly the case with clouds of the cirrus order4 ,at temperatures between this unknown1 but low value1 and the normal freezing-point1 the clouds will be composed of liquid2 but when the particles join together1 snowflakes will result instead of raindrops2 and this will be just as true of alto clouds as it is of the great vaporous mountains of the lower regions of the air which bring falls of snow4 ,the streaks often mixed with alto-cumulus are cirrus threads1 and are1 no doubt1 of exactly the same nature as the tails of cirrus caudatus1 or even the fibres of cirro-filum4 ,the simplest alto cloud is alto-stratus4 ,when this is complete1 so as to cover the sky1 it can be distinguished from cirro-stratus by the absence of fibrous structure1 and by the facts that it never produces any halo or fragment of a halo1 but instead surrounds the sun or moon with a white blur1 or1 if it is thin enough1 with a close ring of coloured light much nearer than a halo1 and with the colours in the inverse order--that is1 with the red furthest from the centre4 ,some of these so-called coron are very beautiful when seen in the black mirror1 and some of those formed around a full moon show quite brilliant tints to the unaided eye4 ,of course1 these meteorological coron have no relation whatever to the true solar corona2 they are simply formed by the passage of the rays of light through the veil of small particles1 and may be easily imitated4 ,take a piece of glass such as a lantern-cover glass1 breathe on it1 and hold it close before the eye while looking at some small source of light4 ,if the dew deposit is thin1 bright colours are shown in a luminous ring surrounding the light1 and the thinner the deposit of dew the larger the ring will be4 ,breathe heavily so as to give a thick deposit1 and the light will be seen to be the centre of a patch of white brightness without any colour4 ,the phenomena are due to what is known as diffraction1 and if the other conditions are unchanged the diameter of the ring is inversely proportional to the size of the particles4 ,purity of colour in these rings is an indication of uniformity in the size of the particles4 ,when the moon is shining through a sheet of alto-stratus1 which thins off to one edge1 very beautiful effects may often be noticed1 and the change from the colourless blur1 when a thicker part of the cloud is interposed1 to the brilliant colours of the corona formed by the thinner edges is very striking4 ,similar phenomena are shown almost equally well by any of the alto clouds1 but cirrus thin enough to produce a coloured corona will generally produce a halo4 ,alto-cumulus of the kind most nearly allied to cirro-cumulus is shown in ,plate #be4 ,the upper part of the picture shows ragged1 irregular patches1 with slight indications of fibrous streaks4 ,the lower portion shows rounder1 ball-like cloudlets1 a few of the larger of which have distinct shadows on the side away from the sun4 ,this plate gives alto-cumulus in a partly formed condition1 but it is not a mere passage form4 ,sometimes exactly such a cloud will float overhead for hours1 showing very little movement and only slow changes of detail4 ,it is therefore a distinct variety1 and may be called alto-cumulus informis4 ,illustration3 ,plate #be4 ,,alto-,,cumulus ,,informis4 ,a less definite form is shown in ,plate #bf4 ,it may also be regarded as only partly formed1 but its construction is quite different1 every part being misty and ill defined4 ,it is a common cloud1 especially in sultry summer weather with still air4 ,under those circumstances1 after a hazy morning1 it may be seen slowly forming during the afternoon1 growing in density as the hours go by1 until it reaches a maximum about five or six o'clock1 after which it melts away1 or settles down into small patches of high stratus4 ,most frequent in summer1 it is by no means rare in autumn and winter1 but still air is essential4 ,from its hazy appearance it may be called alto-cumulus nebulosus4 ,illustration3 ,plate #bf4 ,,hazy ,,alto-,,cumulus4 7,alto-cumulus ,nebulosus47 ,fixity of detail and slow movement characterize both the foregoing forms1 and in that respect our next picture 7,plate #bg7 shows a cloud which is a great contrast4 ,its detached cloudlets are rather flatter and thinner1 and though the cloud as a whole will often persist for hours1 it is undergoing continual change1 and is formed when the air is far from still4 ,cloudlets form and gather into stratiform patches1 which soon break up again and disappear2 and the process goes on here and there1 sometimes accompanied by fairly rapid movement of the patches as a whole4 ,this cloud may be described as alto-cumulus stratiformis4 ,illustration3 ,plate #bg4 ,,flat ,,alto-,,cumulus4 7,alto-cumulus ,stratiformis47 ,we now come1 in ,plate #bh1 to a cloud of singular beauty4 ,it forms rapidly in a clear sky1 its first traces bearing a striking resemblance to cirro-macula1 but the floccules1 instead of remaining semi-transparent or dropping cirrus threads1 rapidly become opaque balls of cloud which lengthen upwards4 ,this upward tendency causes the formed cloudlets to have their longer axes vertical1 which is very characteristic4 ,it might be named alto-cumulus castellatus1 or high-turreted cloud4 ,mr4 ,ley named it stratus castellatus1 or turret-cloud1 but it certainly belongs to the cumulus section of the alto group4 ,thunder weather is the invariable condition for its production4 ,if it is seen1 at least in ,england1 thunderstorms are certain to be recorded not very far away4 ,when this particular photograph was being taken in ,south ,devon1 very destructive storms were recorded in ,brittany and in the ,english ,midlands1 and the anvil-shaped tops of unmistakable thunder-clouds were visible above the horizon while the exposure was being made4 ,illustration3 ,plate #bh4 ,,high ,,turreted ,,cloud4 7,alto-cumulus ,castellatus47 ,another form of almost equal beauty is shown in ,plate #bi4 ,the rounded balls make their appearance as semi-transparent spots upon the sky1 and in their general characters might easily be mistaken for cirro-macula4 ,but a few minutes will be enough to decide the question4 ,the little spots rapidly grow denser1 frequently becoming ragged at the edges2 they never drop down the slender filaments which usually descend from cirro-macula1 and their edges are never denser than their central parts1 which1 it will be remembered1 was a frequent feature of the true speckle cloud4 ,the cloudlets are obviously rounded balls arranged in patches1 which may turn gradually into alto-stratus by their fusion1 or1 after an existence of minutes or hours1 the whole may disappear by a disintegration of each ball1 by its breaking up into a ragged mass and melting away4 ,the altitude at which this cloud forms is between #ejjj and #ijjj metres1 according to measurements made by the writer1 the actual specimen figured being about #gjjj4 ,it is almost as characteristic of thunder weather as the last1 but whereas ,plate #bh shows a variety which is most often seen before #c p4m41 since it only occurs while the cloud planes are rapidly rising1 the one before us may be formed at almost any time of day1 but most frequently occurs in the afternoon4 ,an imperfect form of it is frequently met with about sunset1 in which the rounded balls are not usually so well defined as when the sun is high above the horizon4 ,alto-cumulus glomeratus would be a suitably descriptive name4 ,illustration3 ,plate #bi4 ,,high ,,ball ,,cumulus4 7,alto-cumulus ,glomeratus47 ,if it were possible to take a good typical example of the variety just described and roll it flat1 so that each cloudlet should be reduced to a lenticular shape1 we get a type which seems seldom to appear during the heat of the day1 and to be most frequent about sundown4 ,it consists1 as shown in ,plate #cj1 of distinct cloudlets1 with considerable spaces between them1 and gives the impression of a discontinuous level sheet4 ,but the component cloudlets are much too definite1 and preserve their individuality far too well to suggest any idea of a broken stratus2 the spotted structure is the predominant feature1 while the stratiform arrangement is almost equally plain4 ,alto-stratus maculosus would be a suitable term4 ,it is not so high a cloud as the glomeratus type1 the one shown being at an altitude of about #efjj metres4 ,the plate shows the position of the setting sun1 which is partly hidden behind some dark patches of broken alto-stratus 7fracto-alto-stratus71 the hazy form and boundaries of which form an effective contrast to the shining cloudlets #bjjj metres or so above them4 ,many of our most beautiful sunsets are due to this form of cloud1 particularly in the late autumn4 ,it is a cloud of calm weather1 and often floats apparently motionless1 and undergoing little change1 like flakes of glowing fire against the background of a fading sky long after the sun has disappeared4 ,it is not indicative of thunder conditions1 and it may occur on the margins of an anticyclone4 ,illustration3 ,plate #cj4 ,,mackerel ,,sky4 7,alto-stratus ,maculosus47 ,a lower and coarser form of the spotted alto-stratus is shown in ,plate #ca1 where it is seen through the gaps in a thin sheet of broken stratus4 ,in this case also the sun was getting low in the sky1 being hidden by the denser bit of stratus in the bottom left-hand corner4 ,illustration3 ,plate #ca4 ,,mackerel ,,sky4 7,alto-stratus ,maculosus47 ,alto-stratus does not often1 if ever1 grow from the fusion of the cloudlets of the maculosus type4 ,but it does come from alto-cumulus glomeratus1 and also from a form shown in ,plate #cb4 ,here we have alto-stratus in process of growth4 ,small irregular lumps of cloud forming on the right-hand side of the picture grow larger and more irregular1 begin to fuse together towards the centre1 and on the left-hand side the fusion is almost complete4 ,still1 although the sky is covered with cloud1 the lumpy form is plainly visible4 ,the term 8alto-strato-cumulus0 is suitable1 as it differs from the more frequent and much lower cloud1 which will be described further on as strato-cumulus1 in little else than altitude and general massiveness of texture4 ,this high strato-cumulus is common enough1 too common1 indeed1 in ,england1 as it produces many a dull grey sky both in summer and in winter4 ,in the latter season it is not unfrequent with the cold east winds of ,february and ,march4 ,it is probably the lowest of the alto clouds2 the lowest measurement made by the writer being #ahbh metres at ,exeter1 but lower altitudes seem to have been recorded elsewhere4 ,illustration3 ,plate #cb4 ,,alto-,,strato-,,cumulus4 ,alto-cumulus castellatus1 which is breaking up and disappearing1 is shown in ,plate #cc4 ,it was photographed with a long-focus lens1 so that the scale of representation is about eight times as great as that of ,plate #bg4 ,this view was taken at ,exeter while a thunderstorm was in progress at ,bristol4 ,illustration3 ,plate #cc4 ,,sunset4 7,alto-cumulus ,castellatus ,fractus47 ,,chapter ,v ,,lower ,,clouds ,,the clouds of the lower portions of the atmosphere are formed in regions where water vapour is abundant1 and frequently in easy reach of the strong ascending and descending air currents produced by the varying temperatures and irregular surface of the ground4 ,it is sufficient to recite these conditions to show that these lower clouds will be denser1 larger1 coarser in texture1 and characterized by greater definiteness of form than those we have1 so far1 considered4 ,in the ,international system they are classified thus-- ,group ,c4 ,lower clouds4 7a7 ,strato-cumulus4 7b7 ,nimbus4 ,group ,d4 ,clouds of diurnal ascending currents4 ,cumulus and cumulo-nimbus4 ,group ,e4 ,high fogs4 ,stratus4 ,this is certainly the least satisfactory part of the whole scheme1 and it is not at all easy to see upon what grounds it was adopted by the ,international ,committee4 ,group ,d--cumulus and cumulo-nimbus--do show important differences from the other groups1 though it is often difficult to say whether the sky should be described as covered with strato-cumulus or as covered with numerous small cumulus4 ,it is the separation of stratus--placing it in a group by itself1 and making that the lowest--which is the worst point4 ,as a matter of fact1 stratus may exist at any altitude from sea-level up to such heights that we should not hesitate to call it alto-stratus4 ,indeed1 there is no essential character of alto-stratus which distinguishes it from some of the lower forms4 ,whatever its altitude1 its thickness and the size of its particles may vary in a precisely similar manner4 ,we may have the particles exceedingly small1 when the fog will be dry1 and such a stratus may be so thin as hardly to dim the sun2 or it may be so thick as to completely hide it4 ,on the other hand1 the fog may consist of particles easily visible to the naked eye1 forming the so-called ,scotch mist1 or the 8dry0 fog of ,dartmoor1 which will wet things as rapidly and more thoroughly than a smart shower4 ,when such a fog accumulates to a sufficient depth1 the particles in their fall pick up others1 and the result is a distinct fine rain4 ,this may occur not only near the ground1 but at almost any level below that at which the cloud would pass into the region of cirrus4 ,plate #cd shows three layers of stratus1 in each case much broken up4 ,the highest layer is a good example of alto-stratus maculosus4 ,lower down1 by half a mile or more1 come parts of a grey sheet considerably denser and thicker4 ,it is a matter of taste whether this should be called high stratus or low alto-stratus4 ,there is no test by which the one can be distinguished from the other4 ,lower again come the detached darker clouds1 which are fragments of a sheet of stratus which is breaking up and disappearing4 ,the photograph was taken in the afternoon1 after a wet morning1 and all three layers were probably relics of the great rain-cloud system1 or nimbus1 which produced the rainfall4 ,illustration3 ,plate #cd4 ,,three ,,layers ,,of ,,stratiform ,,cloud ,,after ,,rain4 ,we have referred to the production of fine rain from a thick fog4 ,if now such a thick layer of coarse-grained fog--if we may use such a phrase--is suspended overhead at a moderate altitude1 the result is a drizzling rain underneath1 and the cloud at once becomes a nimbus4 ,when ,luke ,howard adopted the term 8nimbus10 he proposed to employ it1 apparently1 for a vast mass of cloud such as that which forms the rainy region of a cyclone2 a huge pile of clouds containing representatives of all his other types in some unknown but close relationship4 ,it was1 in fact1 a comprehensive term1 and as such there was a good use for it4 ,at present it is applied to any cloud from which rain is falling1 except when the cloud can be identified as a variety of cumulus which is called cumulo-nimbus4 ,but we have already said that a stratus may be a rain-cloud1 and so may other varieties4 ,moreover1 whenever a nimbus breaks sufficiently for us to be able to see its upper surface1 we invariably find that1 if it were viewed from above1 we should1 without a moment's hesitation1 place it in one of the other groups4 ,it is only when we are underneath it we can see its rain-producing character1 and give it the orthodox name4 ,the real fact is that nimbus should be an adjective1 meaning rain-producing1 and not a substantive4 ,however1 it has its allotted place in the ,international system1 and it is better to adhere as far as possible to a defective but widely recognized system until it can be authoritatively amended1 rather than to make an individual attempt to ignore it4 ,the facts are sufficiently obvious1 and the days of nimbus as a type are numbered4 ,the two plates1 ,nos4 #ce and #cf1 are fair typical representations of the clouds usually known as nimbus2 but they are both of them only the under-surfaces of other clouds1 ,plate #cf showing the under-surfaces of a group of heavy cumulo-nimbus all joined together so as to cover the sky1 while ,plate #ce shows a mass of dense strato-cumulus4 ,the rain-cloud is always a form of either stratus or cumulus1 or a combination of the two1 sometimes in further combination with clouds of the alto class1 or even extending upwards to cirro-stratus and cirro-nebula4 ,where it consists of a single layer1 that layer differs from its rainless representative only in greater thickness from base to summit1 or in greater density2 and when there are several distinct layers of cloud1 so that the lowest is shaded by the higher1 rain may fall1 even though they differ in no visible way from clouds which would be rainless if alone4 ,plate #cc is an example4 ,illustration3 ,plate #ce4 ,,rain-,,cloud4 7,nimbus47 ,illustration3 ,plate #cf4 ,,rain-,,cloud4 7,nimbus47 ,nimbus1 indeed1 is not a type-form1 but is merely a typical condition1 and when used as a substantive is only a convenient way of expressing our ignorance as to the real form of the cloud we so describe4 ,the altitude of the base of a rain-cloud may vary considerably4 ,it may be anything from sea-level up to heights which vary with the geographical conditions and with the conditions of temperature and pressure1 but probably in this country never greater than #gjjj or #hjjj metres4 ,rain1 or snow1 often falls from clouds at greater altitudes than these1 but unless in its descent it passes through other lower clouds1 the drops1 as a rule1 will dry up and disappear4 ,the author has often seen quite heavy rain descending from a cloud1 and disappearing completely within a thousand feet or so of the cloud-base4 ,on rarer occasions a still more remarkable thing may be seen--namely1 a shower falling from an upper cloud into a lower1 and none between this lower cloud and the ground4 ,this curious phenomenon can only be explained by supposing that the convection currents which make the lower cloud are strong enough to support the small raindrops4 ,pure stratus is a level sheet of cloud with little variation of thickness1 not ascending every here and there into rounded lumps4 ,its most typical form covers the whole sky with a uniform grey pall1 which may or may not completely hide the sun4 ,such a cloud does not lend itself to pictorial representation4 ,a frequent form1 in which the sheet is more or less broken1 is shown in ,plate #cg4 ,this is a variety which is frequent in the summer mornings1 and generally breaks up and clears away before eleven o'clock4 ,if1 however1 it appears in autumn and winter with layers of alto cloud above1 it may grow denser1 and turn into a stratiform nimbus1 or it may go on drifting overhead for several days without sign of change4 ,illustration3 ,plate #cg4 ,,stratus ,,communis4 ,break up such a sheet of cloud by numerous meandering cracks1 and round off the detached pieces so as to give them a more or less rounded or pyramidal section1 and the cloud becomes strato-cumulus1 typical representations of which are shown in ,plates #ch and #ci1 which depict different parts of the same sky4 ,in ,plate #ch the camera was pointed due west1 and in ,plate #ci it was turned round to the north-west1 so that the two views do not quite meet4 ,illustration3 ,plate #ch4 ,,strato-,,cumulus4 ,illustration3 ,plate #ci4 ,,strato-,,cumulus4 ,plate #dj is a different variety4 ,it is stratiform1 each component cloudlet being rather ragged at the edges4 ,in some ways it resembles cirro-macula and the speckled varieties of alto cloud1 but it is coarser in texture and obviously at no great altitude4 ,the ,international system would call it strato-cumulus1 but ,mr4 ,ley gives a representation from another negative taken at the same time as the type of what he calls stratus maculosus1 a name which seems far more suitable1 since the cloud bears a much closer relation to stratus than to cumulus4 ,in the particular instance figured1 the broken structure did not last long2 the spaces gradually closed in1 and a complete stratus was the result4 ,illustration3 ,plate #dj4 ,,stratus ,,maculosus4 ,strato-cumulus often lasts for hours1 with little or no perceptible change1 but stratus maculosus rarely persists for more than half an hour4 ,the first is a cloud of fairly stable conditions1 the latter is dependent for its existence upon the near approach to critical conditions at one particular level1 and1 as we have said in other cases1 such a critical state is almost always soon passed1 with the result that the cloud either masses into a denser form1 or else breaks up and disappears4 ,if the up and down currents are strong enough to persist1 the result will be strato-cumulus and not stratus maculosus4 ,a kind of stratus which is frequently seen in the daytime is shown in ,plate #da4 ,this is literally a lifted fog1 having been formed about midday1 after ground fog in the early morning4 ,it would be called common stratus1 or stratus communis4 ,when it appears it is a fairly persistent form1 sometimes breaking up or swelling up into strato-cumulus1 but more often splitting into long rolls of cloud1 with margins like those of cumulus4 ,this phenomenon is shown in ,plate #db1 which was taken in ,december at #aa a4m41 on a day which opened with a thick ground fog4 ,a precisely similar cloud is frequent in the early hours of a summer morning1 as a stage in the dispersal of a radiation ground fog4 ,the fog first lifts from the ground1 until it reaches a height of a few hundred metres1 when it splits into the long rolls whose axes are at right angles to the direction of drift4 ,the consequence is very strange if you stand on a hilltop close under the drifting mass1 and look towards the horizon in the direction of drift4 ,the changing shadows give the impression that the clouds are actually rolling along1 though of course no such thing is really taking place4 ,as time goes on the rolls grow larger and the interspaces wider2 then transverse fissures appear1 and gradually the rolls break up into small detached cumulus4 ,cumulus radius1 from the ,latin for a rolling-pin1 might be a suitably descriptive name1 but it should not be forgotten that it is only an intermediate link between stratus and cumulus1 and1 indeed1 is more nearly related to the former1 since it is never produced except on the break up of stratus1 while it may dry up and disappear without reaching the cumulus stage at all4 ,stratus radius would therefore be a better name4 ,illustration3 ,plate #da4 ,,common ,,stratus4 7,stratus ,communis47 ,illustration3 ,plate #db4 ,,roller ,,cloud4 7,stratus ,radius47 ,cumulus is closely related to another form of stratus1 which ,mr4 ,ley has named stratus lenticularis1 but this appears to be so frequently the last stage in a disappearing cumulus that its history will come better later on4 ,it is mentioned here as it is1 after all1 one of the commonest of all forms of stratus1 the form which appears at1 or after1 sunset1 and is one of the few clouds which have an ,english popular name--,fall cloud4 ,plate #dg gives a representation of it1 standing out dark against an evening sky1 with a sheet of alto-stratus far above it in the upper part of the photograph4 ,to sum up1 then1 we have among the lower clouds of more or less stratiform pattern-- ,stratus communis1 or ,common stratus4 ,stratus lenticularis1 the ,fall cloud4 ,stratus radius1 or ,roll cloud4 ,stratus maculosus1 or ,mottled stratus4 ,strato-cumulus1 or ,sheet cumulus4 ,the last leads naturally to the consideration of cumulus and cumulo-nimbus1 while the term 8nimbus0 does not belong to any one type-form1 but sometimes to one1 sometimes to another1 and generally to a mixture of two or more4 ,a good many years ago the writer made a series of measurements of the thickness of detached clouds of the stratus and cumulus types1 such as those which may produce a shower4 ,the conclusions reached in consequence of those determinations have since been amply confirmed by subsequent observations4 ,in winter no rain will fall from a cloud unless it reaches a minimum thickness of at least #ajj metres1 while in summer it must have rather greater thickness4 ,there is one exception1 and that is in winter1 when the temperature is so low that the drop starts on its downward journey as a flake of snow4 ,when this is the case1 rain may fall from a layer of thin lifted fog1 not quite thick enough to hide the blue colour of the sky4 ,but under ordinary conditions of temperature1 if the cloud has a thickness less than #bjjj feet1 or #faf metres1 rain is unlikely1 but if it does come1 the drops will be small and the fall of rain quite trifling4 ,above this thickness the heaviness of the rain and size of the drops increases1 so that if the distance from base to summit be between #bjjj and #cjjj feet1 or #fjj to #ajjj metres1 the fall will be gentle4 ,a thickness of #djjj to #fjjj feet1 or #abjj to #ahjj metres1 gives large drops and a fairly heavy shower1 while1 in summer time at least1 cold heavy rain and hail come from clouds measuring #fjjj to #aj1jjj feet1 or in round numbers #ahjj to #cjjj metres or more4 ,in winter the necessary dimensions seem to be less1 but the rule still holds equally good1 that the rain-cloud does not necessarily differ in any way from the rainless one1 except in thickness1 and that when the requisite thickness is present rain is not always the result4 ,,chapter ,,vi ,,cumulus ,,under the general term cumulus there are grouped the most common1 the best known1 and the grandest forms of cloud4 ,indeed1 beautiful as the cirrus and alto clouds may be1 there is a solid grandeur about the greater forms of cumulus which gives them a beauty of their own quite comparable with the charm afforded by the delicate tracery of their more lofty rivals4 ,cumulus can be divided into several types1 which are best considered in the order of growth4 ,they are all formed in the lower part of the atmosphere1 their under-surfaces varying in altitude from about #fjj metres1 or even less1 up to #cjjj metres1 or slightly more4 ,the writer's own measurements vary from a minimum of #ehd metres to a maximum of #bbhf metres1 with an average of a little more than #ajjj metres4 ,they are described in the ,international system as 8clouds in a rising current10 and there is no doubt the description is correct4 ,each cumulus must be looked upon as simply the visible top of an ascending pillar of damp air4 ,the vapour which makes its appearance in the cloud is present in the transparent air beneath1 and the base of the cloud is simply the level at which that vapour begins to condense into visible liquid particles4 ,since cumulus clouds are caused by ascending currents1 these currents must be brought about either by the general disturbance of the air due to a cyclonic movement1 or by the local irregularities of temperature on the ground produced by the sun's heat4 ,as a matter of fact1 we do get cumulus produced in great abundance in the rear of every cyclone1 and we get them also under the conditions of still air and hot sun1 which specially favour evaporation and the development of differences of temperature4 ,the cyclone cumulus may come at any hour of the day or night1 though comparatively rare between midnight and the morning4 ,heat cumulus is generally formed during the afternoon1 and it is only under relatively uncommon conditions that it persists during the night4 ,if the cloud has not grown to very great size it usually begins to break up and disappear about sunset1 but if it has grown to the enormous dimensions of a summer thunder-cloud it may go on growing1 piling mass on to mass1 until it generates a thunderstorm1 even in the hours of early morning4 ,in the case of some of the higher kinds of cloud1 we are not able to give any certain account of the mechanics of their production from a study of those clouds themselves4 ,we have already referred incidentally to some of the speculations as to their origin and some of the facts definitely known1 but considerable light can be thrown on the genesis of all the varieties of cirro-cumulus and alto-cumulus by a careful study of their larger and more accessible representatives of lower regions4 ,the cyclone cumulus does not differ in any essential from the clouds of calm weather4 ,the only difference is that the uprising currents are perhaps partly eddies1 and the rate of fall of temperature with ascent is often more rapid4 ,given any mass of air at a particular temperature1 it can take up and hold in the form of invisible vapour a fixed quantity of water1 and no more4 ,when it holds the maximum possible it is said to be saturated4 ,if it is nearly saturated it would be called damp2 if far from saturated1 dry4 ,now1 the warmer the air the larger the quantity of vapour necessary to saturate it1 so that if a quantity is saturated at a high temperature1 and is then cooled1 it will no longer be able to retain all its moisture in the invisible form1 but the surplus quantity will make its appearance as liquid particles1 that is to say1 as mist or cloud4 ,similarly1 if a quantity of air is not fully saturated at its particular temperature1 and is then cooled1 it will approach nearer and nearer to saturation1 and if the process is continued long enough the result will be cloud formation4 ,all clouds1 without exception1 are produced by exactly such cooling of air containing water vapour1 first to the temperature at which the quantity it contains is the maximum possible1 and then beyond that point4 ,now1 if we start with very warm air1 and cool it #a degree1 we decrease its vapour-holding power1 and the decrease per degree grows less and less as the temperature falls4 ,suppose1 for instance1 we have air saturated at #fa degrees and cool it to #fj degrees1 the quantity of vapour condensed will be equal to the difference of holding power4 ,suppose1 again1 we have air saturated at #ca degrees and we cool it to #cj degrees1 the quantity of vapour condensed will again be equal to the difference of holding power2 but this quantity will be very greatly less than in the former case4 ,cooling air saturated at #fa degrees to #fj degrees might produce a dense cloud2 but applying a similar reduction of #a degree to air saturated at #ca degrees1 if we take the same volume of air1 will only produce a very much thinner result4 ,here we see one good reason why the highest clouds are the thinnest and the alto clouds of intermediate density4 ,the necessary cooling may be brought about in several ways4 ,firstly1 the air is capable of radiating its heat into space1 and therefore of cooling4 ,but we know little of the laws which govern atmospheric radiation1 and presumably1 if cloud could be produced by such means1 it ought to make its appearance most frequently in the small hours of the morning before sunrise4 ,we are1 however1 unaware of any variety of cloud which answers those conditions1 unless it be the ground fogs which so often form during the night2 and these1 we know1 are certainly due to the chilling of the air by contact with the ground1 which has been cooled by radiating away its heat4 ,on the contrary1 it is well known to astronomers that the sky is1 on the whole1 clearer and freer from clouds after midnight than in the earlier hours of the night--a circumstance which is particularly unfortunate for the amateur star-gazer1 who has to be up and about at the same time as the rest of the working world4 ,cooling by radiation we may then dismiss as a cause of cloud formation of no great efficacy1 and certainly one which has little to do with the production of cumulus4 ,cooling by contact with a cold body is another and more potent cause4 ,we often see it in a mountain district1 where a frost-bound peak stands facing the wind with glittering snow-slopes on which the sun is shining1 while a long tongue of cloud hangs like a banner on its leeward side4 ,in such a case it is easy to understand how the air sweeping by the icy mass is chilled below its saturation point2 but as it passes on1 the chilled portions become mixed with the rest1 and the cloud evaporates again4 ,it is not quite so easy to see how far this cause is responsible for the clouds which are formed when the warm damp air of the ocean drifts over a comparatively cold land4 ,it is probable that the contact chilling is in this case only part of the explanation1 and that other causes co-operate4 ,the mixing of warm damp air with cold has often been adduced as a cause of clouds4 ,no doubt it might be1 and some of the stratiform types may possibly be formed at the junction between a warm damp stratum of air and a cold one1 but no example is certainly known4 ,it may also be a contributing cause in producing the sharply defined upper surfaces of some cumulus or strato-cumulus clouds1 but these are in the main most certainly due to the chief cause of cloud production--namely1 what is known as dynamic cooling4 ,if a quantity of air exists under a certain pressure and at a certain temperature1 on reducing the pressure it will expand1 and in the act of expanding it will become cooler4 ,this may easily be illustrated with an air-pump4 ,let a damp sponge or a piece of wet blotting-paper stand under a glass receiver over an air-pump until the air has become damp4 ,if the apparatus is in a darkened room1 and a powerful beam of light from a lantern is sent through the receiver1 the damp air will be seen to be quite clear2 but a stroke or two of the pump removes some of the air1 the remainder is chilled by its own expansion1 and a dense cloud is precipitated4 ,if this cloud be viewed closely1 it will be seen to be composed of minute particles1 which1 on looking towards the light1 glow with the colours of a corona4 ,in a few minutes the cloud will disappear1 but it can be recalled again and again by successive strokes of the pump1 getting thinner and thinner as the air gets more and more rarefied2 an illustration of a second reason why the high clouds are thinner than the lower4 ,some years ago ,mr4 ,john ,aitken showed that if the damp air used in this experiment were carefully filtered1 so as to remove all foreign particles1 no cloud was produced1 and the introduction of a puff of unfiltered air was attended by immediate condensation4 ,the deduction was that vapour1 even below its saturation temperature1 cannot produce cloud unless nuclei of some sort are already present1 presumably dust particles4 ,later on it was shown by ,mr4 ,shelford ,bidwell and others that gaseous particles1 such as those produced by the burning of sulphur1 would serve the purpose1 and that the brush discharge from an electrified point was in some mysterious way particularly effective4 ,it has recently been shown by ,mr4 ,c4 ,t4 ,r4 ,wilson that causes such as the radiations of radium1 or the impact of ultra-violet rays1 acting on the air itself1 splits up some of its particles into the smaller bodies known as ions1 and that these are efficient nuclei4 ,these experiments open up many most interesting questions1 but1 unless it is to explain the extreme density and darkness of a thunder-cloud1 they do not seem to play any important part in determining the forms to be assumed4 ,nuclei in sufficient abundance are probably always present at any height which can be reached by enough vapour to form a cloud4 ,now1 if we have a quantity of air1 say at sea-level1 damp but not saturated1 and it is caused to ascend1 either because it is warmer and therefore lighter than the surrounding air1 or for some other reason1 as it moves upwards the pressure upon it will decrease1 it will expand1 and in the act it will be steadily cooled4 ,this cooling may after a time bring it down to the same temperature as the rest of the air at its particular level4 ,if so1 it will no longer be lighter1 and the ascent will come to an end4 ,but before this state of affairs is attained it may have reached its saturation point1 and cloud production will begin4 ,it is true that the rarefaction of the air tends to enable it to retain more vapour than it could if it were cooled without change of density4 ,the temperature of the air being fixed1 its holding power increases with decrease of pressure4 ,but this increase is much less than the diminution due to cooling1 and the result in nature must be similar to what we can see happen under the receiver of the air-pump4 ,the condensation of water introduces another factor of great importance4 ,it has just been said that the ascending air may be cooled so rapidly as to be reduced to the same temperature as the rest of the air at that level1 and if so the ascent will end4 ,clearly the cessation or persistence of the upward motion depends upon whether the diminution of temperature per #ajj metres of ascent is most rapid in the rising column or in the air outside it4 ,as long as the ascending air is warmer than that outside1 but at its own level1 so long will ascent continue4 ,now1 as long as no condensation was taking place1 the rate of cooling would follow a simple law which produces a cooling of #a degree for about #ajj metres of ascent2 but as soon as water vapour begins to pass into the liquid form1 a large quantity of heat is set free1 and the rate of cooling is consequently greatly lessened4 ,cloud production tends1 therefore1 to accelerate ascent1 and the greater the amount of condensation1 the more important will this consideration become2 though1 on the other hand1 when once the cloud is formed1 it tends to stop the rising current by shading the air and ground beneath it4 ,on an ordinary day the rate of decrease of temperature as we ascend is rather less than the value given above1 and uprising currents are soon checked4 ,if they do extend far enough to reach cloud production1 the clouds will be small1 forming the smallest variety of cumulus4 ,this is shown in ,plate #dc4 ,small irregular uprising currents have just been able to reach far enough up to have their summits tipped with cloud4 ,illustration3 ,plate #dc4 ,,small ,,cumulus4 7,cumulus ,minor47 ,after the foregoing explanation1 it is easy to see why at a given time the floating cloudlets should have a common base level4 ,this is the height to which the air must attain before reaching its saturation temperature4 ,each cloudlet marks an uprising current1 and the intervals show the position of the counterbalancing descending streams4 ,a larger variety is shown in ,plate #dd4 ,in this the level base and generally pyramidal shape is shown1 and also the hard1 rounded upper surface4 ,the thickness of this cloud was about #ejj metres4 ,when clouds like these are visible1 they may be the beginning of larger ones1 and the only way to judge whether they are likely to develop into rain- or shower-clouds is to watch them4 ,if they are seen to be growing larger1 and particularly if detached fragments are developing into clouds1 further growth is almost certain1 and rain is probable4 ,illustration3 ,plate #dd4 ,,cumulus4 ,if great towering masses are making their appearance with little dark fragments between them1 as shown in ,plate #de1 then smart showers may be confidently expected4 ,the cloud figured was a shower-cloud1 and the distance is seen through the veil of falling rain4 ,the height and thickness of this particular cloud were measured just after its photograph had been taken4 ,its base was #abjj metres above the ground1 and its summit was #aejj metres further4 ,its thickness from summit to base was1 therefore1 not much short of a mile1 and the total contents of the cloud were probably between one and a half and two cubic miles4 ,the upper contour is hard and rounded1 as in the smaller cloud of ,plate #dd1 but the whole cloud is much larger4 ,illustration3 ,plate #de4 ,,large ,,cumulus4 7,cumulus ,major47 ,we have already explained that there seems to be a definite connection between the thickness of such clouds and the amount of precipitation from them4 ,small cumulus1 less than #abj metres thick1 rarely produces rain1 and nothing like a heavy shower is likely unless the thickness exceeds #djj metres4 ,in winter1 especially in hard frost1 snow crystals may fall from the smallest cloud1 even from little fragments only a few metres thick1 but the quantity of water so precipitated will1 of course1 be small4 ,as long as the top of the cumulus is rounded and clearly defined1 the conditions of arial equilibrium are stable1 and the growth of the cloud has been brought to an end by a stoppage of the ascending current4 ,in ,plate #de the ascent has been hindered both by the mechanical action of the falling raindrops and by the cooling of the lower parts of the ascending column by the descent into it of the cool drops from its colder upper part4 ,this is probably one of the chief reasons why a shower-cloud never maintains its activity as a rain producer for more than a very limited period4 ,as the cloud drifts over the landscape1 it seldom maintains its showery character for more than ten or twenty miles1 often for much less4 ,cumulus1 like any of these three1 is a cloud of the daytime4 ,it generally begins about ten or eleven o'clock in the morning1 grows larger until about four o'clock1 and then begins to break up and disappear4 ,after the ascending currents have ceased1 the component cloud particles slowly settle down into the warmer air beneath1 until the mass has lost its proper pyramidal form1 and has become an irregular cloud1 such as is shown in ,plate #df4 ,this is known as degraded or fracto-cumulus4 ,illustration3 ,plate #df4 ,,fracto-,,cumulus4 ,one consequence of the arrest of the uprising currents is the formation of lenticular patches of stratus1 called by ,mr4 ,ley stratus lenticularis4 ,this is often formed about sunset1 and has been named fall cloud1 from its appearance at the fall of night4 ,the name is appropriate in another way4 ,the ascending currents having ceased1 the cloud particles slowly subside until they dry up in some warmer stratum4 ,the water vapour does not continue its descent1 but slowly diffuses in all directions1 and if the fall of cloud particles is sufficient1 this stratum1 which is approximately coincident with the base of the original cumulus1 soon becomes saturated1 and further particles which fall into it remain visible4 ,this saturated zone will slowly sink lower and lower with the descent of the particles1 until it reaches regions in which the temperature is high enough for the whole to be evaporated without reaching saturation point4 ,evening stratus in calm weather always goes through this sequence of changes4 ,it usually forms at1 or soon after1 sundown1 and begins to break up and disappear as the stars are becoming visible in the darkening sky4 ,plate #dg shows a specimen of this evening stratus4 ,illustration3 ,plate #dg4 ,,fall ,,cloud4 7,stratus ,lenticularis47 ,a curious feature is sometimes shown on the underside of a thick cloud1 which is probably due to the upper part of the ascending column having been carried beyond its position of equilibrium by its own inertia1 and then falling back again in the teeth of the still rising lower part4 ,the result is to give the base of the cloud an appearance of a number of rounded masses hanging downwards below the cloud1 very suggestive of the idea that the cloud is upside down4 ,such an event will not often occur1 and when it does the conditions are quite wanting in stability1 and the consequent features will be very transient4 ,when the base of a cumulus or cumulo-nimbus is so affected1 the cloud is known as festooned cumulus1 or cumulus mammatus4 ,a precisely similar structure may be seen under strato-cumulus1 or even thick stratus4 ,in some countries it seems to be frequently observed1 but in ,england it is so uncommon that the writer has only noted it about a dozen times in twenty years1 and on no one of these did it last long enough to allow of its portrait being taken4 ,it is an indication of very disturbed conditions1 and is usually followed by heavy rain4 ,when cumulus clouds are formed in air which is steadily moving as a whole1 that is to say1 when there is a steady breeze1 they have a very decided tendency to follow each other in long lines4 ,it may often be noticed that in a particular place with a certain direction of wind these long processions follow definite tracks in relation to the geographical features4 ,the phenomenon does not seem to have been recorded except in hilly country1 but has frequently been observed by the writer4 ,it is not the same thing as the formation of stationary belts of cloud transverse to the wind4 ,these cumulus float along with the movement of the air1 and the question to be answered is1 why should they follow each other so persistently1 and why should the intervening belts of sky be so continuously free from cloud4 ,if we consider that the warm damp air which supplies them is drawn from the ground1 it seems that any cause which tends to direct this warm stratum into definite channels1 as it is carried on by the wind1 will be a competent cause of the whole phenomenon4 ,this we find in the presence of lofty hills which stand in the way of the warm surface winds1 causing them to follow more or less the general trend of the valleys1 and so delivering the rising convection currents of cloud-producing air at the same spot4 ,it is easy to conceive that other causes1 such as a difference in temperature or dampness of neighbouring tracts1 resulting from whether they are bare or wooded1 marshland or sandy plain1 might equally suffice2 or might1 at least1 powerfully co-operate with1 or counteract1 the effect of hill and vale4 ,but in any case it is plain that the geographical conditions to the windward of the place of observation not only may affect the occurrence and distribution of cloud1 but if the wind is steady it is difficult to see how they could avoid affecting it4 ,another puzzling phenomenon1 sometimes presented by cloud and fog1 is that our instruments for detecting humidity show that the air within them is not always fully saturated4 ,it seems probable that this is due to such cloud or fog having begun the process of drying up1 or that in some way not fully understood the presence of the cloud particles after they have first come into existence may cause the withdrawal of some of the moisture from the intervening damp air4 ,the surface of each minute droplet exerts a pressure on its interior similar to the pressure exerted by the film of a soap-bubble on the air within it1 and it is conceivable that some of the uncondensed vapour from outside may diffuse through this enclosing surface film1 and be retained there in consequence of the pressure4 ,if this is so1 and subsequent investigation can alone decide the matter1 it will follow that when once cloud production has begun it will be continued until the air between the cloud particles is reduced so far below its saturation point that the tendency of the drops to evaporate1 that is to say1 for the imprisoned water to escape through the confining film1 balances the retaining pressure4 ,this consideration1 however1 is quite incompetent to affect the general explanation of cloud formation which has been given4 ,its result would be to carry condensation a little further than the exact saturation point1 and to retard equally slightly the subsequent evaporation of the cloud particles4 ,we have spoken of the typical cumulus as having a roughly pyramidal shape1 and if the horizontal movement of the air is small1 the loftiest point of the cloud will be situated approximately above the centre of its base4 ,but if the horizontal movement increases in velocity1 so that the top is in a more rapidly moving stratum than the base1 it will lean forward in the direction of movement4 ,this is a very common phenomenon1 being generally shown by cumulus on a windy day4 ,on much rarer occasions the converse occurs1 and the top of the cloud lags behind the base1 the explanation being a lessening of the velocity of the wind as the height above ground increases4 ,but such conditions rarely occur1 and when they do they are due to local eddies and affect only a limited area4 ,hence such clouds are isolated1 and indicate a disturbed state of the air and uncertainty of weather4 ,the clouds which lean forward are formed under conditions which are spread over wide districts1 such as the rear of a large cyclone1 and cumulus of that kind may follow one another across the sky for hours or even days as long as the wind persists4 ,so far we have considered only the round-topped types of cumulus--those which mark the tops of ascending currents whose ascent has been stopped at a comparatively early stage1 or those whose ascent is still in that early stage1 though the upward movement has not yet come to an end4 ,the full story of the growth of a cumulus is identical with that of the youth of a cumulo-nimbus1 the later stages of which we will consider in another chapter4 ,,chapter ,,vii ,,cumulo-,,nimbus ,,grandest of all clouds are the huge mountains of vapour which are the parents of summer thunder-storms4 ,they are at once distinguished from ordinary cumulus by their upper parts1 which sometimes reach beyond the region of the alto clouds high into the realm of cirrus1 and extend outwards as a broad disc1 which is occasionally indistinguishable from the cirro-nebula and cirro-stratus which form the van of a cyclone cloud canopy4 ,indeed1 there seems to be no essential dividing line between a large cumulo-nimbus and the cloud pile of a small cyclone1 and no real difference between them except their size4 ,as a matter of fact1 the term cumulo-nimbus would only be given to the cloud when a large fraction of the whole can be seen at once4 ,in dealing with common cumulus1 it has been pointed out that the cessation of the uprising convection currents which determines the maximum height to which the cloud will grow is due to the rate of cooling within the ascending column being greater than the rate of cooling outside it4 ,it follows that when the ascending current has reached a certain height it will1 as a whole1 be just as heavy as an equal column outside4 ,ascent must then cease4 ,the equilibrium of the air in such a case is said to be stable1 and the condition of such stability is simply that the general rate of fall of temperature per #ajj metres of ascent is less than the rate of cooling dynamically produced in an ascending current4 ,if1 however1 the general rate of fall of temperature is greater than that produced dynamically1 the consequence will be that the upward tendency of the rising air will increase as it moves upward1 and the taller the column becomes the greater will be the difference of weight between the inside and outside columns4 ,in such a case the equilibrium is said to be unstable1 and the result will be the production of cumulo-nimbus4 ,just as cumulus may be divided into heat cumulus and the clouds of the rear of a cyclone1 so cumulo-nimbus may be divided into the same two groups4 ,in the case of the heat thunder-clouds the instability of the air is effected by the rapid heating of its lower layers in contact with the ground1 those lower layers being so quickly warmed that there is not time for them to become mixed with the overlying air in which the rate of decrease is normal4 ,if there is much wind we rarely get cumulo-nimbus1 because the heated air is mixed mechanically with the overlying parts1 and the rate of decrease is approximately normal throughout4 ,calm air and hot sun are then one set of necessary conditions for the production of instability4 ,but it is well known that thunder-showers and lesser examples of cumulo-nimbus are by no means infrequent in the rear of a cyclone1 and such storm clouds are usually attended by considerable wind4 ,they are1 as a rule1 much smaller than those produced by heat1 but they have the same form1 and are evidently due to instability in the lower part of the air1 and the question is how can that condition be produced4 ,in order to find the answer it is necessary to refer to the temperature phenomena of a cyclonic area4 ,if a cyclone be divided into four quadrants by two lines drawn through its centre1 one in the direction in which the system is travelling and the other at right angles to it1 then the front right-hand quadrant is the warmest and the rear right-hand quadrant much colder4 ,the cumulo-nimbus clouds of a cyclone are limited to the first part of this cold quadrant1 that is to say1 to the portion of the storm in which a great volume of cold air is flowing over a district which has just been warmed and wetted by the preceding part4 ,the result is that1 the air being warmed by contact with damp ground at a temperature many degrees above that of the air itself1 we have produced exactly the same unstable state at a low temperature as we have at a high temperature in the case of heat storms4 ,the lower temperature of the whole is enough to account for the smaller volume of the cloud1 and that in turn explains why cyclone thunderstorms are1 generally speaking1 on a much smaller scale than heat storms4 ,the life history of a cumulo-nimbus is easily studied on a suitable day4 ,the rapid heating of the lower layers of air causes them to expand bodily1 and as they do so they lift the overlying air1 frequently in broad domes or waves4 ,the first result is the expansion of these upper zones1 which are lightened by the flowing away of still higher layers4 ,expansion means chilling1 and sooner or later its effects become visible in the formation of cirro-stratus1 cirro-cumulus1 or alto-cumulus4 ,simultaneously the heated air near the ground begins to rise up in tall columns1 while the cooler air from a little higher descends to take its place4 ,soon patches of lower cloud appear1 at first hazy and indistinct1 but gradually shaping themselves into cumulus with hazy base and rounded summits4 ,these rapidly assume the typical pyramidal shape1 with level base and sharply contoured top1 and so far there is little to distinguish them from an ordinary cumulus 7see ,plate #dh74 ,but watch them carefully4 ,here and there some will be growing taller than their fellows1 and as they grow their rate of growth increases until the top begins to show signs of spreading outwards4 ,rapidly the bulging summit throws out long fingers of cloud1 radiating from the central column almost as if propelled by some repulsive force4 ,at first1 these fingers are merely projecting lumps of cloud with rounded ends1 but in a few minutes they undergo a sudden and striking change4 ,the whole summit becomes frayed out1 drawn out into long radiating lines1 which thin off against the blue sky exactly like the edges of a sheet of cirro-stratus4 ,false cirrus is the name commonly given to this1 but there seems no valid reason why it should be regarded as 8false40 ,the top of the cloud rapidly spreads horizontally1 forming a disc of cirriform cloud1 which sometimes spreads several miles ahead of the rest of the storm4 ,meanwhile1 the original cumulus column loses all its deep folds and convolutions1 and other round-topped cumulus arise around it until the completed system consists of a more or less disc-shaped mass of cumulus1 with a common base1 rising higher and higher towards some central point1 where these are connected1 by an uprushing column of vapour1 to an upper disc with cirriform margins4 ,illustration3 ,plate #dh4 ,,thunder-,,clouds ,,forming4 ,in ,plate #di we have on the left hand a specimen in which the outspreading is just beginning1 and the same cloud is shown half an hour later on the left of plate #ej4 ,a complete cumulo-nimbus in full work is shown on the right of ,plate #di1 and the same appears on the right in ,plate #ej4 ,illustration3 ,plate #di4 ,,thunder-,,clouds4 7,cumulo-nimbus47 ,illustration3 ,plate #ej4 ,,thunder-,,clouds4 7,cumulo-nimbus47 ,these clouds were thunder-clouds1 the larger one being a smart thunderstorm with heavy hail4 ,they were photographed in the evening1 and in the second picture the sun was just below the horizon4 ,but1 to continue the story of a thunder-cloud1 we always find that after a time the cirriform top flattens out and gradually subsides1 and this is usually accompanied by a descent of the cloud base to a lower level4 ,meanwhile1 it frequently happens that the whole series of phenomena is repeated in one of the attendant cumulus4 ,plates #ea and #eb are also two views of the same cloud at different times4 ,in ,plate #ea we have the main part of the storm on the right1 while on the extreme left a lower part of the cloud is rising rapidly into a tall dome4 ,in ,plate #eb the central top has lost its cirriform margin and has distinctly flattened1 while the left-hand dome has risen much higher and is beginning to throw out the projecting bits4 ,illustration3 ,plate #ea4 ,,thunder-,,cloud4 7,cumulo-nimbus47 ,illustration3 ,plate #eb4 ,,thunder-,,cloud4 7,cumulo-nimbus47 ,the hard-topped cumulus which fringe the lower disc1 and the vast pile of cirriform and hazy cloud which forms the centre of a cumulo-nimbus1 are shown in ,plate #ec1 which represents part of the side of a great thunder-cloud4 ,in this case the diameter of the lower disc was about #ae miles1 and the upper disc was rather larger4 ,the uprising column in the middle was about #g miles across1 and the height from base to summit about #c miles4 ,the whole system contained between #ajj and #aej cubic miles of cloud4 ,when photographed it was over the northern part of ,salisbury ,plain4 ,lightning played repeatedly between the back of the white cumulus and the hazy mass behind it1 and the rumble of thunder was all but continuous for nearly half an hour as the great cloud passed by4 ,illustration3 ,plate #ec4 ,,the ,,flank ,,of ,a ,,great ,,storm4 ,this was an unusually large cloud for this country1 but specimens of #aj1 #bj1 or #cj cubic miles are quite common4 ,now for the explanation of the series of events4 ,to begin with1 we have the production of an ordinary cumulus1 but the equilibrium is unstable1 the growth of the cloud1 therefore1 becomes more and more rapid1 and the rapid condensation adds to the instability until the rising column is so much lighter than an equal column outside that a powerful updraught is created1 strong enough for a time to hold up the raindrops or even hailstones4 ,at length the condensation is complete1 the upper part of the cloud consisting of snow crystals exactly like those of any other cirrus4 ,in the mean time1 the rapid ascending current necessarily involves an indraught from around1 and consequent descending currents to supply it4 ,the result is to set up a circulating system1 moving inwards along the ground1 upwards in the central column1 and outwards in the upper disc4 ,the downward currents are sometimes shown by a curling over of the edges of the upper disc1 but the phenomenon is not often seen1 as the descending movement is generally enough to dry up the cloud particles4 ,the rapid rising of damp air drawn from the ground brings about rapid condensation and heavy rain4 ,the large size of thunder-drops is almost certainly due to the fact that it is only the larger drops which can fall in the teeth of the strong updraught4 ,but when these drops begin to fall1 and still more when cold hailstones begin to fall through this ascending air1 it becomes chilled from top to bottom1 and the column is broken or even stopped altogether4 ,the frozen particles which make up the top subside gradually1 and the chilling of the air immediately below the cloud brings the saturation level nearer the ground1 and we say the cloud base descends4 ,the arrangement of a thunder-cloud into the upper and lower discs with a connecting uprush gives to the typical cloud a shape something like that of an anvil when seen sideways1 but in the larger clouds the disc-like form is more obvious4 ,in any ordinary thunderstorm the great majority of the discharges of lightning play between the two discs1 and the larger the cloud the more frequent these are4 ,such discharges as pass between the cloud and the earth come exclusively from the base of the lower disc if the cloud is large1 and generally follow immediately after or simultaneously with one between the two discs4 ,the phenomena of lightning are intensely interesting1 but the purpose of these pages is confined to the study of clouds and cloud forms1 and it would be going beyond our scope to discuss either lightning or hail4 ,both are1 however1 so closely related to cumulo-nimbus that they can hardly be passed over in silence4 ,one thing is certain1 and that is that neither the electrical developments nor the hail has anything to do with the growth of the cloud4 ,on the contrary1 both are consequences of the cloud1 the hail being due to the great altitude1 and consequent low temperature of the upper part of the cloud1 and also to the violent uprising currents within it2 while the electrical phenomena are due to either the enormous amount of condensation1 or to friction due to the rapid uprush1 or more probably to the fact that considerable differences of electrical condition exist in the distant parts of the air connected by the cloud1 and between which its circulating currents move4 ,these differences are known to exist at all times1 and we cannot here discuss their origin4 ,the formation of cumulo-nimbus and cumulus is dependent upon the presence of a large amount of water vapour4 ,it is worth while to consider whether the atmospheric movements which bring about the condensation could exist without moisture4 ,wherever we find differences of temperature between neighbouring places we must get currents of hot air rising from the warmer spots1 and compensating descending currents around them4 ,but we have pointed out that if the rate of cooling as we ascend in the still air is less than the rate at which an ascending current will be dynamically cooled1 such a rising current will come to rest4 ,if1 on the other hand1 the rate of cooling in the ascending current be less than in the still air1 the equilibrium will be unstable1 and a violent uprush will result4 ,now1 in a climate such as our own1 where the lower regions of the air contain large quantities of water vapour1 any considerable rise brings about more or less condensation1 and that condensation is attended by a liberation of very large quantities of heat1 which retard cooling in the ascending current1 and so facilitate the production of instability4 ,but if this cause is put aside it is still possible to have a similar circulation4 ,when discussing the causes of instability1 it has been pointed out that the prime condition was an unusually rapid rate of fall of temperature in still air1 such as may be produced by hot sunshine4 ,now1 these conditions are exactly those which will give rise to the phenomenon of the mirage1 and which reach their fullest development in great desert districts when the air is still4 ,again1 it has been pointed out that the causes which bring showers and thunderstorms to an end include the chilling of the lower parts of the ascending column by the descent of cold rain or hail from above4 ,we may also add the shading of the underlying ground by the cloud itself1 and the absorption of heat in the partial evaporation of some of the rain during the lower part of its fall4 ,in a desert district the arising currents are so dry that even a very great ascent does not often result in visible cloud2 and when it does1 the cloud is produced at so great a height that the air is too rarefied to produce anything much denser than thin alto-stratus1 from which no falling droplets could reach the earth4 ,it seems1 then1 that there will be no such automatic check on the growth of the circulating system1 and it will go on growing in volume and intensity indefinitely4 ,as a matter of fact1 this is not the case4 ,a different check does come into operation1 but not until the indraught and updraught have become so powerful as to draw up the dust and sand and generate a sandstorm1 the weight and shade of which1 in time1 destroys the circulating currents which uplifted them4 ,since1 however1 condensation is a considerable factor in producing instability1 we should expect that such sandstorms would be rarer than thunderstorms are in an equally hot but well-watered district1 which is the fact4 ,again1 since rain and cloud are checks upon such systems1 we should expect the sandstorm systems to be larger and far loftier than thunderstorms1 and to consist of far more violent atmospheric movements4 ,this also is the case1 and when we know that some of these disturbances have the dimensions of a cyclonic storm1 it is easy to understand how the finest dust may be raised to vast altitudes1 into the great upper currents of the air1 by which it may be borne hundreds of miles before returning to the ground4 ,it is thus that the dust of the ,african deserts is carried across the ,mediterranean to ,europe1 and the yellow loess from ,mongolia even to the eastward of ,japan4 ,,chapter ,,viii ,,wave ,,clouds ,,reference has already been made on more than one occasion to the remarkable rippled or wavy structure sometimes assumed by clouds4 ,the waves may be of almost any dimensions1 from the broad bands into which a sheet of cirro-stratus or of alto-stratus is sometimes divided1 down to the most minute ripples4 ,sometimes they are ranged in long straight lines1 sometimes they are bent into sharp angles1 and sometimes curved in very elaborate patterns2 but whether they be large or small1 straight or curved1 no one can see them and fail to conclude that they must be due to an action more or less analogous to the causes which produce waves on the sea or ripple marks upon the sand4 ,wave clouds occur at all heights where clouds are formed4 ,the break-up of a lifting fog into roller clouds is probably the lowest example1 but it may more frequently be seen in higher clouds of the alto or cirrus kinds4 ,a low example is given in ,plate #dj1 which represents stratus maculosus1 and which has already been described4 ,a higher type is shown in ,plate #ed1 which is a wave-like arrangement of alto-cumulus4 ,rather higher come the long zig-zag bands of ,plate #ee1 in which the stratiform arrangement is more obvious1 and which would be best described as a wave-form of alto-stratus4 ,these two plates form striking contrasts4 ,the clouds shown in the first are distinctly of the cumulus order1 and a prominent feature is the way in which the right-hand side of each wave has a clear-cut rounded contour like that of the upper edge of a small cumulus1 while the left-hand edge of each band is frayed out into a ragged fringe4 ,the whole cloud was moving slowly in a direction nearly1 but not quite1 at right angles to the waves1 and the fringed edge formed the rear4 ,it is evident that this peculiar structure must be due to a series of narrow waves intersecting a plane in which the air is just on the point of producing alto-cumulus4 ,if there were no such waves1 the little uprising currents1 with their intervening down currents1 would be irregularly distributed1 and all the wave disturbances have had to do is to arrange them4 ,the consequence is that as the waves pass along the stratum the air is alternately raised and lowered4 ,where it is rising condensation takes place1 where it is falling evaporation results4 ,illustration3 ,plate #ed4 ,,crested ,,alto ,,waves4 7,alto-cumulus ,undatus47 ,illustration3 ,plate #ee4 ,,alto ,,waves4 7,alto-stratus ,undatus47 ,the cloud1 like most other wave clouds1 did not retain its features for any length of time1 but the gaps closed slowly in as the cloud-bands increased in size1 until a sheet of alto-stratus was produced4 ,since the time of day was the morning1 it is almost certain that the plane of saturation was rising in accordance with the general law1 which is that the planes of condensation rise steadily1 until about two or three o'clock in the afternoon1 and then slowly descend4 ,in ,plate #ee each band is much flatter and less dense4 ,they are just as evidently formed by wave movements intersecting the plane of condensation2 but this was formed in the evening when the sun was nearing the horizon1 and at a time when the cloud planes are as a rule rapidly descending4 ,among the alto clouds wave-forms sometimes persist for a fairly long time1 and in this case the bands moved steadily onward in a direction equally inclined to their length and breadth1 that is to say1 from the bottom left-hand corner of the photograph to the top right-hand corner4 ,as they passed across the sky new bands kept on making their appearance at about the same spot1 each band persisting with little change until it had passed out of sight4 ,going much higher up into the region of cirrus1 we meet with the most minute and delicate ripple clouds4 ,some of these have already been referred to4 ,they are connected with either cirro-macula1 cirro-cumulus1 or cirro-stratus1 just as the coarser textured waves we have been considering are connected with alto-cumulus or alto-stratus4 ,in ,plate #ef we have an example in which we can see the stages in the process4 ,nearest to the zenith we have cirro-cumulus1 which is here and there irregularly distributed1 but is generally arranged in delicate ripples1 which are variously curved4 ,nearer the horizon the troughs of the waves are filled in1 and sheets of cirro-stratus are the result4 ,here1 again1 the wave-form is evidently not typical4 ,it is an arrangement of either cirro-cumulus or cirro-stratus1 produced by the intersection of the plane of condensation by a series of wave movements4 ,illustration3 ,plate #ef4 ,,cirro ,,ripples4 7,cirro-cumulus ,undatus47 ,the arrangement is1 however1 so striking a feature when it is well shown that any description of the cloud which contains no reference to the waves is manifestly incomplete1 and this would be best effected by adding the word undatus or waved to the name of the cloud4 ,plate #ed will then be alto-cumulus undatus1 ,plate #ee alto-stratus undatus1 and ,plate #ef would be described as cirro-cumulus undatus1 passing into cirro-stratus undatus and cirro-stratus4 ,in popular language ,plate #ee might be called alto waves1 ,plate #ed crested alto waves1 and ,plate #ef cirro ripples4 ,if we are satisfied that the wave clouds are due to a wave movement intersecting a plane of incipient cloud formation1 the whole question of their mode of production resolves itself into two parts--how is that plane of incipient condensation produced8 and how can we account for the intersecting waves8 ,the first question has by far the greater importance1 since it amounts to asking for a general explanation of the production of high clouds1 especially the forms of cirro-cumulus1 cirro-stratus1 cirro-macula1 and the corresponding alto varieties4 ,there are1 again1 two divisions also to this question4 ,how does the water vapour reach the stratum in sufficient quantity to saturate it8 and when condensation takes place1 why does it so frequently assume the characteristic mottled and granular forms like crowds of little cumulus clouds arranged in one level8 ,this last sentence gives the clue4 ,they are1 in truth1 little cumulus clouds1 and must be formed in exactly the same way as their vastly larger prototypes of lower regions4 ,it has been explained that low cumulus is the result of large upward moving air columns or convection currents1 each one being initially caused by the heating of the vapour-laden air near the ground1 and each uprising column being supplied by cooler descending air which flows down in the intervening spaces4 ,it has also been explained that these movements result in changes of temperature1 which tend to check those movements and restore the original equilibrium4 ,suppose this to occur1 as it constantly does1 without any column reaching sufficiently high to produce a cloud4 ,there will be no visible effect1 but1 nevertheless1 an important change has taken place4 ,every ascending current has lifted some water vapour with it to a higher level1 and the descending drier air has come down in contact with the ground or damper air to become equally charged with moisture in its turn4 ,the process will be repeated again and again1 and at one level after another1 so that the water vapour travels ever higher and higher4 ,this process of interchange between ascending and descending air has been called by ,mr4 ,ley inversion1 but the term does not seem very suitable1 and interconvection would be better4 ,the two opposite currents pass through each other1 as if the ascending air gathered itself into definite channels1 and passed through holes in the descending mass like the passage of water upwards through a descending plate of perforated metal4 ,moreover1 just as the holes in such a descending plate might have any size1 so that the ascending streams might vary in breadth from the finest hair to a column of huge diameter1 in exactly the same way the ascending columns of air may vary from the smallest imaginable size to the great cumulo-nimbus currents4 ,it is the little currents which account for the constant quiver of the margins of any object which is viewed through a large telescope by day1 and for the haze1 so characteristic of a hot day1 which makes distant objects seem ill-defined and in a state of continual tremble4 ,the rays of light in passing through the intersecting streams are bent a little1 now this way1 now that1 as the air currents sway to and fro4 ,the near neighbourhood of the ground is not essential4 ,as long as the temperature of the air at any level is rising1 so long interconvection must occur4 ,the process will be independent of the presence or absence of wind4 ,all that wind can do is to mix up the air at different levels1 breaking the system of currents and reducing it to1 so to say1 a finer texture1 or producing eddies1 if strong enough1 which direct the currents and gather them into definite channels4 ,the final result in any case is that1 with rising temperature1 water vapour is steadily borne upwards from the ground4 ,as it ascends the air becomes cooler1 and yet retains its water vapour4 ,when the rising currents are large they mix little with the descending dry air1 and on reaching a certain level condensation takes place1 and we have the beginning of a cumulus4 ,if they are of a more moderate size they will ascend less rapidly1 the admixture with descending air will bear a larger proportion to the whole1 and the plane at which condensation will begin will be higher1 and then each small column will be tipped with a ball of alto-cumulus4 ,make the interconvection currents smaller still1 and the cloud plane will be lifted yet higher1 and we shall have cirro-cumulus or cirro-macula4 ,now1 the more even the distribution of temperature on the ground the less the probability of coarse interconvection1 and the same is true of any higher stratum of air1 provided it is free from disturbing influences from outside4 ,if1 therefore1 we have large currents near the ground1 ending1 as they must1 in cumulus1 it has already been explained that these clouds stop the action1 and the general system of large currents will be restricted to the region in which they occur4 ,at some distance above the lower clouds the only difference will be that water vapour has been brought up to their level in great abundance4 ,smaller systems of interconvection can then exist1 and so we may have the spectacle of several layers of cloud--cumulus capping the great currents of lower regions1 alto-cumulus forming the summits of the smaller currents of intermediate regions1 and cirro-cumulus floating far above both4 ,frequently it happens that before the ascent of vapour has gone quite far enough to produce a cloud1 other causes co-operate1 and the cloud makes its appearance suddenly over considerable patches of sky4 ,the most potent of these is a fall of the barometric pressure1 which is brought about by some of the air far above the region of even the highest clouds flowing away to some other district4 ,the air at all lower levels being thus relieved of the superincumbent pressure1 immediately expands1 and is thereby cooled throughout4 ,consequently1 if at any level it was near its point of saturation1 it will be carried beyond that point1 and cloud will rapidly make its appearance over a large part of the sky1 possibly at more than one level4 ,stratiform arrangements will be the rule2 but if interconvection is going on at the time1 its presence will be betrayed by a granular or cumuloid structure4 ,interconvection clouds should then be most frequent1 and best formed when the air as a whole is still or moving slowly 7so as not to create great eddies71 when the temperature is rising rapidly1 and when the barometer is making a sudden fall4 ,all these conditions are met in thunder weather1 and at the time when a summer anticyclone is giving way4 ,it will be remembered that many of the most beautiful forms have been described as forming under one or the other of these very conditions4 ,a second contributing cause1 and one which tends to make the condensation in patches or long broad bands ranged roughly at right angles to the direction in which the air is moving1 has been referred to earlier4 ,it is the passage of the air over an undulating country2 the up-and-down movements of the lower air being transmitted upwards to great altitudes1 as ever broadening and flattening waves4 ,if the upper air is flowing more rapidly than the lower1 these broad waves may be far ahead of their real cause1 which will1 therefore1 quite escape recognition1 but the phenomenon is constantly to be detected in the arrangement of the lower clouds4 ,two instances in the writer's experience will suffice4 ,it was desired one morning to measure the altitude of some small clouds which were passing from the north-west at a height of probably between #bjjj and #djjj metres1 over a hill only about #aej metres higher than the valley in which the apparatus was fixed4 ,in order to make the measurement1 it was necessary for the cloud to cross the valley and appear in the same field of view as the sun1 according to the method that will be described further on4 ,but in order to cross the valley the air had to descend1 and so1 of course1 had the cloud stratum1 though to a less extent4 ,but small as the descent was1 it was enough to dry up the clouds entirely1 and for more than a couple of hours the clouds came sailing over the hill1 disappearing entirely1 and then reforming so far beyond that no measurement was possible1 since not one single fragment came near enough to the position of the sun1 which remained shining brightly through a broad clear gap between two patches of cloud-strewn sky4 ,on another occasion considerable preparations had been made for some photographic observations during an eclipse of the sun4 ,the observatory stands on the eastern side of the valley of the ,exe1 which is flanked on its western side by a long ridge of hills going up to #hjj feet above the sea4 ,beyond these hills lies the deep1 narrow valley of the ,teign1 and beyond that the granite ramparts of ,dartmoor1 #ajjj feet above the sea4 ,the wind was blowing gently across the two valleys1 and shortly before the eclipse began a broad strip of thin cloud formed above and rather towards the eastern side of the ,exe valley1 just where the sun was1 while at the same time the sky was practically clear half a mile further east1 and bright sunlight was streaming down on the ridge between the two rivers a few miles towards the west4 ,the cloud was never thick enough to quite hide the sun1 so that the eclipse was easy to watch with the naked eye2 but in spite of fairly rapid movement of the cloud masses as they drifted before the sun1 they kept on forming in just the same place1 and completely prevented the carrying out of the programme planned4 ,it is almost certain that the phenomenon was brought about by an upward moving wave marking the place where the level of approaching saturation was upheaved by the disturbance caused by crossing the two valleys and intervening ridge4 ,these two instances are not quoted as examples of a rare occurrence1 but as definite simple instances of a phenomenon which may be constantly observed1 and as proof that the conformation of the ground does exercise an influence upon the distribution of cloud4 ,but no irregularities of the ground will suffice to explain the minute waves and ripples which have been described at the beginning of this chapter4 ,these must be due to wave disturbances in the air itself4 ,they have been explained as due to two different currents of air1 either a warm damp current flowing over a cold one1 or vice vers4 ,now1 such an occurrence as a warm damp current flowing over a cold one must be very rare1 though it is impossible to deny that it might occur4 ,the immediate contact of a cold current above a warm damp one is equally unlikely1 unless the general atmospheric condition were greatly disturbed1 which is the same thing as saying that wave clouds would not occur4 ,they are most frequent at just those times when interconvection has freest play1 and this is amply sufficient to account for a plane of saturation without any necessity for a hypothesis of two layers of air at different temperatures all but producing cloud at their junction4 ,no convincing evidence of cloud production by such means has yet been adduced1 and it is better to rely upon causes which we know do operate than to call in theories as to what might possibly happen4 ,this is one of those points in the study of clouds which need investigation1 and until proof is forthcoming it is better to say that the admixture of two strata of air might conceivably produce cloud1 but most forms can be accounted for by other causes of which we have more positive evidence4 ,still1 the wave clouds are due to waves1 and there seems no other way of accounting for them than the supposition of gentle differential currents4 ,but if such currents occur the ripples and waves will not be limited to a definite surface1 so to say1 of contact1 but will be propagated upwards and downwards for considerable distances from the level of greatest disturbance4 ,whether1 therefore1 the level at which the natural operation of interconvection has produced saturation is high or low in this region1 the result will be the marshalling of the ascending and descending elements of the convection system in the characteristic waves4 ,the differential currents1 then1 which cause the waves must not be conceived as producing those waves at a surface of contact1 nor must the currents be thought of as separated by any definite surface1 but rather by a region of variable but usually considerable depth1 in which the air is disturbed by a series of small slow eddies and oscillatory movements4 ,when the waves are parallel straight lines the air currents may be really portions of a whole1 having the upper part more rapid than the lower4 ,in such a case the direction of movement should be at right angles to the cloud lines4 ,if the upper current differs in direction as well as velocity1 the direction of movement of the clouds will be intermediate1 and will resemble that of the upper or lower current1 according to their relative distances from the plane at which the clouds are formed4 ,the behaviour of the clouds will depend upon the relative shares in their production borne by interconvection pure and simple and by the wave oscillations4 ,if the stratum is one in which cloud would actually be formed independently of the up-and-down movements1 all this will be able to do will be to arrange the cloudlets at their birth1 and these will then continue to exist1 drifting with the general horizontal movement of the air like any other cloud of the same order4 ,on the other hand1 if the production of cloud is dependent upon the vertical oscillations1 the cloudlets or lines of cloud will move with the air waves1 and their rate of motion and direction of motion will be determined by the rate and direction of the waves1 which may be quite different from that of the air at that stratum as a whole4 ,the ascending waves will be marked by lines of cloud generally rounder and better defined on their advancing sides1 while the descending troughs will be marked by clear intervals4 ,wave movements of the necessary kind are frequently very complicated1 and it is not by any means a rare occurrence to see the wave lines in one part of the sky at all sorts of angles with similar lines in other parts1 or even to see two or more sets of waves at different altitudes crossing one another4 ,either phenomenon is always accompanied by rapid changes in the cloud1 and the rippled structure is short-lived4 ,this was the case with the clouds shown in ,plate #ed4 ,plate #ec1 on the contrary1 shows great uniformity in the wave lines1 and although the vertical oscillation is probably the main cause of condensation1 the form was unusually persistent4 ,irregular patches of wave disturbance1 affecting a plane occupied by cirro-stratus vittatus1 are shown in ,plate #eg4 ,in this case the wave systems only touch the cloud plane here and there1 and the places of contact varied rapidly4 ,it is pretty clear from this photograph that the idea of the waves being formed at a surface of contact between two diverse currents will not suffice4 ,the bands of the cirro-stratus are for the most part unbroken and unaffected2 it is only here and there that the wave region touches them4 ,illustration3 ,plate #eg4 ,,waved ,,cirro-,,stratus4 7,cirro-stratus ,undatus47 ,the conclusions at which we have arrived are simple1 and there is little room for doubt as to their main correctness1 but there are numerous minute features presented by these beautiful cloud patterns which await interpretation1 and they reveal complicated oscillatory movements in the air which are difficult to account for1 whether we seek their originating causes or the mechanics of their motions4 ,,chapter ,,ix ,,cloud ,,altitudes ,,during an extended experience of cloud photography1 it was found that it was quite possible to get pictures which showed the cloud detail even when the sun was in the field of view4 ,sometimes the solar image was reversed1 but if the exposure was very short this was not the case4 ,in such photographs the structure of the cloud was exceedingly clear and sharply defined quite close to the sun4 ,indeed1 the intense illumination seemed to reveal minute details of internal arrangement which could not be detected in similar clouds some distance away4 ,the methods which had been employed for the measurement of cloud altitudes elsewhere have already been briefly referred to4 ,some of them required two observers1 who were equally responsible1 each of them having to direct his apparatus or camera to the same point of the cloud1 and to record the exact direction in which the instrument was pointed4 ,the instruments1 if accurate1 were costly1 and there were many opportunities for error in reading the graduated circles which gave the directions4 ,moreover1 in most of these methods the two observers were connected by telephone1 and had to agree on the exact point towards which their instruments should be directed2 either the exact point of the cloud1 or the precise direction as shown by the mounting of the camera or other instrument4 ,at ,kew some of these sources of error were avoided by fixing the two cameras with the axes of the lenses and centres of the plates in a vertical position and exposing the two plates simultaneously4 ,the ,kew observations were not long continued1 and for some years the only measurements in progress were those carried out abroad1 particularly at the ,blue ,hill ,observatory and at ,upsala4 ,the experience gained in photographing clouds in order to record their forms suggested a way in which many of the sources of error in previous measurements of altitude could be avoided1 especially by simplifying and reducing the operations at the moment of making the observation4 ,if two cameras are placed at the opposite ends of a measured base line1 whose direction is known1 and if they are both pointed towards the sun1 on making the exposures by electrical means at the same moment1 the position of the image of the sun upon the plate gives the direction in which the cameras are pointed4 ,it will be in the same direction as seen from both ends of the line4 ,now1 if we note the time at which the exposure is made1 this with the date gives all that is required for ascertaining the sun's position in the sky1 and is1 therefore1 the only exact observation which need be made at the time of taking the photographs4 ,mistakes are almost impossible1 as each plate contains its own record of the sun's position1 and even if some of the plates should get mixed the images of the clouds will generally suffice to pair them properly4 ,for general measurements there is one grave defect in the method1 and that is that it can only be used when the sun and cloud can be got into the same field of view4 ,but with the higher varieties of cloud this is generally possible1 and it was just these higher sorts about which knowledge was least certain1 and which it was proposed to study4 ,an initial difficulty was the finding of a level site1 flat land being very uncommon in ,devonshire1 but fortunately a suitable place was found in some artificially levelled ground close to ,exeter1 belonging to the ,london and ,south ,western ,railway ,company4 ,it was a stretch of ground intended to be covered with sidings1 but had not been finished1 and had become overgrown with grass1 stunted sallows1 and other wild plants4 ,being railway ground1 it was1 comparatively1 though by no means entirely1 free from mischievous and inquisitive people4 ,the next point was a suitable camera4 ,it must have fairly long focus in order to give a large image1 and therefore large displacement2 it must be capable of being pointed in any direction and clamped there2 and it must be capable of standing considerable extremes of temperature and variations of dampness1 as it was intended that they should be kept on the spot in wooden structures1 which served for stands as well as to contain the apparatus4 ,the pattern finally decided upon is represented in ,plate #eh1 which shows one of the cameras pointed up to the sky and standing on one of the stands4 ,these cameras were to take plates of whole plate size1 two double dark slides of the ordinary pattern being attached to each4 ,illustration3 ,plate #eh4 ,,camera ,,for ,,measuring ,,altitudes4 ,the camera looks rather complicated1 but it is really simple4 ,its body consists of front and back1 each attached to a central part by a short bellows and sliding on a base board1 to which it can be clamped by screws of the usual pattern4 ,the central part carries trunnions1 such as are used for looking-glasses1 which swing in sockets carried by two upright supports1 so as to give the whole free motion in a vertical plane4 ,in order to be able to fix it firmly at any angle1 the base board of the camera body carries on its underside a thin board projecting beneath it and forming a segment of a circle whose centre would be the horizontal axis through the trunnions4 ,the board passes between the jaws of a small wooden clamping vice in front1 which is carried by the square base to which the uprights are fixed4 ,the whole is firmly made of well-seasoned pine1 and has stood well the hard usage of half a dozen years4 ,there is no focusing screen4 ,focusing was done with great care once for all1 and then a coat of hard varnish was put over all the adjusting screws4 ,a small view-finder is attached to one side1 and it was by this that the camera was pointed in the desired direction4 ,in order to lessen risk of mistake1 it was so arranged that the two slides belonging to one camera would not fit the other4 ,the lenses1 of #ah inches focus1 and giving sharp detail all over the plate1 were carefully matched1 and the focus adjusted until the images given by them when placed side by side appeared to coincide exactly4 ,they were provided with iris diaphragms1 which were shut down to an aperture of a quarter of an inch1 and with shutters which could be released at the same moment by an electric current1 acting through the electro-magnet shown under the lens on the front of the camera4 ,the shutters were of the kind known as the 8,chronolux10 which will give any exposure from the sixty-fourth of a second up to three seconds4 ,but it was found in practice that the highest speed was sufficient and gave satisfactory results4 ,of course1 there was no idea of adjusting matters on each occasion so as to get the best possible negatives capable of yielding good prints4 ,measurement was the object1 and if the negative showed the sun and sufficient cloud detail for the identification of cloud points1 that was all that was wanted4 ,the shutters gave a good deal of trouble at first4 ,their sliding parts were made of ebonite1 and when the cameras were left in their stands with an ,august sun shining down upon them1 everything inside got very hot and the ebonite warped2 but the difficulty was got over by substituting aluminium4 ,the two camera stands were placed #bjj yards apart1 and were connected by a line of telegraph wire carried on short poles4 ,at each end of the wire an insulated connecting piece was brought down to the camera stand1 and to the batteries and other apparatus4 ,the current which was sent through this wire by pressing a contact at one end of the line did not directly make the exposures2 but two similar relays were brought into action1 and each of these sent the current from a local battery of ,leclanch cells through the electro-magnet on the camera and made the exposure4 ,after development the two negatives showed the image of the sun1 not far from the centre of the field of view1 and the cloud whose altitude was required4 ,since this was taken from two different points of view1 the negatives were not alike1 but the distances between the centre of the sun's disc and any special point of the cloud were different4 ,for instance1 if the cloud were east of the sun1 with its edge just apparently touching the solar image as photographed from the eastern station1 then the negative taken from the western end of the base would show an interval of clear sky between the two1 which would be greater as the cloud was lower4 ,it often happened that after developing the plates the image of the sun was lost in a black blur1 but it was easy to reduce this part of the image by local application of a reducing agent#c by means of a paint-brush1 until the disc became clear enough4 ,two lines were then drawn on the negative1 one vertical and the other horizontal1 intersecting each other at the centre of the sun's image4 ,these lines served as the starting-points for exactly measuring the distance from their point of intersection to any selected point of the cloud4 ,the distances could generally be determined to a fiftieth or a hundredth part of an inch1 and their difference was1 of course1 dependent upon the direction of the sun relative to the base line and the altitude of the cloud1 but for low level clouds the difference was sometimes so great that no pair of corresponding points could be detected1 while it was often as much as an inch4 ,with higher clouds the differences were smaller1 but unless the sun was very low in the sky1 either east or west1 the displacements of the cloud image were great enough to give reliable measures4 ,specimen prints from pairs of negatives are shown in ,plates #ei and #fj4 ,illustration3 ,plate #ei4 ,,print ,,from ,a ,,negative ,,used ,,for ,,measuring ,,altitude4 ,illustration3 ,plate #fj4 ,,pair ,,of ,,prints ,,showing ,,the ,,displacement ,,of ,,the ,,cloud4 ,the processes by which the measurements are worked out are laborious1#d and consist of two parts1 the first being the determination of the exact position of the sun from the date1 hour1 and latitude and longitude of the place1 and the second1 the determination of the position of the cloud4 ,two points which represent the same part of the cloud are selected1 and their respective distances from the two lines drawn through the sun are measured as accurately as possible4 ,now1 a certain distance on the negative corresponds with a definite angular displacement1 and a scale can be constructed showing how much should be added to or subtracted from the sun's position to get the exact position of the cloud4 ,this being done1 it is then a simple piece of trigonometry to deduce the actual height of the cloud above the place of observation4 ,the work of computation1 however1 was greatly lightened by the fact that many of the pairs of negatives showed more than one layer of cloud2 thus ,plate #ei1 which is a fair specimen1 shows three layers1 and1 consequently1 one determination of the sun's position sufficed for three distinct results4 ,for the highest clouds the displacements were1 of course1 small1 and could only be made with certainty of a correct result within about three hours of noon4 ,earlier than #i a4m41 or later than #c p4m41 the sun was too nearly in a line with the two stations1 or too low in the sky1 to give a sufficient displacement of image4 ,a base line of #djj yards instead of #bjj would have been better for the high clouds4 ,but1 on the other hand1 when low level clouds are viewed from two different spots their outlines may seem so changed that it may be impossible to identify a pair of corresponding points1 and the same difficulty may also arise when high clouds are seen through a gap in a lower stratum4 ,the longer the base line the more frequent and more obtrusive would this perspective difficulty become1 so the distance of #bjj yards between the stations was adopted as a convenient mean4 ,the method of making the observations was simple4 ,each observer was provided with some signal flags1 by which the necessary communications were made in accordance with a simple code4 ,call the two observers ,a and ,b1 and suppose ,a directed the operations4 ,he watched the sky until a favourable opportunity seemed to be approaching4 ,he then signalled to ,b1 and both cameras were turned to the sun1 the dark slides were inserted1 the shutters set1 and everything made ready4 ,signals were then interchanged1 to signify that preparations were complete1 and when ,a saw that the edge of the cloud had reached a suitable position to be in the same field of view with the sun1 the contact key was pressed and the plates simultaneously exposed4 ,at the moment when this was done the time was noted4 ,several observations were thus made in a short time4 ,measurements were carried out as opportunity allowed over four consecutive seasons1 from the beginning of ,april until the end of ,october4 ,during the last of the four years1 the site had become less convenient owing to an extension of the railway work1 and early in ,november the series was brought to an abrupt conclusion by a heavy gale1 which snapped off all the poles carrying the connecting wire4 ,but by that time #dbc measurements had been obtained1 the great majority of which referred to clouds of the cirrus and alto groups4 ,the general results may be tabulated thus1 giving heights in metres3-- ---------------------------------------- ------------------- ,number of ,maximum ,minimum ,mean observations4altitude4altitude4altitude4 ---------------------------------------- ,cirrus #eh #bg1dac #d1aad #aj1bcj ,cirro-stratus #fd #ae1ejc #c1hdj #i1edj cumulus #fc #aa1fgi #c1feg #h1fbd ,alto-cumulus #hc #i1cij #a1hbh #e1cdh ,cumulus top #db #d1ehb -- #c1jjf base #dh #a1iei #ehd #a1bij ,strato-cumulus #bg #f1ibf #hbc #b1bdh ,cumulo-nimbus top #ae #f1dji #b1jjd #h1jjb base #ae #b1bhf #gff #a1jde ---------------------------------------- ------------------- ,these values are not very different1 on the whole1 from those which have been arrived at elsewhere1 and in making a comparison it must be borne in mind that there is always a little want of precision in cloud nomenclature4 ,as a whole1 the ,exeter maxima are greater than the foreign ones1 and this is very markedly so in the case of cirrus1 for which the ,american highest record is #ad1icj metres1 the ,swedish record is #ac1cgf1 while the ,exeter value is #bg1dac metres1 or about #ag miles4 ,but this extreme measurement1 and several others unusually large1 were made in one morning1 a day of very hot damp weather1 when cloud formed at seven different levels3 cumulus at a height of #ai miles1 alto-cumulus at #ci miles1 cirro-cumulus at #dg miles1 cirro-stratus 7,no4 #a7 at #h miles1 cirro-stratus 7,no4 #b7 at #if miles1 cirrus at #aae miles1 and cirrus excelsus at #ag miles4 ,by about half-past one in the afternoon the sky was completely overcast with dull grey clouds1 which cleared off at half-past four1 and at half-past five in the evening the cirrus had fallen to #gi miles1 and the cirro-cumulus to #dc miles4 ,if this one day's observations had been omitted1 the ,exeter maximum would only have been little more than #ajjj metres above the record from across the ,atlantic1 but #ajjj metres is a height worth noting4 ,while the ,exeter maxima are all rather greater1 we find the minima for cirrus1 cirro-stratus1 and cirro-cumulus are rather less than at the foreign stations2 that is to say1 that clouds are formed over ,devonshire both at lower and at higher levels than seems to be the case in ,massachusetts or ,sweden4 ,it seems probable that this is due to a greater humidity on our western coasts1 such as we should suppose would be the case from their position and the prevailing winds and ocean currents4 ,if so1 we should expect the great convection clouds to be larger4 ,thus1 at ,exeter1 out of only fifteen examples of cumulo-nimbus1 the top varied from #bjjd metres to #fdji1 with an average base level of #ajde4 ,at ,upsala the maximum was #eigj and the minimum #adjj1 with an average base level of #adjj4 ,the mean thickness of the ,swedish clouds was only #adjj metres1 while that of the ,devonshire specimens was more than #bjjj metres4 ,again and again1 during the progress of these measurements1 it was found that the greatest altitudes and the richest development of the higher varieties occurred towards the end of a spell of fine calm weather1 when convection had had free play day after day4 ,a slight fall of the barometer1 only the hundredth part of an inch1 would usually1 under those circumstances1 bring about abundant formation of high clouds1 frequently of the undatus kind4 ,all the cumulus clouds1 by which we mean to include alto-cumulus and cirro-cumulus1 are most frequent when the levels of condensation are rising1 while the stratiform clouds are an indication of no vertical movement or of active descent4 ,pure cirrus is indicative rather of movement in a horizontal direction1 and may occur when the condensation levels are stationary1 or when they are rapidly changing either way4 ,in broken weather the natural movements of the atmosphere and of its vapour are masked and disturbed by the strong eddies brought by the cyclonic systems4 ,it not unfrequently happens that the region of disturbance does not reach up to the level of the highest cirrus1 or1 what is more probable1 the cyclonic system leans so far forward that we may have in its rear the upper clouds floating quietly far above the comparatively shallow region of disturbance1 while in front the upper part of the storm system projects above undisturbed air4 ,the frequent appearance of cloud almost at the same time at more than one level is at first rather difficult to understand1 but it will be noticed that when this occurs the barometer almost invariably falls4 ,now1 if we suppose that the air is nearly saturated at more than one level1 and that the whole is then bodily relieved of some of the superincumbent mass1 so that the barometer falls1 the mass of air will at once swell up1 being cooled from top to bottom simultaneously1 and wherever it is damp enough cloud will be formed4 ,the converse is equally true4 ,if we have cloud at several levels1 and the whole is compressed by the addition of more air above1 which is the case when the barometer rises1 that compression will be accompanied by the generation of heat and the consequent disintegration and disappearance of the clouds4 ,,footnotes3 #c ,ferricyanide of potassium and hyposulphite of soda4 #d ,from the declination of the sun corrected for variation and from the known latitude1 the meridian zenith distance is calculated4 ,from the ,greenwich time1 the longitude1 and the equation of time1 the hour angle is obtained4 ,now1 if ,h be the hour angle1 ,d the reduced declination1 and ,m the meridian zenith distance1 the sun's altitude may be calculated by the formula-- log versin ,h ,l cos lat4 ,l cos ,d-#bj log n1 where n is a natural number1 and n vers ,m covers alt4 ,again1 to find the azimuth-- vers sup4 7lat4 alt47-vers polar dist4 m1 where m is another natural number1 and log m ,l sec4 lat4 ,l sec4 alt4-#bj log vers azim41 reckoned from the south4 ,hence the position of the sun is ascertained for both negatives4 ,by actual measurements on the plates and reference to a previously constructed scale the position of the cloud as seen from each camera is next determined1 and the angle subtended by the base line at a point ,x vertically beneath the cloud is calculated4 ,if ,a and ,b are the stations1 and a and b the angles from them respectively1 the distance ,,ax is given thus-- log ,,ax ,l sin b-,l sin ,,axb log ,,ab1 and the height h of the cloud above ,x is given by-- log h log ,,ax ,l tan alt4-#aj4 ,,chapter ,x ,,cloud ,,nomenclature ,,since a considerable number of new terms have been suggested in the foregoing pages1 it may be convenient to collect them and tabulate them1 so as to show their relation to those already recognized by the ,international system4 ,in the atlas put forward by the committee1 sixteen varieties are recognized by distinct names1 and these are drawn up in tabular form with appropriate abbreviations for use in making records4 ,the names are-- ,cirrus4 ,ci4 ,cirro-stratus4 ,ci4 ,s4 ,cirro-cumulus4 ,ci4 ,cu4 ,alto-cumulus4 ,a4 ,cu4 ,alto-stratus4 ,a4 ,s4 ,strato-cumulus4 ,s4 ,cu4 ,nimbus4 ,n4 ,cumulus4 ,cu4 ,cumulo-nimbus4 ,cu4 ,n4 ,stratus4 ,s4 ,fracto-cumulus4 ,fr4 ,cu4 ,fracto-nimbus4 ,fr4 ,n4 ,fracto-stratus4 ,fr4 ,s4 ,stratus-cumuliformis4 ,s4 ,cf4 ,nimbus-cumuliformis4 ,n4 ,cf4 ,mammato-cumulus4 ,m4 ,cu4 ,during our survey of these groups we have found that some of them include clouds of many shapes1 which must be due to very diverse conditions4 ,it follows that if observations are to be made on the occurrence of these special kinds1 with a view to arriving at a thorough understanding of the circumstances to which they owe their forms1 it becomes necessary to devise a code of names and symbols whereby an interchange of ideas and records may be rendered possible4 ,specific names have been proposed as each form was considered1 and it only remains to sum them up concisely4 ,subsequent observation1 particularly in other climates1 may show that further additions should be made2 but if the principle of specific names be once admitted1 it will be easy to fill any omission4 ,,group ,,cirrus4 ,under the general head of cirrus we have found nine distinct forms-- #a4 ,cirro-nebula 7,ley7 7,plates #b and #c74 ,cirrus veil4 ,characterized by comparative absence of structure and by the formation of halo4 ,ci4 ,na4 #b4 ,cirro-filum 7,ley7 7,plate #g74 ,thread cirrus4 ,built up of fine long threads1 straight1 curved1 or crossing1 but free from hazy curling or flocculent structures4 ,ci4 ,f4 #c4 ,cirrus excelsus 7,plate #e74 ,high cirrus4 ,characterized by great altitude1 thinness1 irregular branching structure4 ,ci4 ,ex4 #d4 ,cirrus ventosus 7,plate #f74 ,windy cirrus4 ,characterized by curving branches leaning forward in the direction of movement1 and other long curving streamers lagging behind and below4 ,fluffy parts are usually present1 and mark the origins of the long curling fibres4 ,ci4 ,v4 #e4 ,cirrus nebulosus 7,plate #i74 ,hazy cirrus4 ,characterized by the absence of sharply defined lines1 fibres1 or streamers2 all parts of the cloud being hazy1 and suggestive of other varieties of cirrus out of focus4 ,ci4 ,neb4 #f4 ,cirrus caudatus 7,plate #h74 ,tailed cirrus4 ,characterized by small hazy or fluffy heads behind or below which hang long streamers1 which taper away more or less to a point4 ,the tails are sharply defined1 and so are the edges of the heads4 ,ci4 ,ca4 #g4 ,cirrus vittatus 7,plates #ab and #ac74 ,ribbon cirrus4 ,characterized by formation in long bands of cloud1 sometimes made of parallel long fibres with cirrus haze linking them together1 sometimes consisting of a long bundle of fibres1 from which others diverge at an angle as shown in the plate4 ,ci4 ,vt4 #h4 ,cirrus inconstans 7,plate #aj74 ,change cirrus4 ,characterized by a peculiar ragged1 wavy appearance4 ,it is generally only the beginning or the end of a mass of cirro-stratus or cirro-cumulus1 but occasionally it vanishes shortly after its appearance1 without reaching the further stage4 ,ci4 ,in4 #i4 ,cirrus communis 7,plate #aa74 ,type cirrus or common cirrus4 ,characterized by short irregularly curling fibres collected together in considerable patches4 ,no definite arrangement into any of the forms already described4 ,ci4 ,com4 ,,group ,,cirro-,,stratus4 ,under this group the cloud usually shows some structure1 being apparently built up from a massing together of detached forms at a common level4 ,when this is so it should be described by adding the specific name of the detached form most nearly related4 #a4 ,cirro-stratus nebulosus 7,plates #c1 #d1 and #ad74 ,hazy cirro-stratus4 ,characterized by absence of visible structure4 ,ci4 ,s4 ,neb4 #b4 ,cirro-stratus communis 7,plate #af74 ,common cirro-stratus4 ,characterized by the presence of short curling fibres matted together4 ,ci4 ,s4 ,com4 #c4 ,cirro-stratus vittatus 7,plate #eg74 ,ribboned cirro-stratus4 ,characterized by being made up of long stripes or bands of cloud4 ,ci4 ,s4 ,vt4 #d4 ,cirro-stratus cumulosus 7,plate #ag74 ,flocculent cirro-stratus4 ,characterized by an obscurely granular structure4 ,ci4 ,s4 ,cu4 ,many forms of cirro-stratus are arranged in waves or ripples4 ,this is indicated by attaching the word undatus1 or waved1 after the ordinary specific name1 or the letter ,u after the abbreviation4 ,,group ,,cirro-,,cumulus4 ,divisible into three species4 #a4 ,cirro-macula 7,ley7 7,plate #bc74 ,speckle cloud4 ,characterized by semi-transparency1 by the fact that the particles are frequently whiter and more opaque on their edges4 ,a patch of cirro-macula always looks like a thin sheet which has curdled4 ,ci4 ,ma4 #b4 ,cirro-cumulus nebulosus 7,plates #bj and #ba74 ,hazy cirro-cumulus4 ,characterized as rounded balls of semi-transparent cloud1 but ill-defined and hazy4 ,no shadows4 ,ci4 ,cu4 ,neb4 #c4 ,cirro-cumulus 7,plates #ah and #ai74 ,characterized as opaque rounded balls clearly defined1 but showing no shadows on their under sides4 ,ci4 ,cu4 ,com4 ,wave forms again are indicated by the addition of the word undatus4 ,,group ,,alto ,,clouds4 ,divisible into nine species4 #a4 ,alto-stratus4 ,high stratus4 ,a uniform veil of cloud showing no details of structure except local variation in density in patches4 ,rarely dense enough to completely hide the sun1 or even the full moon4 ,a4 ,s4 #b4 ,alto-stratus maculosus 7,plate #cj74 ,mackerel sky4 ,characterized as numerous nearly equal and small lenticular patches ranged on a level and about equi-distant from each other4 ,a4 ,s4 ,mac4 #c4 ,alto-stratus fractus 7,plate #cd74 ,patches and bits of cloud of irregular shape1 but resembling broken bits of a level sheet4 ,a4 ,s4 ,fr4 #d4 ,alto-strato-cumulus 7,plate #cb74 ,intermediate between alto-stratus and alto-cumulus4 ,a4 ,s4 ,cu4 #e4 ,alto-cumulus informis 7,plate #be74 ,characterized as more or less rounded cloudlets interspersed with ragged bits of cloud and occasionally with streaks of cirrus1 the cloudlets showing no clear-cut outlines1 but having distinct shadows4 ,a4 ,cu4 ,in4 #f4 ,alto-cumulus nebulosus 7,plate #bf74 ,hazy alto-cumulus4 ,a4 ,cu4 ,neb4 #g4 ,alto-cumulus castellatus 7,plate #bh74 ,turret cloud4 ,a high cloud resembling a number of tall narrow cumulus clouds on a very diminutive scale4 ,the cloudlets show distinct shadows1 are very opaque1 and their upper margins are sharply defined4 ,vertical axes longer than the horizontal ones4 ,a4 ,cu4 ,ca4 #h4 ,alto-cumulus glomeratus 7,plate #bi74 ,characterized by the roundness and regularity of the cloudlets1 which have sharp margins1 cast distinct shadows1 and have their axes about equal in all directions4 ,a4 ,cu4 ,gl4 #i4 ,alto-cumulus communis4 ,small high cumulus of the ordinary pyramidal pattern4 ,a4 ,cu4 ,com4 #aj4 ,alto-cumulus stratiformis 7,plate #bg74 ,flattened cloudlets gathering into small detached sheets4 ,a4 ,cu4 ,s4 ,lower clouds4 ,,group ,,stratus4 #a4 ,stratus communis 7,plates #cg and #da74 ,in its most typical state1 stratus consists of a sheet of cloud of approximately uniform thickness4 ,the most common form1 however1 does vary considerably1 though usually dense enough to hide the sun4 ,portions of such a sheet would take the same specific name1 unless the portions are very small and ragged1 which would be expressed by adding the word fractus4 ,s4 ,com4 #b4 ,stratus maculosus 7,plate #dj74 ,formed either by the appearance of cloud in lumps1 which are always lenticular in shape1 and ultimately join together to form a stratus1 or by the break up of the typical stratus4 ,s4 ,mac4 #c4 ,stratus radius 7,plate #db74 ,roll cloud4 ,formed during the break up of a low stratus1 which separates up into a number of parallel lines of cloud4 ,s4 ,r4 #d4 ,stratus lenticularis 7,plate #dg74 ,fall cloud4 ,formed by the collapse of cumulus or strato-cumulus4 ,a cloud of evening1 easily recognized as lenticular patches4 ,s4 ,l4 #e4 ,strato-cumulus 7,plates #ch and #ci74 ,a term applied to either a stratus which has thickened every here and there into cumulus1 or a number of cumulus which have joined together so as to show a nearly continuous common base4 ,s4 ,cu4 ,,group ,,cumulus4 #a4 ,cumulus minor 7,plate #dc74 ,small cumulus4 ,cumulus clouds so small as to present the appearance of rounded lumps1 no definite pyramidal form or flattened base4 ,cu4 ,mi4 #b4 ,cumulus major 7,plates #dd and #de74 ,large cumulus4 ,characterized by a flattened base and rounded clear-cut upper surfaces4 ,cu4 ,ma4 #c4 ,cumulo-nimbus 7,plates #di to #eb74 ,storm cloud4 ,characterized by the expanded1 anvil-shaped1 or disc-shaped top1 cirrifying at its edges4 ,,general ,,terms ,nimbus1 a term applied to a cloud from which rain is falling4 ,when the form of the cloud is visible1 the term should be attached to that belonging to the cloud4 ,it may1 however1 be used as a substantive alone when there is nothing to show from what sort of cloud1 or combination of clouds1 the rain is falling 7,plates #ce and #cf74 ,nimbus is either heavy stratus1 massive strato-cumulus1 or a combination of these with stratiform clouds above1 and possibly ragged masses of fracto-cumulus below4 ,n4 either alone or after the sign of the cloud4 ,fracto- is a term placed as a prefix before the name of a cloud to indicate that the cloud has ragged irregular margins1 as if it had been more or less torn to pieces4 ,it is sometimes less awkward to append the word fractus after the name of the cloud4 ,a convenient abbreviation would be to write ,f4 after the name of the cloud4 ,undatus1 or waved1 should always be added to the name of any cloud which shows the arrangement so described4 ,,chapter ,,xi ,,cloud ,,photography ,,reference has been made in the first chapter to the fact that those who wish to make a photographic study of clouds must follow a special course of procedure4 ,for every photographic purpose there is some particular process or some special kind of apparatus which is better fitted for the end in view than any other1 and half the difficulty in attaining success is to find out the best tools and the best methods4 ,there is no difficulty whatever in securing excellent photographs of heavy grey clouds1 or of clouds which stand out dark against a twilight sky4 ,any camera and any plate can be used1 and in an experienced hand will ensure success after a few trials1 but except under these special conditions1 cirrus1 in all its varieties1 the alto clouds1 and even many of the lower ones1 present a real difficulty due to two causes4 ,in the first place1 they and their surroundings are so brilliant that a very short exposure is sufficient1 far shorter than would be needed for a sunlit landscape2 and in the second place1 the actinic value of the light they reflect is very little greater than that received from the background of blue sky4 ,when so minute a difference comes to be represented in the monochrome of the ordinary photograph1 the eye fails to appreciate it1 and all the finer details are lost4 ,now1 if proper care is taken in the development of a negative1 satisfactory results may be attained even if the exposure is twice as great1 or only half as great1 as it should have been to get the best result4 ,but if the exposure is four or more times the best duration1 the negative will generally yield but poor contrasts1 if any result at all can be coaxed out4 ,again1 if the exposure is only a quarter or less of the ideal time1 little or no image will come out4 ,suppose1 now1 we have a brilliant object1 and the correct exposure for the plate and aperture of lens employed should be one-fiftieth of a second2 if we make an error either in judging or in effecting the exposure1 which amounts to one twenty-fifth of a second too much1 we get the negative exposed three times as much as it should be4 ,suppose1 again1 the object is less brilliant1 and the correct exposure should be one-fifth of a second1 an equal error of one twenty-fifth will make little difference4 ,but in photographing cirrus and such clouds1 if we used the same plates and the same lens apertures as we employ for ordinary landscape work1 we should want exposures of the order of those given by a focal plane shutter1 and a mistake either in judging or in making the exposure1 of even the hundredth part of a second1 would be fatal to good results1 and would probably completely spoil the plate4 ,evidently one of our first steps must be to lengthen the correct exposure4 ,there are four ways in which this can be done--by using a slow-acting plate1 by lessening the aperture of the lens1 by putting some transparent screen in front of the lens to shut off some of the light1 and1 finally1 by pointing the camera1 not at the cloud itself1 but at its image in a black mirror4 ,of these1 of course the slow plate and small aperture are the simplest to adopt1 and all the cloud studies shown in the illustrations to these pages have been taken on plates prepared for photo-mechanical purposes or for transparencies4 ,there seems to be nothing to choose between these two brands4 ,orthochromatic1 isochromatic1 double-coated1 and many other special types of plate had previously been tried1 both with coloured filters in front of the lens and without them1 without showing any marked superiority over an ordinary plate of low rapidity4 ,at last the photo-mechanical plates were tried1 and the efforts made to get satisfactory cloud portraits1 which had previously been marked only now and then with satisfactory results1 became uniformly and continuously successful4 ,if the slow plates are exposed in the camera without either a screen or the black mirror1 the diaphragm should be reduced to a small size and the exposure suitably adjusted4 ,the length of exposure may generally be judged by looking at the image on the focusing screen1 and reducing the aperture until the picture shows its detail easily4 ,then1 regarding the picture as that of a sunlit sea or distant landscape1 judge the necessary exposure by the brightness of the image4 ,no definite rule can be given4 ,the light varies enormously from day to day1 and hour to hour1 and especially with the position occupied by the cloud relative to the sun4 ,thus1 working with a lens of six inches focus and an aperture of a quarter of an inch1 the exposure may vary from the quickest snap of a ,thornton-,pickard roller blind to as much as a quarter of a second1 or even more4 ,again1 using a lens of eighteen inches focus and an exposure of a fiftieth of a second1 the necessary aperture might vary from an eighth of an inch up to an inch and a half4 ,but if we suppose that we are dealing with an ordinary bright summer sky between #i a4m4 and #e p4m41 and that the clouds are cirrus or cirro-cumulus1 an aperture of about one thirty-second of the focal length will probably give some sort of image with a snap-shot exposure4 ,at first the failures will be many1 but a little practice will soon enable very respectable pictures to be taken by varying either the diaphragm or the speed of shutter4 ,heavier clouds of the alto types will need rather longer exposure or larger aperture4 ,the lens may be of any kind1 as long as it gives a well-defined image1 but there are many advantages in using one of the rectilinear type provided with an iris diaphragm4 ,a rapid lens is not needed2 indeed1 it has been pointed out that slowness is a very great desideratum1 and if the camera is provided with a rapid lens it must be ruthlessly stopped down4 ,for general cloud purposes the best kind of lens is a wide-angle rectilinear1 but many occasions will present themselves on which a lens of longer focus will be wanted in order to give more insight into the details of some specially delicate clouds4 ,if the lenses are good1 and the focusing is accurate1 enlargements will go a long way towards revealing the minuter structures1 but the results can never be quite so well defined as a direct photograph in a long camera4 ,a shutter will be essential1 and it should be one which opens in the middle1 or which travels across the lens4 ,the shutters which are ingeniously contrived to give more exposure to the lower part of the picture than to its upper part are useless for the purpose in view4 ,it should have some latitude of exposure1 from about one-sixtieth of a second up to a full second or more4 ,then as to the camera4 ,any light-tight camera will do1 and1 as the objects will all be at a great distance1 it may very well be a fixed-focus one1 or may be kept set up and fixed in focus for a distant object4 ,if not1 on setting it up it should be focused on the horizon or most distant object possible1 and not on the cloud itself4 ,as1 however1 the clouds present themselves at all heights above the horizon1 even in the zenith1 it becomes necessary to have some means of pointing the camera in such directions4 ,to a certain extent the ordinary stand does allow of tilting1 but a special support which will allow the camera to be fixed firmly in any position is of the greatest convenience4 ,if the study is meant to be at all prolonged1 the best plan is to make a suitable camera1 once for all1 which can be left in fixed focus1 so as to be always ready1 and which can be directed with equal ease to any part of the sky1 from the horizon to the zenith4 ,if it is intended to use a black mirror1 then a special mount becomes almost essential4 ,many of the most delicate of the photographs reproduced here have been taken with a camera of peculiar pattern1 the structure of which is shown in ,plate #fa4 ,the lens is an ordinary rapid rectilinear1 and the stop used was generally one-sixteenth of the focal length4 ,the shutter is a light slip of aluminium1 which can be drawn across from side to side at any desired pace4 ,the body of the camera is mahogany1 with a bellows part for getting correct focus1 but when once this was obtained the back was clamped to the tail-board and a little varnish brushed over the clamping screws4 ,illustration3 ,plate #fa4 ,,cloud ,,camera ,,for ,,studies4 ,the camera swings on a couple of screws1 which act as trunnions4 ,these pass through two upright arms1 which spring on either side from the base board1 which is attached to the ordinary camera stand4 ,this base board can be rotated into any horizontal position desired1 and the camera can be tilted through any vertical angle by swinging it between the uprights1 and can be clamped by tightening the two trunnion screws4 ,these screws are so placed on the front of the camera that the lens and its attachments on the one side nearly balance the back part of the camera on the other side1 and so lessen the danger of slipping4 ,supported in front of the lens by light brass-work is the black mirror1 made of a very dark glass optically worked on the front face4 ,it is a curious fact that1 although bits of plate-glass blackened on the back seem to the naked eye to give a single image of sufficient truth1 if such a mirror is placed in front of the camera the second faint image formed by reflection from the blackened surface is almost always to be detected4 ,moreover1 the lens with its large aperture at once detects irregularities in the surface of the glass1 which are quite imperceptible through the narrow limits of the pupil4 ,black glass1 with a truly worked surface1 is essential then1 but the surface need not be of the high order of excellence required for mirrors used for telescopic work1 since the first image is not1 as a rule1 intended to be highly magnified4 ,the mirror is held so that its surface makes an angle of about #cc degrees with the axis of the lens1 and the block carrying shutter and mirror can be turned round into any position by slipping it round the lens mount as an axis4 ,the mirror thus always retains the correct angle4 ,the action of the mirror is to a large extent due to mere diminution of brightness1 but it also partly extinguishes the blue light of the sky without exerting any such influence on the white light from a cloud4 ,this is due to the fact that the blue light of the sky is partly polarized1 while that reflected from the cloud is not4 ,now1 polarized light which falls upon a black mirror held in a particular position is not reflected by it4 ,this position depends upon various circumstances1 but one condition is that the reflected ray must make an angle of about #cc degrees with the surface of the glass4 ,the amount of the polarized component of the blue light varies greatly1 but is at a maximum at all points #ij degrees away from the sun4 ,this1 then1 is the best possible position for photographing a cloud1 as the whole of this polarized component may be suppressed by adjusting the mirror to the proper position1 and then the most delicate cirrus fibres stand out brilliant on an almost black background4 ,the black mirror could with some advantage be replaced by a ,nicol's prism mounted between the components of the lens1 so that it could be turned in any position2 but ,nicol's prisms are expensive1 and such an arrangement would cost many times the sum sufficient for an excellent mirror1 and then would narrow down the field of view in a very inconvenient way4 ,with this apparatus exposures of a tenth to a fifth of a second were usually required for high clouds in bright daylight1 while longer times1 up to a second1 might be required under less actively actinic conditions4 ,the exposure having been made1 the next step is development4 ,now1 every practical photographer has his own pet formula1 his own particular favourite among the numerous developing compounds now on the market4 ,it is1 therefore1 rather a thankless task to offer advice as to which should be selected4 ,in all probability as good results may be got by other methods and other formul1 and the description which follows must be understood rather as an account of the process actually adopted1 than advice as to that which should be chosen4 ,the developer used has been always pyro and ammonia1 made up in accordance with the formula-- ,pyro #cj grains ,potassium metabisulphite #cj ,ammonium bromide #cj ,water #aj ozs4 ,but if much work was anticipated the solution was made up in a more concentrated form1 and diluted to this strength of #c grains of pyro per ounce for actual use4 ,the ammonia solution is prepared by mixing #c drams ammonia fortiss4 with #bj ozs4 of water4 ,in developing it is necessary to remember that our object is to make the most of a very small difference in effect4 ,the plate is first flowed over with a mixture of sufficient developer1 with not more than a quarter of its bulk of the ammonia4 ,if the cloud should flash out in a few seconds add more of the pyro solution1 but unless the exposure has been much overdone this will not happen4 ,if the image begins to appear after from thirty to forty seconds it is probable that the best result will be reached by leaving it alone1 but if there is any hanging back of the detail another quarter bulk of ammonia should be put into the glass1 the developer mixed with it1 and the whole returned to the developing dish4 ,if no image appears after about forty seconds1 add more ammonia as above described1 and leave for another forty seconds1 and so on1 until by this method of trial the right quantity of alkali for the particular exposure has been ascertained4 ,the development must never be hurried1 or the background of sky will blacken too soon1 and in some cases it may take a quarter of an hour or more to get enough density on the cloud4 ,but as a general rule the image is fully out in about two minutes1 and the plate is then washed and fixed in the usual way4 ,if a black mirror is used there will seldom be any necessity for intensification1 but if not1 it may frequently be required1 especially for the more delicate kinds of cirrus4 ,indeed1 the image may sometimes be so thin that the common process of intensification by mercury and ammonia does not give density enough4 ,if that seems at all likely to be the case1 it is wiser to use the formula known as ,monckhoven's1 since that simply adds silver to silver instead of replacing the silver image by some other body1 and the process can consequently be repeated more than once1 if sufficient density is not secured by the first application4 ,the formula does not seem to be very often used1 so it may be best to quote it4 ,a4 ,potassium bromide #ajj grains ,mercuric chloride #ajj ,water #aj ozs4 ,b4 ,potassium cyanide 7pure7 #ajj grains ,silver nitrate #ajj ,water #aj ozs4 ,place the washed negative in ,a until it has gone white1 then rinse it well and transfer to ,b1 in which the image turns to a velvety black4 ,after washing1 the process can be repeated4 ,intensification is1 however1 only a way of saving photographs which cannot be secured again4 ,if the first photograph of a particular variety of cloud is not satisfactory1 it ought at least to tell the operator where he had gone wrong1 and a second attempt should produce a better result than any image built up by chemical action on an imperfect base4 ,there is nothing novel in any of these methods1 and there is no doubt that other formul would be as good2 but the one thing essential is to have a developer whose action can be held under control1 and to apply that developer in such a way that very considerable over-exposure will not result in the ruin of the plate4 ,if a number of photographs have been taken in about the same part of the sky1 and within a short time of each other1 then the correct proportions of developer and alkali will be nearly the same for all1 but the first of such a batch will always have to be attacked in the cautious step-by-step method4 ,patience and perseverance1 backed by a steady refusal to be discouraged by the failures which are at first inevitable1 are as certain to be crowned by success as they are in other studies4 ,the workers are few1 and there is much to be done2 for it is mainly to those who will photograph the higher clouds1 and so trace the stages of their growth and decay1 that we must look for the data which will enable us to solve the problems they present1 and so enlarge the narrow boundaries of our knowledge of some of the most beautiful things in ,nature4 ,,references #a4 8,international ,atlas of ,clouds0 7,atlas ,international des ,nuages74 ,hildebrandsson1 ,riggenbach1 and ,teisserenc de ,bort4 ,paris4 #ahif4 ,this is the atlas referred to in the text4 ,the letter-press is short1 and is repeated in ,english1 ,french1 and ,german4 #b4 8,annals of the ,astronomical ,observatory of ,harvard ,college40 ,vol4 ,,xxx4 ,observations made at the ,blue ,hill ,meteorological ,observatory4 ,part ,,iii4 ,measurement of ,cloud ,heights and ,velocities4 ,by ,h4 ,h4 ,clayton and ,s4 ,p4 ,fergusson4 ,part ,,iv4 ,discussion of the ,cloud ,observations4 ,by ,h4 ,h4 ,clayton4 ,this last gives a very concise account of all the different proposals which have been made for the systematic naming of clouds4 #c4 8tudes ,international des ,nuages40 #ahif-ahig4 ,observations et ,mesures de la ,sude4 ,i41 ,,ii4 ,publication de l',observatoire ,mtorologique de l',universit ,roy4 d',upsala4 ,h4 ,h4 ,hildebrandsson4 ,an account of the ,upsala observations referred to in the text4 #d4 ,quarterly ,journal of the ,royal ,meteorological ,society4 ,helm ,wind4 ,marriott4 #ahhf and #ahhi4 ,the ,thickness of ,shower ,clouds4 ,clayden4 #ahhf4 ,methods of ,cloud ,measurement4 ,ekholm4 #ahhh4 ,cirrus ,formation4 ,clayton4 #ahij4 ,nomenclature of ,clouds4 ,hildebrandsson4 #ahhg4 ,abercromby4 #ahhg4 ,wilson-,barker4 #ahij4 ,gaster4 #ahic4 ,scott4 #ahie4 ,a ,new ,instrument for ,cloud ,measurement4 ,ekholm4 #ahic4 ,calculation of ,photographic ,cloud ,measurements4 ,olsson4 #ahid4 ,the ,motion of ,clouds4 ,shaw4 #ahie4 #e4 ,reports of the ,british ,association4 ,reports of the ,committee on ,meteorological ,photography4 ,clayden4 #ahia to #aijj4 ,the reports for #ahif and #aijj refer mainly to the measurements described in the text4 #f4 8,cloudland40 ,clement ,ley4 ,the work in which ,mr4 ,ley set forth his proposed scheme4 #g4 8,a ,popular ,treatise on the ,winds40 ,ferrel4 ,not a 8popular0 work in the usual sense1 but contains lucid descriptions of the mechanics of the atmosphere4 #h4 ,there are many excellent text-books on meteorology1 all of which deal more or less with the movements of the atmosphere and the formation of clouds4 ,,index ,abercromby1 #aj ,aitken1 #c1 #ia ,altitude of clouds1 #ai1 #adi ---- of rain-clouds1 #gf ,alto clouds1 #ei1 #aei ,alto-cumulus communis1 #afa ---- castellatus1 #ff1 #afj ---- glomeratus1 #fg1 #afa ---- informis1 #fd1 #afj ---- nebulosus1 #fe1 #afj ---- stratiformis1 #fe1 #afa ,alto-strato-cumulus1 #gj1 #afj ,alto-stratus maculosus1 #fh1 #afj ,ascent of vapour1 #abd ,atlas1 ,international1 #aa ,band cirrus1 #da ,bidwell1 ,shelford1 #ib ,black mirror1 #ad1 #agc ,blue ,hill1 #bj1 #ach ,cameras1 #adj1 #aga ,change cirrus1 #cg ---- of velocity1 #ce ,cirriform top of thunder-cloud1 #aaj ,cirro-cumulus1 #de1 #aei ---- nebulosus1 #eb1 #aei ,cirro-filum1 #cc1 #aef ,cirro-macula1 #ec1 #aei ,cirro-nebula1 #bf1 #bg1 #aee ,cirro-stratus1 #de1 #aeg ---- communis1 #dg1 #aeh ,cirro-stratus cumulosus1 #dh1 #aeh ---- nebulosus1 #de1 #aeh ---- vittatus1 #aeh ,cirro-velum1 #bd ,cirrus1 #ba1 #aee ---- altitudes1 #cj1 #adi ---- communis1 #dj1 #aeg ---- caudatus1 #cd1 #aef ---- excelsus1 #ca1 #aef ---- inconstans1 #cg1 #aeg ---- nebulosus1 #cf1 #aef ---- ripples1 #ea ---- ventosus1 #cb1 #aef ---- vittatus1 #da1 #aef ,cloud altitudes1 #adi ---- nuclei1 #c ---- photography1 #ag1 #afe ,condensation1 #ic ,cooling by contact1 #hi ---- by expansion1 #ib ---- by mixture1 #ij ---- by radiation1 #hh ,development1 #age ,differential currents1 #acc ,diffraction1 #fc ,dimensions of clouds1 #aab ,dry fog1 #gc ,equilibrium1 stable and unstable1 #ajf ,exeter measurements1 #acg ,exposure1 #afg ,fall cloud1 #ha1 #hh ,false cirrus1 #aaj ,fog particles1 #fj ,fracto1 #afd ,gravitation1 #d ,great waves1 #abi ,halos1 #ba1 #bb1 #bc ,heat cumulus1 #he ---- thunderstorms1 #ajg ,hildebrandsson1 #aj ,hoar frost1 #fj ,howard1 ,luke1 #i1 #aj ,intensification1 #agg ,interconvection1 #abe ,international ,code1 #aa ---- ,committee1 #aj ---- ,system1 #ab1 #ga ,ions as nuclei1 #ib ,irregularities of ground1 #ci ,kepler1 #d ,kew1 #ach ,ley1 ,clement1 #bd1 #cd1 #ee ,lightning1 #aad ,lower clouds1 #ga ,mammato-cumulus1 #ii ,meteorological ,conference1 #i ,methods of computing altitudes1 #adc ,munich1 #i ,newton1 #d ,nimbus1 #gd1 #afc ,nuclei1 #ib ,photographic methods1 #ag1 #afe ,rain-clouds1 altitude1 #gf ----1 thickness1 #ha ,rate of fall of temperature1 #id ,sandstorms1 #aag ,saturation1 #hg ,scotch mist1 #gc ,slow plates1 #afg ,spread of condensation1 #eg ,stable equilibrium1 #ajf ,strato-cumulus1 #gg ,stratus1 #gb1 #afa ---- communis1 #gg1 #afa ---- lenticularis1 #hj1 #ih1 #afb ---- maculosus1 #gh1 #afb ---- radius1 #gi1 #afb ,subsidence of cloud top1 #aaa ,surfusion1 #fa ,swing stand1 #aga ,turreted cloud1 #ff ,tycho ,brahe1 #d ,types1 #h ,umbra and penumbra1 #bi ,undatus1 #afd ,unsaturated cloud1 #aja ,upsala1 #aj1 #ach ,wave clouds1 #dg1 #aai1 #acc ,wilson1 ,c4 ,t4 ,r41 #ib ,,the ,,end ,,printed ,,by ,,william ,,clowes ,,and ,,sons1 ,,limited1 ,,london ,,and ,,beccles4 99 99 99 99 99 ,transcriber's ,note ,variations in hyphenation 7i4e4 thunderstorm and thunder-storm7 have been retained4 ,the following apparent typographical errors were corrected3 ,page #aga1 8focussed0 changed to 8focused40 7it should be focused on the horizon7 ,page #agc1 8aperature0 changed to 8aperture40 7the lens with its large aperture7