J DEW-PONDS By WILLIAM MACDONALD, M.S. Agr., Sc.D., Ph.D. DRY FARMING : THE CONSERVA- TION OF SOIL MOISTURE Illustrated. (Farmers' Library) Crown 8vo, 6s net THE CONQUEST OF THE DESERT Fully Illustrated Demy 8vo, 75 6d net DEW. PONDS HISTORY, OBSERVATION AND EXPERIMENT BY EDWARD A. MARTIN, F.G.S. AUTHOR OF " THE STORY OF A PIECE OF COAL," " A BIBLIOGRAPHY OF GILBERT WHITE OF SELBORNE," " GLIMPSES INTO NATURE'S SECRETS," "NATURE- CHAT," ETC. LONDON T. WERNER LAURIE, LTD. 8 ESSEX STREET, STRAND [All rights rgsevved] \ PRINTED AT THE BALLANTYNE PRESS LONDON 6- EDINBURGH PREFACE ALTHOUGH much has been written about Dew-Ponds during late years, it has been written chiefly from a theoretical point of view. Being favoured with a liberal grant from the Government Grant Committee of the Royal Society, I was enabled to give a three-years* course of experiment to the subject. In preparing this book of results for publi- cation, I have drawn largely on my papers read before the Research Department of the Royal Geographical Society, the British Asso- ciation, and the South-Eastern Union of Scientific Societies, and to those my acknow- ledgments are due. When I commenced my experiments I had a strong leaning in favour of the theory of the 5 PREFACE replenishment of these ponds by dew, but I was soon led to abandon this idea, and, although there is evidence to show that considerable condensation takes place into high-level ponds other than rain, dew has, I submit, little or nothing to do with it. EDWARD A. MARTIN SOUTH NORWOOD CONTENTS CHAP. PACE I. THEIR AGE AND HISTORY 9 II. THEORIES OF DEW-POND ACTION 34 III. VARYING MODES OF DEW-POND CONSTRUCTION 83 IV. EXPERIMENT AND OBSERVATION 112 V. SUMMARY AND CONCLUSIONS 182 APPENDIX 202 INDEX 207 ILLUSTRATIONS TO FACE PAGE SECTION OF FORMER POND NEAR CHOLSEY Frontispiece SHEEP WATERING AT UPPER STANDEAN 36 CHANCTONBURY RING AND POND 60 RlDDLESDOWN POND 60 SH AMBLED EAN BOTTOM POND 114 WATERHALL POND 114 DlTCHLING BOSTEL POND 152 STANMER DOWN POND 152 EWEBOTTOM HILL POND 176 UPPER STANDEAN POND 176 ROTTINGDEAN POND 1 94 SHEEP AT UPPER STANDEAN POND 194 8 CHAPTER I THEIR AGE AND HISTORY THE term "dew-pond" has been applied to certain high-level ponds, the chief charac- teristic of which is that, as a general rule, they do not fail to give a supply of water when other ponds at lower levels have dried up. The source of the water in low-level ponds is apparent. They are fed in general by surface drainage, by brooks and ditches, or by the drainage from a road or high shelving ground around. The sources of supply in these cases are visible, and there, is no doubt as to their origin or their replenishment. In the case, on the other hand, of ponds on high ground, where there are no brooks or ditches, and little if any drainage, certain of them have been found either never to have run dry DEW-PONDS within the memory of man, or only very occasionally. It has appeared, therefore, to those who have used such ponds, that there must be some means of replenishment which is not recognized in the case of ponds at lower levels, and, as the grass on the high ground around such ponds has been observed in the hottest weather to be thickly covered by dew after nightfall, it has been assumed, although not proved, that the ponds also received a deposit of dew, and so, with the customary habit of assuming where proof is absent, such ponds, about which the mystery hangs, have come to be termed " dew-ponds. " All ponds on high ground are not called dew-ponds, but there seems to be no agreement as to what constitutes a dew-pond, even amongst those who use the term for some ponds whilst denying the title to others. I think there are reasonable grounds for believing that certain high-level ponds do owe their maintenance to some other source than the rainfall, although the rainfall is a 10 THEIR AGE AND HISTORY considerable factor in filling all ponds. The point we have to remember is that rain-fed ponds at low levels run dry, as a rule, more rapidly than those on high ground, and this statement has been attested in the past by very many observers, from Gilbert White onward. If this were not a fact, we may assume that the term " dew-pond " would never have arisen, since, as I hold, it was invented to describe those ponds whose replenishment was a mystery, and for whose filling rainfall alone was seen to be insufficient. We start, then, with the postulate that certain high- level ponds maintain their supply in the driest, hottest weather, long after those in the plains below have more or less completely dried up. Investigators into the subject of high- ground water-supplies are rarely at a loss in mountain districts for a solution of the problem before them. As a general rule the higher the ground the more mossy and boggy it becomes. Not only is there greater ii DEW-PONDS precipitation, but the nature of the rocks is such as to bring about, with the aid of the flora which is encouraged to grow thereon, the prevention of complete percolation and loss of rainfall. And even when the water does not collect into tarns, pools, or ponds, the valleys will generally be found to contain running water near at hand on which to draw for a supply. Where the problem has been to discover the source of supply of water for ancient encampments, we find that, apart from wells, even when springs themselves did not rise within their boundaries, as was sometimes the case, the occupants did not, as a rule, have to go far to find the sought- for and much-needed water. Whatever may be the real origin of the water in dew-ponds, a necessity for such ponds was not felt where the soil was of so retentive a nature as to hold up naturally the rainfall, or where the tilting of under- ground strata brought out water by means of springs. Or rather, shall I say, no necessity 12 THEIR AGE AND HISTORY was felt of giving to such ponds a particular name of their own, since the source of their supply was apparent. There was no mystery about the filling up of a mountain tarn. The stream which ran through many a district would obviate the necessity of making ponds. And where these sources were absent, the surface drainage could easily be diverted to feed such ponds as might be made. But in a chalk district things were different, more especially where the chalk reached a height of over 400 feet above the level of the surrounding country, and where no spring would be likely to break out. There is, of course, much comparatively flat chalk- covered country, where natural ponds form, by reason of the covering of the flinty debris, sometimes clayey, and sometimes more or less sandy, which has resulted from the denudation of long-ago existing tertiary beds over the chalk. But the home of the dew- pond is on the chalk heights where there are absolutely no springs, and where the chalk, 13 DEW-PONDS if left alone, absorbs, or at least allows of percolation of all the rain which falls upon it. It is on the higher grounds of the chalk downs where dew-ponds are found, and it is here also where there is frequently but little debris of lost tertiaries to form a soil of more than a couple of inches in thickness, the turf as a matter of fact frequently resting right on the chalk. In such places it was found necessary to give an artificial puddling to the chalk before it could be made to hold water. It was found, too, that finely-divided chalk proved in the long run as good a material for making the puddle required as was clay itself, and this discovery was no doubt brought about simply by virtue of the necessity of finding some cheaper material than clay, which had generally to be brought from a distance, and dragged several hundreds of feet uphill. It is, though, quite conceivable that the idea of puddling with chalk may have been derived from simple observation of the chalk 14 THEIR AGE AND HISTORY streams which cut their way through chalk downs. If these had not puddled their own bottoms they would have long since been swallowed up in the chalk, as indeed we have reason to believe many were, although it is only fair to admit that much of their sediment is brought down from other than chalk districts, and that their beds contain a mixture of chalky and clayey ooze. To hark back to the times when dew-ponds first came to be made, we shall probably find ourselves in prehistoric times : and in the discussion of the matter of their first invention, we have only to consider the ranges of chalk hills which almost invariably stretch along within a few miles of the sea. Although the dew-pond idea was no doubt carried inland with a shifting population, and utilized so long as the inhabitants remained upon, or whenever they reached, the unretentive chalk subsoil and continued to encamp thereon, yet the invention must, I think, have come about while the migrating palaeo- 15 DEW-PONDS lithic or neolithic races still remained upon the chalk downs which stretch westward from Kent and Sussex. If we knew more of the early palaeolithic races, we might find that even as early as their times dew-ponds were made. If we accept the continental divisions into which palaeolithic flint imple- ments have been divided, reckoning among them the beautifully and finely-chipped arrow-heads, knife-blades, etc., which are even now the wonder of students of flint- chipping, then the palaeolithic age must have been one of great length, possibly some 400,000 years, it has been suggested, and the tribes which brought the chipping to such perfection could not have been few in numbers. So, in order to obtain a water-supply for the places where they encamped, it is quite con- ceivable, and almost probable, that they were really the first to utilize the puddled hole in which to collect water. Here and there where the clay formed, as it does frequently, a pocket in the chalk, palaeolithic mammals 16 THEIR AGE AND HISTORY far and wide would know where a " lick " could be obtained in times of drought, and so also would palaeolithic man ; and a race that had the natural brain-power to make a Solutrean implement would not lack the necessary intelligence to adopt the plan of artificially puddling their own ponds. The trampling of cattle round the " pans " would of itself puddle them, and make them addi- tionally retentive, and the very simple basis of a dew-pond would thus be laid down. All the arguments in favour of attributing the invention of the dew-pond to neolithic peoples seem to me to be just as applicable to the earlier races. The difficulty is to apply our imagination to those earlier races of whom we know so little. We have no reason to think that palaeolithic man was any more fond of low-lying forest- land than neolithic man, or, for the matter of that, than Roman or Saxon on their first entrance into this country. The downland of south-eastern England formed for all of B 17 DEW-PONDS them almost natural highways into the in- terior. I do not think that the barer portions of the downs were ever much less bare of vegetation than they are at the present day, and the dew-pond was then a necessity, as now. I believe the downland was the home of the earliest races, whether palaeolithic or neolithic, and it seems to me that the balance of probabilities is on the side of regarding the many earthworks of the downs as places of more or less permanent habitation, and not merely places of refuge in times of tribal storm or stress. In both ages the water-level on the habitable chalk ranges was probably much the same, and we can conceive of no causes which were then in existence to lead to a lowering of the water-level, such as since has been seen in the chalk. Forests had not yet been cleared. There was probably a greater rainfall in both ages than there is now, and whatever may have been the reasons why neolithic peoples seldom made wells would have applied with equal force to 18 THEIR AGE AND HISTORY the palaeolithic peoples who preceded them. Probably the simple reason was that it was so much easier to trample out a dew-pond than to dig laboriously a well to the required depth. It was probably much easier to make a protecting rampart around a pond, than to make a vertical hole in the earth, and who cares to say that some of our earth- works may not have been laid out in the first place by the earlier people. It should be noted that Pitt-Rivers, in his notes on the Winklebury Camp excavations, 850 feet above sea-level, speaks favourably of the idea that these highly-placed camps may have been watered by springs which then ran at a higher level than now. And, of course, if there were a probability of this, we should have here important evidence in favour of some dew-ponds having been filled at one time by springs. But this could never have been so in the case of those ponds which are really at the very summit of the downs. Gilbert White referred to the fact that the 19 DEW-PONDS water-line in chalk was always found at the same level in all the wells in his district, although recent observations go to show that the water-line follows the contour of the chalk hills. We know that since so many private wells and borings have tapped the chalk under London, the water-level has been steadily sinking. The chalk is sometimes likened to a sponge in the way in which it soaks up water, and it will not yield surplus water until it has itself been saturated. But then, if the water-level be lowered, as we know it has been lowered, the chalk would still remain saturated if we grant it this soaking power, although above the water- level it would not yield a supply which could be tapped by well-sinkers. In the olden days it would not have been any more likely to have given rise to springs on the highest ground than now, and little more than the mere surface drainage, or that part which remained after percolation, would have gone to fill the ponds. Pitt-Rivers also points out that in 20 THEIR AGE AND HISTORY many chalk districts " there are high springs which run only in the winter, when the hills have sapped up the winter rains, and retained them, like sponges, at the higher levels/' (" Excavations in Cranborne Chase/' vol. ii. p. 237.) But this can have no reference to summit-ponds, and if these springs merely flow because the water which supplies them cannot sink into saturated chalk, then the ponds which they do feed have no special reason to be called " dew-ponds " at all. I have in my mind's eye, as I write, the picture which the ancient downland dwellers must have seen as they looked northward from the top of Sussex downs. Here as they sent out scouts in the early morning, and cast their gaze out over the Weald, they would have seen, with the rising sun, the scattering of the night-born mist-wreaths which still remained, whilst the lowlands would still be be covered by a sea of fog. No church spires would then have risen above the level of this lowland fog. Only here and there a forest- 21 DEW-PONDS giant would push its head above the billowy white. A small hill here and there would have been seen, but the upward side of the sea of fog would have been seen to rock and sway like a restless ocean. Then the sun, rising in its might, would begin to dispel the cloud of whiteness. The forest would com- mence to be visible. The denizens of the woodlands would begin to awake. Wolf and bird would again be heard, and our flint- using ancestors would perhaps be thankful that necessity did not bid them often penetrate the forest below, or to make it their home. We must needs admit that before neolithic man and woman occupied the downs, other earlier races had for thousands of years made their homes there also. We need not follow the many geological speculations which have been made to estimate the number of actual years during which palaeolithic man occupied England. It is not for us here to discuss the question of the length of the Ice Age, 22 THEIR AGE AND HISTORY nor to fix on a date when the glaciers of Wales or even the ice-sheet of southern England passed away. We know that man was inter- glacial, he was probably pre-glacial, and he may even have been pre-pleistocene. All are agreed that before neolithic man came over from what is now the Continent, possibly resting his feet on the dry bed of the English Channel, other races had been here before him, for a period perhaps forty times as lengthy as that of his own period. No one, I believe, supposes that evolution of an ape- like man from a man-like ape took place in Britain, and that being so, we must conclude that the first human occupants here travelled from elsewhere, and probably from warmer and sunnier climes. If we believe in tertiary man, it may be that many of the relics of his occupation passed away when the chalk- covered dome of the Weald also passed away. There is no reason to think that his latest representatives had made no advance in civilization on what remained of the chalk 23 DEW-PONDS downs, when neolithic man came to stay, and to oust him and his race from his ancient possessions. But what he knew, and what inventions he had stumbled upon in the course of his centuries of groping darkly after better things, would be likely to be assimilated by the new-comers, and the necessity for a downland water-supply would be felt by the one just as much as by the other. The birth of the dew-pond may, indeed, have been coeval with the birth of an intelligent race, even though his intelligence was but little above that of brute creation. Let us say, then, that man found that his needs were at first met by naturally-puddled pools where four-footed beasts came to quench their thirst. Animals were brought down by his flint-headed or wooden javelins whilst they were so engaged. The beaver will make his dam, and surely when need arose, man would adapt to his own ends what he saw done by the beasts of the field ; not only would he daub himself with clay, but he 24 THEIR AGE AND HISTORY would daub his water-hole with the same or similarly tenacious material to prevent the rain-pool from soaking away. The essence of the dew-pond is the puddled waterproof bottom. The much-used en- trances to Cissbury fort probably came to be for long periods together veritable dew-ponds. We are not going to suppose that neolithic people, even with the high civilization to which they attained, made their ponds with elaborately laid straw foundations, such as has been thought by some to be essential in the construction of a dew-pond. Greater antiquarians than I have shown that all suggested means of water-supply, save ponds, for encampment life on the downs have been carefully weighed in the balance and found wanting. The weight of evidence is con- vincing that ponds only could be relied on for an all but constant supply, and as most frequently the highest spots on the downs were selected for encampments, it came to be that the best possible spots 25 DEW-PONDS were frequently selected for the founding of ponds. If we are right in assigning the dew-pond idea to our flint-implement progenitors, it is legitimate to ask why so little has been heard of them until comparatively recent times. References to the subject during the twentieth century have been quite numerous, and backward thence the subject can be traced in various writings for about a couple of centuries, when, however, oblivion drops her curtain. Johnson, Allcroft, the Hubbards, are twentieth-century writers, and their writings will merit greater attention later on. Cornish in 1902, and Miall and Reid in 1901, were attracted by the apparent mystery which surrounded the subject, and had their say thereon. But when we pass back into the nineteenth century we find that the in- terest then was more practical than scientific, and references were few. The most im- portant is, no doubt, the treatise by H. P. Slade in 1877, in which the basis of argument 26 THEIR AGE AND HISTORY goes to show that dew-ponds are really " artificial rain-ponds/' This was three or four years before my personal interest was drawn to the subject by my old schoolmaster during a well- remembered walk across the Falmer downs. The statement was made that a pond which was visited on the way was fed entirely by dew. I permitted myself to ask a question mentally as to what became of the rain which fell into it. Clutterbuck in 1865 gave important in- formation on the subject, and Vancouver speaks of them in 1813. Walter Johnson has discovered a work by Christopher Packe, bearing date 1743,* in which a pond is described, near Brabourne, in Kent, which was said to be 100 " perpendi- cular feet " above the springs below ; it was never found to be empty, and it occurred to the author to find out the cause. He thought that there was a " condensation of * " Folk- Memory," by Walter Johnson, F.G.S., p. 307. 27 DEW-PONDS elevated vapours/' and although this ex- planation was, I think, strictly correct, it is as good as saying that he knew very little about it. But it showed the awakening of a mind trying to unravel a secret of nature. Packe did not call it a " dew-pond.' ' Nor did Gilbert White, who wrote in 1776, calling the attention of his correspondent to the little ponds on the chalk hills which were not known to fail, even when the ponds in the vales were dried up. I shall say more about his remarks later. But the phenome- non appears then to have long been recognized. The word " dew-pond " does not appear to have been used prior to 1813, and since then, although some ponds have borne that name, others have been known as " mist- ponds " or " fog-ponds." What of the time between the era of the neolithic encampments and the eighteenth century ? I think that it is too much to imagine that any of the existing dew-ponds are actually those which were in existence 28 THEIR AGE AND HISTORY in the earlier times. Of course, one can never speak with certainty. Others may express a contrary opinion. But it is for such to bring forward some sort of evidence which will carry weight. Certainly no one can say that any particular pond has never failed. It may last for a hundred years with no apparent change, but unless attended to and watched over ,even the best-laid ponds " gang" at length "agley." Scattered along the downs are numerous depressions which appear once to have been ponds, but neglect has caused them to leak and run dry. The life of downland ponds is so long that they seem almost to be perennial. Hence, when once a pond has been successfully made, and it passes in turn into the hands of successive farmers, records of its making are lost, and, in view of its apparent perennial nature, it comes to be neglected. Then, when finally it ceases to be waterproof, generally by grasses and rushes growing their roots through the puddle, it gradually dries 29 DEW-PONDS up, and the secret of its formation not being remembered, it soon reaches the stage of becoming covered with downland grass, and an empty hollow alone remains. So little is done to most ponds nowadays in the direction of repairs that it is scarcely to be wondered at that so many empty pond sites are seen. During the progress of excavation in a chalk quarry near Cholsey, Berks, what may well have been the site of an old dew-pond was cut into. In the section shown in the illustration, a deposit filled the bed of the dew-pond, in which were found numerous specimens of three species of helix, and of cyclostoma elegans. No freshwater shells were found in it, but after the drying-up of the pond the land molluscs mentioned may have found a home in the well-sheltered hollow. Mr. R. R. Hutchinson, to whom I am indebted for the photograph, and for calling my atten- tion to the matter, informs me that he found absolutely no trace of vegetable remains, such as straw, at the base of the section. 30 THEIR AGE AND HISTORY With the comparatively settled times which were already in being prior to the Roman invasion, no doubt the population of the downs began to drift towards the Weald and to make clearings therein. The windy discomfort of the high downs would readily be abandoned for the lower ground as soon as a fair amount of security from wild beasts could be obtained there. Water would be there abundant, and the necessity of repeating on the lower grounds the ponds on which so much had previously depended would not be felt. With the flocks on the downs be- coming more and more scanty under the protection of the empire, many ponds no doubt were neglected, and dried up, never again to be sought after, until the time, as Johnson has suggested, when the downs once more came to be used as sheep-walks, during the wool-growing times of the mediaeval sheep-masters. I think that there is fair probability that some of the existing ponds do actually date from the revival of wool- Si DEW-PONDS growing, following the immigration of the Flemings. There is the possibility that some of them may have been continuously in use ever since. That a very large number have been lost to view completely is probable in view of the extension of the cultivation of the downland which went on during the many continental wars in which Great Britain has been engaged, and during the existence of the corn-laws. For the same reason many earthworks have been undoubtedly levelled by the farmer's plough. But this is a very different thing from maintaining that some ponds are exactly the same as those which watered palaeolithic or neolithic flocks and peoples. If under Roman protection a large population began to dwell in clearings in the woodlands, to just the same degree would be brought about the neglect of upland ponds. When a pond after a gradually dwindling process extending over many years finally became permanently unable to hold water, and bearing in mind the shallowness 32 THEIR AGE AND HISTORY of such ponds, the growth of vegetation, and the working of worms, would in the course of the centuries between, say, the first and the thirteenth, be in a fair way towards becoming filled in, and perhaps with the exception of the banks, in a few cases, would be almost unrecognizable as pond-depressions. And, one may be permitted to ask, is it to be seriously believed that tracks made by cattle in neolithic times can still be detected as leading to and from such ponds? In this respect I fail to see anything but what has taken place save within the last two or three centuries. I should like it to be proved other- wise, but my wish shall not at any rate be the father to the thought. Earthworks have been thrown up on the downs, and kitchens for troops excavated in quite modern times. Ponds may, too, have been dug, and the passage of troops and all that accompanied them during the civil wars of England will certainly have made many of the so-called cattle-tracks which are sometimes attributed to prehistoric times. c 33 CHAPTER II THEORIES OF DEW-POND ACTION THE chief interest in the subject of " dew- ponds " lies in the mystery of their mode of replenishment. In such ponds we have a fairly constant water-supply in spite of their being situated, as a rule, high above all visible means of replenishment, in places where no springs can possibly break out, where evapora- tion must proceed to a greater extent than in the valleys, and where they frequently give water to an immense number of cattle, which are thus saved the necessity of being led or driven down from their high feeding-grounds to the brooks or ponds in the valleys below. A dew-pond is contained in a circular de- pression, as a rule on a chalk subsoil, and may measure from 30 feet to 70 feet across the 34 THEORIES OF ACTION widest part. Every pond on a high situation is not necessarily a " dew-pond/' and the ponds which are being built nowadays on the chalk hills, are placed in such a situation, I am informed by a practical pond-maker, that there should be some sort of natural runnel to convey surface drainage into the pond. The object of course is to collect water, and advantage of anything which is conducive to this object will naturally be taken. But such a pond is not. a dew-pond, and we have now more particularly to consider those ponds which have no connexion at all with surface drainage. They are frequently at the very summits of hills, and, although some are placed as stated by Clement Reid, on the northern slopes of the downs, yet some of those I have seen are a couple of miles south of the escarpment of the South Downs, and so are not, in these cases, indebted to the condensation of vapour that takes place on the edge of the escarpment, which occurs 35 DEW-PONDS when a very wet wind is blowing in from the sea. Tyndall's observations on the cloud which always appears on the lee side of the Matterhorn are interesting in considering ponds on the northern slopes, the cloud appearing to be stationary, but being in reality fresh and continued condensations of aqueous vapour. The chief point to be remembered in con- nexion with the position of these ponds is that they stand on high ground, and that many of them retain a quantity of water, when the brooks and the ponds in the lower country around are running dry. So, as Johnson and Wright remark, the strange spectacle is often seen of villagers going up the hill to fetch water for the houses and granges in the vale.* I have made some effort to ascertain the origin of the well-known nursery rhyme concerning the adventures of Jack and Jill, but, although no useful information is forth- * " Neolithic Man in N,E, Surrey," p. 47. 36 < si THEORIES OF ACTION coming, I have little doubt that it originated in a time when the necessity was more frequent than now, of sending " up the hill to fetch a pail of water/' although it is hoped, not always with the same disastrous consequences. Gilbert White's mention of the little ponds on the downs around Selborne is an early reference to this class of pond, but he does not actually call them " dew-ponds/' so that the name may have come into use subsequently to his time. He says : " Now we have many such little round ponds in this district ; and one in particular on our sheep-down, 300 feet above my house, which, though never above 3 feet deep in the middle, and not more than 30 feet in diameter, and containing, perhaps, not more than two or three hundred hogs- heads of water, yet never is known to fail, though it affords drink for 300 or 400 sheep, and for at least twenty head of large cattle beside. This pond, it is true, is overhung with two moderate beeches, that, doubtless, 37 DEW-PONDS at times afford it much supply ; but then we have others as small that, without the aid of trees, and in spite of evaporation from sun and wind, and perpetual consumption by cattle, yet constantly maintain a moderate share of water, without overflowing in the wettest seasons, as they would do if supplied by springs. By my journal of May 1775, it appears that ' the small and even considerable ponds in the vales are now dried up, while the small ponds on the very tops of hills are but little affected/ Can this difference be accounted for from evaporation alone, which certainly is more prevalent in bottoms ? or, rather, have not those elevated pools some unnoticed recruits, which in the night-time counterbalance the waste of the day ; with- out which the cattle alone must soon exhaust them ? " The question which we have to consider is the nature of these high-ground recruiting agencies. Briefly they are (i) Mist, (2) Dew, (3) Rain, and it is desirable to ascertain the 38 THEORIES OF ACTION relative values of these agencies. To take dew first. What is dew ? This may seem a simple question to answer, but like all apparently simple things, it becomes more complicated the more one considers it. Many meteorologists still maintain the old theory, which is certainly the popular theory, that dew is formed by precipitation of the aqueous vapour already existing in the lower layers of the atmosphere, when the radiation of heat from the earth has caused its surface to be in the condition to chill below the dew- point the layer of saturated air in contact with it. Precipitated moisture may appear in the form of dew, hoar-frost, mist, fog, or cloud, but in dew and hoar-frost there is precipitation without a cloudy intermediary. Freest radiation of heat from the earth's surface takes place when there are no clouds to reflect to earth the heat which it gives off at night. If there are no clouds, the chilling of the ground and of the layer of air 39 DEW-PONDS in contact with it will be considerable, and the temperature may be reduced to the dew- point. Dr. W. C. Wells, a London physician, was, we are told, the first properly to define what dew is. He published his well-known " Theory of Dew " in 1814. He exposed certain flocks of wool on the ground, and found that moisture was deposited upon it, if freely exposed to an unclouded sky. If, however, he placed the wool under a table or other covering, then no moisture was found upon it. The first was not, perhaps, an original observation, for it was at least as old as some of the older portions of Old Testament history.* It was evident that the wool had to be exposed to the freest radiation, in order that dew might be deposited on its upper surface. Wells maintained that dew was formed by the cooling, below a certain point, of air which was in immediate contact with the * Judges vi, 37, 38, 40 THEORIES OF ACTION soil and was approaching a condition of saturation. During the day the earth is receiving heat from the sun, and a wave of heat gradually penetrates by conduction into the soil, and this may be to a depth of as much as 4 feet Below this point, observations go to show that diurnal variation of temperature ceases to be noticeable. As the sun declines in the west, terrestrial radiation takes the place of solar radiation, and, in accordance with physical law, the earth begins to radiate the heat in all directions. This radiation in- creases after sunset, and causes a rapid reduc- tion in the surface temperature, until dew- point is reached. The air, we must remember, is scarcely warmed at all by radiation, whether from the sun or the earth, although with its enormous vertical height, it absorbs some 50 per cent, of the sun's heat which passes through it. The earth's radiation after nightfall does little towards warming the air, whilst, on the other hand, it does DEW-PONDS much to cause it to become cooled by contact with the chilled earth. The heat rises through the pores and interstices of the soil, and the soil, reaching a low temperature itself, cools, by contact, the narrow stratum of air resting upon it. If this stratum be fairly moist, it will, by reduction of its temperature, very quickly attain the point of saturation. It will reach saturation-point in obedience to the well-known law that a high temperature will allow it to hold a larger quantity of aqueous vapour than will a low temperature (as a matter of fact, it is, of course, from the operation of this simple law that precipitation in any shape or form takes place). Then on the saturation-point being reached, any further reduction of temperature will cause dew to be deposited, first on the most power- ful radiators, such as wool, hair, straw, and vegetation in general, and then on substances of less radiating power, such as the under- sides of stones. The deposition may, however, be interfered with by buildings or lofty trees, 42 THEORIES OF ACTION by a cloudy sky, by a very damp atmosphere, as this checks radiation, or by a high wind. A wind prevents the air remaining in contact with the soil long enough to give any material reduction in its temperature. It must, too, be remembered that the very condensation of vapour which results in dew, will at once set free a certain amount of latent heat, and this fact alone may tend to arrest the cooling of the air, so that, after a certain amount of dew has been deposited, the temperature may again rise above dew- point. This theory of dew presupposes that the immediate source of the deposition is from vapour which already exists in the atmosphere, but this theory was contested by Edward Blythe in 1836. In his edition of Gilbert White's " Selborne," he criticized a remark which White makes, that the moisture which occasions the clouds in the morning, " being no longer rendered buoyant by the warmth of the sun, falls down again in dews." 43 DEW-PONDS Blythe says that " the true theory of dew is that it rises from the ground and does not fall, as is the common opinion. Much moisture is at all times continually ascending from the earth, which passes off during the day in invisible vapour, but is condensed by the chills of the night and appears as dew." This may be called the " upward theory," as compared with Wells' theory, which may be called the " downward theory." That dew which rises is at least as important in its results as that which falls, was fully demonstrated by Dr. J. Aitken some years ago. Moisture rises out of the ground with the radiant heat at night-time and, provided there are no clouds to interfere with radiation, the moisture will be chilled below dew-point. The clearer and more cloudless the night, and the smaller the movement of the air, then the greater and freer the radiation, and therefore the greater the deposition of dew. This theory has also been demonstrated by experiment, since a tray, being inverted over 44 THEORIES OF ACTION the soil on certain nights, will be found to be covered with dew only on the underside, and at the same time the undersides of stones are also moist. Experiments illustrative of both opinions have been carried out by me repeatedly, and both are correct under certain conditions. But there is yet another form in which so-called dew may be met with, and that is by reason of the transpiration of plants, and what is sometimes called the " sparkling dewdrop " may be nothing but an exhalation from the leaves of a plant. We know that at night-time the stomata of leaves close and transpiration is very much diminished, but the absorption by the roots which has gone on during the day, and by which food from the soil is drawn up, con- tinues at night-time although to a much diminished extent. During the day, when the stomata are open, a sunflower 3^ feet high, exhibiting a surface of 5600 square inches, was found in summer to exhale no less than 45 DEW-PONDS a pint of water a day, and at night-time so much water frequently accumulates in a plant as to force drops out of the leaves as well as out of other parts. This is particularly noticeable with some saxifrages and grasses, and when in the early morning leaves are found covered with dewdrops, it is not easy to see which form of dew is responsible for the phenomenon. We have, therefore, three sources of dew : 1. From the condensation of moisture already existing in the air at nightfall. 2. From the condensation of that which rises with radiant heat from the earth. 3. From the exhalation of plants. From an economic point of view we may, if we please, consider either of these forms of dew to be of use to us, if they happen to be present, and to aid in replenishing our ponds. In regard to the last-mentioned form of dew, we may practically put it out of court 9 as the greater the quantity of vegetation around a pond, the more it will be likely to THEORIES OF ACTION interfere with free radiation from the soil, and, at all events, the total amount of dew exhaled by the plants can go but a very short way in the direction of filling a pond. Being themselves, however, good radiators, the deposition of true dew upon them may be considerable, and this may be useful if it drains into the pond. In regard to Dr. Aitken's theory, it will be seen at once that dew deposited in the way shown can only arise from moisture which has arisen with the radiation of the pond's own heat, and consequently its deposition can only result from earlier evaporation at the expense of itself, except that, if the sloping margin of such a pond be broad and waterproof any dew which may form thereon, under this theory, may possibly trickle into the pond and so add its quota thereto. We are therefore left practically to dew formed in accordance with Dr. Wells' theory, for the replenishment of our ponds, if we are to grant that dew has anything to do 47 DEW-PONDS with it at all. There can be no doubt, I would add, that the term " dew " has been very loosely used in the past, practically anything in the way of condensed moisture, other than rain, being classed as dew. We can form some idea of the nature of vegetation as a good radiator of heat by looking at the deposition of hoar-frost upon various objects when dew-point is below zero. Then hoar-frost and not dew is formed, and the amount of frost is a valuable guide to show us the extent of dew-deposit. If we notice isolated logs of wood on a frosty morning, or the palings surrounding a field, we shall often see that these will be covered with frost, although the stones of a road will be dry. Now notice where the palings are the thickest in consequence of the back-stays which go to support them. If the frost has not been sufficient to cause the whole wooden front to be covered with glistening crystals, you will see that the frost is thickest just where the crossings of the 48 THEORIES OF ACTION wood take place, or if there is no frost where the wood is in single planks, it will yet be seen that where the planks are double some frost will be found ; and in this way, where otherwise the back-stays of a fence would be invisible, their pattern is thus made visible. Now the greater the radiation of heat the greater the frost, and this observation goes to show that the greater the thickness of wood the greater the radiation of heat. Now examine another portion of a fence, the cross- sections of the rough upright supports, or examine the cross-sections of the trunks of trees, where they have been cut down ; you will not see, as a rule, a mass of hoar-frost on the top, but, strangely as it seems, a series of concentric rings, and these exactly imitate the rings of annual growth of the tree. Now the rings of growth are made up of two parts, the early or spring growth, and the later or autumn growth. The former is open and thin and permeated with elongated vessels, the latter is thick and dense and has practically D 49 DEW-PONDS no visible vessels at all. You will at once notice that the frost-rings follow the thick and dense wood, that is to say, the autumn layer ; so in this case the denser the wood the greater the radiation, and the chilling, and the deposition. I should like at this stage to introduce a further remark of Gilbert White's in regard to the deposition of frost on the surface of bogs, where they conceal fallen trees beneath the surface. He says : " Old people have assured me that, on a winter's morning, they have discovered these trees in the bogs by the hoar-frost, which lay longer over the space where they are concealed than in the surrounding morass. Nor does this seem to be a fanciful notion, but consistent with true philosophy. Dr. Hales saith, ' That the warmth of the earth, at some depth under the ground, has an influence in promoting a thaw, as well as change of the weather from a freezing to a thawing state, is manifest, from this observa- 50 THEORIES OF ACTION tion, viz., November 29, 1731, a little snow having fallen in the night, it was, by eleven the next morning, mostly melted away on the surface of the earth, except in several places in Bushy Park, where there were drains dug and covered with earth, on which the snow continued to lie, whether those drains were full of water or dry ; as also where elm-pipes lay under ground ; a plain proof this, that those drains intercepted the warmth of the earth from ascending from greater depths below them ; for the snow lay where the drain had more than 4 feet depth of earth over it. It continued also to lie on thatch, tiles, and the tops of walls/ (See Hales' "Haema- statics," p. 360.) QUERY, Might not such observations be reduced to domestic use, by promoting the discovery of old obliterated drains and wells about houses ; and in Roman stations and camps lead to the finding of pavements, baths, and graves, and other hidden relics of curious antiquity ? " I suggest as a result of these observations, DEW-PONDS that, if we want a non-conducting material and one which, at the same time, is a good radiator, we cannot do better than pave the bottom and sides of our dew-ponds with closely-fitting logs of wood. Now one more observation : Straw, it would seem to follow, in common with other forms of vegetable life, should be a good dew- producer. In India, ice is obtained in shallow pans laid at night-time on rice straw, while the temperature of the air is above 32, and, from the manner in which the ice is produced, it is clear that the effect of the straw is to cut off from below the heat which would other- wise rise at nightfall from the earth. Dr. W. C. Wells,* in his " Essay on Dew," quotes the descriptions given by Sir Robert Barker and Mr. Williams in the Philosophical Trans- actions (vols. Ixv. and Ixxxiii). In the former case excavations were made to a depth of 2 feet in the open plain, and their bottoms * Wells' "Essay on Dew," ist edition 1814, p, 137. 52 THEORIES OF ACTION were covered with dried sugar-cane or stems of Indian corn, to a thickness of 8 inches or a foot. In the process described by Mr. Williams, little mounds of earth, 4 inches high, were made around square flats with 4 feet or 6 feet sides. In these were laid dried straw, or sugar-cane haulm. Ice was formed when the air was still, but wind prevented its formation completely. Dr. Wells repeated the experiments in 1812, in the garden of a friend " in Surrey, distant about three miles from the bridge over the Thames at Black- friars," using in each case clean, dry straw, and proved that the formation of ice was due to " the radiation of heat to the heavens." One experiment in particular is of interest at the present moment. Amongst the pans in which he placed water he placed one that was dry. He found the " inside of its bottom to be as much colder than the air as the water was in the other pans before ice appeared in them." The dry pan, " in the course of the night, attracted moisture, which was after- 53 DEW-PONDS wards converted into a film of ice " (pp. cit. p. 117). In this pan we have the essence of the theoretical dew-pond. It should be noted that the straw had always to be kept dry. The problem that would remain, if this plan were adopted in England, would be to ascertain how the water which might result from deposition of dew in such pans, could be retained and prevented from being evaporated during the day. We must bear in mind that, from time to time, various observations have been made with a view to measure the amount of dew- fall, and Mr. G. Dines has given, as a result of his observations, an average of 1.397 inches, or on the grass alone 1.022 inches, and " making a liberal allowance for contingencies, it may be fairly assumed that the average annual deposit of dew on the surface of the earth falls short of 1.5 inches." If our ponds are recruited to any great extent by dew, there must be a greater deposition on the surface of the water already in the pond, 54 THEORIES OF ACTION or on its margin, than that which is found on the grass. It has been contended that in making a dew-pond, a layer of straw is placed below the puddled floor, and that this prevents the rise of heat from the earth after night- fall. If it acted in such a way it may be granted that the mass of water would remain fairly cool if laid down in winter. But to be an effective non-conductor of heat it must be kept dry, and this is an absolute impossibility beneath the moist puddle of a pond. I have tried it, and I can speak from experience. The ice which is obtained in Bengal at night-time in shallow trays, can only result from a change of straw each night. The straw must be dry. Then the low temperature is obtained, and this is below freezing-point. But if you leave it for a few hours in the sunshine which follows, of course the atmosphere will quickly lick up the moisture which it had disgorged. So with the dew-pond. Supposing that the straw 55 DEW-PONDS were efficient in the way suggested, and the water went below dew-point. Suppose that the straw were in some way arranged that it kept dry. If it were a bad conductor of the earth's heat at night-time, it would be a bad conductor of solar heat in the daytime. So, instead of the sun's heat having free access to the earth beneath the pond, it would be in part arrested, and would raise the temperature of the water higher than it would otherwise reach. The commencement of the succeeding night would therefore see the pond with a greatly-increased temperature, and we may safely say that such a pond would scarcely ever at any time during the night approach dew-point. In my experi- mental pond I could not keep the straw dry. I dug it up to see. When I laid it down the second time, I laid planks of wood between the straw and the puddle. By this means I hoped to keep it drier, and possibly it was because of this that there were one or two occasions on which dew-point was 56 THEORIES OF ACTION reached in the water. But even with these precautions against the straw becoming wet, there must always be the moistening resulting from hydrostatic pressure from below, as pointed out by Walter Johnson. This has been well shown, too, by Professor Oliver, in endeavouring to account for the growth of the flora of shingle-beaches. It has been suggested that in a similar way water from below might rise to a pond, but as the bottom of a pond is made as watertight as possible, and the margin is also made as hard and watertight as possible, if any such action took place, it is certain that in the driest weather much more water would pass down- ward than ever came upward. We now turn to another possible, and I may perhaps say a probable, source of origin for the water in the ponds. In a private letter from the maker of some ponds on the " Duke of Norfolk Downs " and on Amberley Mount, it is stated that the highest parts are chosen, as they are 57 DEW-PONDS " more exposed to the weather " than lower down, the inference being that they are filled by the moisture-laden winds blowing in from the south-west, no consideration being given whatever to any artificial attempt to attract dew-precipitation. But as R. H. Scott says, dew can never appear when there is much wind, for the air cannot remain long enough in contact with the soil for any material reduction of its temperature, and consequent condensation of moisture to take place.* The " weather " referred to can only, there- fore, be mist or fog. It is stated that some old Surrey people in the neighbourhood of Worm's Heath, do not apply the term " dew-ponds " to their remarkably constant sources of water, but call them " mist-ponds," and herein I think we have the source of at any rate a good deal of the water in these ponds. I have watched, and have also felt, the mists which roll over the downs in Sussex, * Int. Sci. Series, vol. xlvi 58 THEORIES OF ACTION permeating every object that they touch, and causing one's garments to become wet and clammy. If this happens to the clothing on a warm body, it does not seem to follow that the surface of the pond must of a necessity be at a low temperature in order to condense water from the mists. Clement Reid states that the " open downs, even in the middle of summer, receive much heavier dews than might be expected, or than are met with on the lowlands. Thick mists often cling to their tops for several hours after sunrise, while the plains below are already dry and sunny. It is only by ob- serving the large amount of moisture inter- cepted and dripping from the overhanging boughs as the sea of mist drifts slowly past, that one can realize how prolific a source this must be." It is impossible, I think, to refuse to admit that the positions of successful ponds favour the contention that fogs and mists do add largely to their supply of water. The 59 DEW-PONDS observations of the shepherd whom Mr. Cornish employed, if reliable, are rather striking in this respect. After a night of fog, the surface of his pond was found on January 18 to have risen ij inches ; the next day, following another fog, gave 2 inches ; and on January 24 an inch was measured.* It was not recorded what was the principle on which the bottom of the pond was laid. I am informed that shepherds on the South Downs at sunrise look out for banks of mist on the ground, for where these banks settle they point out the best sites for pond-construction. The most ancient of the ponds in the neighbourhood of Chanctonbury Ring is bare of all vegetation, except a few rushes in the water, but it has a good wide margin. A half- mile away is the Ring itself, and, although the downs themselves may be free from fog, the trees will often be enshrouded by a " mist- cap/' when it is practically hidden from view. There is no doubt that a great supply of * C. J. Cornish, "Naturalist on the Thames." 60 CHANCTONBURY RING AND POND RIDDLESDOWN POND, SURREY No raised bank around it THEORIES OF ACTION water could be obtained from the " Ring/' if proper collecting arrangements were made. Although a great portion of the clump of trees is enclosed, a portion of the vallum and the enclosed space is open to the public, and the trees are of considerable height. The whole area reeks with moisture, and both the atmosphere and the sodden soil give evidence of the water which might be rendered available for economic purposes. This brings us to a consideration of the question as to how far vegetation will assist in the upkeep of a pond. Reid thinks that an overhanging tree on the south-western side, such as a stunted ivy-covered thorn, or an oak, or a bush of holly, may be of great value in drawing vapour from the atmosphere, especially whenever a sea-mist drifts in, and remarks that, " in early morning or towards evening, there is a continuous drip from the smooth leaves of an overhanging tree/' He also appears to think that the condensation on the surface of the water itself will be 61 DEW-PONDS considerable, although roads adjoining may be dry and dusty. Of the pond on the sheep- down above White's house at Selborne, it is stated that " this pond is overhung with two moderate beeches that, doubtless, at times afford it much supply.*' If we admit the efficacy of overhanging trees, there should be no necessity for limiting their number to one or two. It would be best in such circumstances to do all that might be possible, supposing the trees would grow in these situations, to arrest the mist, and to decide once and for all to sacrifice the radiation necessary for the deposition of dew. The condition of many of the Sussex roads, where the hedges are high on both sides, shows clearly how these check radi- ation and evaporation, and the various drift-ways, or cattle-ways, at the foot of the downs, with tall trees or bushes on both sides, show well also how retentive of moisture the trees cause them to be. Iso- lated trees standing on the bare downs, 62 THEORIES OF ACTION do not, as a rule, seem to be in a very flourishing condition. But the trees at Chanctonbury Ring have a good soil above the chalk. This is used now for puddling new ponds, and it is probable that the ancient vallum on which many of the trees now grow is made up from this clay collected from pockets in the chalk-surface around, thus giving the trees a soil similar to that of some parts of the Surrey Downs, where, in contrast to the South Downs, large treeless tracts are few and far between. I think we may say that if vegetation can be induced to grow around a pond it will be beneficial to it. Attention has been called to the efficacy of vegetation in arresting moisture on the high ground of Table Mountain, by Professor Henslow, and by Mr. W. Parkinson, in the Ligurian Alps.* Whenever a south-west wind blows across Table Mountain, a blanket of white mist is * Knowledge, September and October, 1907. 63 DEW-POND S frequently found to cover the mountain, even in the dry season, and Dr. Marloth endeavoured to ascertain the amount of precipitation which was derived from this mist, in order to account for the luxuriance of the vegetation at the top. In November 1901, he says, " I took two 5-inch rain- gauges to the top of Table Mountain, and placed them about mid-way between the east and west ends of the upper plateau. One I left open in the usual way, the other one I surmounted with a framework repre- senting a bunch of reeds. The arrangement consisted of two rings of 5 inches diameter, which were connected by four rods of stout wire, the whole frame being i foot high. Pieces of wire-netting were fixed inside the rings and reeds were drawn through the meshes and fastened with thin wire. The frame was then inserted into the other rain- gauge, fitting into its opening by means of a narrow socket. Four wires attached at op- posite sides, and fastened to stones near by, THEORIES OF ACTION protected the frame against the fury of the wind. I had consequently one ordinary gauge and one with an imitation bunch of reeds, one foot high." The following Table gives the results of the readings from December 21, 1902, to February 15, 1903 : Period Open Gauge Gauge with Reeds Differ- ences Dec. 21, 1902-Jan. i, 1903 Jan. i-n, 1903 Jan. 11-18, 1903 Jan. i8-Feb. i, 1903 Feb. 1-15, 1903 Total . 1.04 3-51 0.42 15.22 14.64 16.66 (full) 16.66 16.66 ,; 15.22 14.64 15.62 13.15 16.24 4-97 79.84 74.87 " This result is, that from December 21, 1902, to February 15, 1903, that is, in fifty- six days, the gauge with the reeds had con- densed a quantity of moisture equivalent to 79.84 inches of rain, and that quantity was recorded, although the last three times the gauge had overflowed. It is not too much E 65 DEW-PONDS to assume, that as the season of the south- east clouds extends over double that time, there would consequently be a condensation of moisture, exclusive of rain, of at least 150 inches during the dry season alone." We find in fact that the reed-gauge collected sixteen times the quantity of water contained in an open gauge, between these dates. This is a remarkable result. Professor Henslow kindly furnished me with a copy of Dr. Marloth's paper, and I notice that besides attributing to this abundance of moisture the luxuriance of the vegetation of the mountain, he speaks of it as the reason " why there are little lakes, even late in summer, on the top of Table Mountain, as well as close to the summit of Du Toit's Peak." This is additional evidence of the important part played by low vegetation around a pond in arresting and condensing aqueous vapour. It is desirable that the mists brought on to the South and North Downs by the south-west winds may be 66 THEORIES OF ACTION encouraged to deposit their vapour by sur- rounding ponds with shrubs and bushes, or even short reeds embedded in their margins. Mr, Parkinson in noticing a somewhat similar phenomenon in Italy wrote as follows : " This year has been one of exceptional drought in southern Piedmont. Now a special feature of these Ligurian Alps is the extraordinary fertility of the soil at altitudes which, elsewhere in Italy, would be nearly barren. On Mindino, over 6000 feet above sea-level, the cattle pasture, and when we visited Galero, over 5000 feet, the peasants were cutting hay on slopes so steep as to be hard to climb. Corn and potatoes flourish on these crests, and the magnificent forests of chestnuts that cloak the mountain-sides remain green long into the autumn. " But what excited our curiosity and called for explanation was the fact that though all the springs were dry at the foot of the hills, at an astonishingly short distance from the DEW-PONDS summits there were abundant springs of delicious water, so fresh as to numb the hands. " The question that called for solution was, whence comes the inexhaustible supply which enables springs so near the summits to hold out whilst all those in the valleys have com- pletely failed ? " Looking at the mass of the mountain which yields the water, that is to say the volume of rock above the spring head, it seems impossible, even supposing the summit a hollow reservoir, that it can contain a supply equal to the output for a period of drought as great as that of the present year. I should mention that the snowfall here last winter, on which the peasants count for their supply of water in summer, was far below the average/' The conclusion was that the thick covering of verdure on the hills gathers water from the moisture-laden wind, and the low misty clouds which covered the mountain-tops. 68 THEORIES OF ACTION Fogs an dmists condense on the particles of dust which are always found in the at- mosphere. In mists such as are found on the chalk downs, the particles of moisture- covered dust are comparatively large, and fall by gravity. In response to some questions which I raised, Dr. John Aitken wrote to me as follows : " In all conditions with which we are ac- quainted the water-vapour in the atmosphere only condenses on dust-particles as nuclei. It does not matter whether the condensation takes the form of fog, mist, or cloud. In- deed, all these forms of condensation only vary in the amount of vapour condensed on each nucleus. We call it fog when the amount of vapour is small on each nucleus, and when condensation is mainly due to the affinity of the dust-particles for water. This form of condensation generally takes place in air not quite saturated. In clouds the amount of water condensed on each particle DEW-PONDS is greater, and it is grown by the air being or tending to being saturated. When this supersaturation is continued for some time the particles grow to a larger size, and become what we call mist. In clouds up to say 6000 or 7000 feet the highest I have observed in the cloud-particles may be either small or large : in the latter condition they are just like what we call mist at low level. With the aid of an instrument I have seen and counted the particles, which are always falling, and the number was counted on a micrometer. The only difference noticed in the two cases was that in one the particles were small and fell slowly, while in the other they were larger and fell quicker (as mist), but there is never any difference between the two other than in the size of the drops, and that might be anything from almost invisible to fine raindrops. "We do not know of condensation ever taking place in the atmosphere on anything but dust. Some have advocated ions as 70 THEORIES OF ACTION nuclei, in order to explain atmospheric electricity. But ions cannot act save in a dustless atmosphere, and that has not been proved to exist in our atmosphere." He adds, "A sheet of water is of all things the last one would select for collecting dew. Its specific heat is so great that it stands losing a great quantity before it is cooled to the dew-point, and it is not likely to do that during a night, if there is any depth of water." It will be seen that Dr. Aitken will not allow that ions can act as condensation nuclei save in a dustless atmosphere, and this has not been proved to exist in ours. I think that if the dust-particles on the downs were counted, they would be found to be comparatively few. Dust in the air shows itself on a hot day by adhering to perspiration, necessitating frequent washing, or by accumulating in the nostrils. But during my sojourn on the downs, one never seemed to become dirty. DEW-PONDS Nothing was more remarkable than the clean hands and face that were maintained throughout the whole of the day. There was little, if any, dirt to remove. The number of dust-particles must have been very few, although no actual count has, so far as I know, been taken of them. A mist-condens- ing around such a paucity of particles would be coarse and heavy, each particle getting a comparatively large share of the moisture, and this tendency would be increased by the natural affinity of saline particles for moisture. In fact, condensation will take place on salt- dust even before the air is saturated. When such combined salt-and-water-dust passes over an impervious soil-space, much will be deposited in the form of water on coming within the cooling influence of the earth after nightfall, and will be retained where there is a pond-area made to receive it. The small spherules of atmospheric vapour which have formed around particles of dust as nuclei, have been measured by Assmann, 72 THEORIES OF ACTION who found them, on the Brocken, to range from .006 mm. to .017 mm., and by Kaemtz, who found in a fog at Jena diameters of .014 mm. to .035 mm. " When the spherules grew to have a diameter of .04 mm., the fog or cloud became wet and rain began/' The rate of fall of these particles into a pond-area would, of course, be extremely small, owing to the resistance of the viscosity of the air. It would, of course, be considerably less than that of the fall of a raindrop. Wiesner con- cluded by experiment that the largest raindrops could not exceed a diameter of 7.2 mm., and these are constantly splitting up as they fall. Flash-light photographs show that their shapes are constantly changing, and Dr. G. C. Simpson has shown by experiments undertaken by the Meteorological Office of India, that the main cause of atmospheric electricity is to be found in the separation which accompanies the break- ing up of raindrops.* Allusion has already been made to the * Proc. of Royal Socy. t February 4, 1909. 73 DEW-PONDS probability of mist-particles being deposited in a pond by gravity or entanglement as the moving mists pass over it. But there is another way by which such particles may reach a pond-area in which already there is a certain amount of water. The cooling of the earth around a pond is by radiation, but a mobile mass of water becomes cooled by a different process. On the cooling of the upper layer of water this sinks and gives place to warmer water supplied from below. Such a process continues until the whole reaches a uniform temperature of about 40 F., unless the night is too short to allow of this substantial reduction of temperature. As a matter of fact this reduction scarcely ever happens during the months when there is dew-deposition upon the grass around. Consequently the water surface is, as we have seen, almost invariably warmer than the air over it, and the convection-currents of warmer water from below prevent the whole from reaching the low temperature which 74 THEORIES OF ACTION may be reached on the grass around. The surface of the water is almost invariably warmer at night than that of earth around, and the air over the land will have a tendency to become lower in temperature than that resting on the water. The colder and heavier air will then flow over the sides of the pond into the area of the pond and tend to take the place of the warmer air. At the same time the mixture of the cold with the warm air will, if they contain vapour in due pro- portions, result in some condensation over the pond, which will probably go to replenish it. In extreme circumstances, fog will result, but it seems possible that in some cases condensation may be so rapid when the temperatures have become very greatly different, that large mist-particles may be rapidly formed near to the surface, and deposited immediately in the pond. Water- surfaces have been proved to be, as a rule, warmer than the air which rests upon them, especially on those dewy nights which follow 75 DEW-PONDS days of great heat. I cannot help thinking that in this we may see the cause of those rises in the levels of ponds which have been recorded after dewy nights. Not that true dew will have been deposited on the water, but the circumstances have been such as may result in the simultaneous production of dew on the grass, and another form of condensation on the water. Economically speaking, it matters but little whence comes the water of the hill-tops, so long as ponds, constructed with a fair amount of care, will collect it. If we encourage more mist-deposition by planting small shrubs around, we might interfere with the deposition of dew, but on the other hand, as dew is a minor factor, as I hold, then one or two overhanging trees will be beneficial, these being on the south- west side of the pond, in the track of the moisture-laden winds. The claim that dew alone is the great cause of the permanence of such ponds receives a THEORIES OF ACTION shock from an experiment conducted by J. G. Cornish at Lockinge, in Berkshire, and recorded in C. J. Cornish's " Naturalist on the Thames/' The temperature of the water in a dew-pond on Lockinge Downs on July 16, 1901, was 20 F. higher than the temperature of the air. Dew, could not, therefore, have been deposited, since the differences of tem- perature would probably have been main- tained throughout the night, or rather the differences would probably have been ac- centuated. This would be in accordance with the principle that water, although it takes longer to warm, yet when once it acquires a certain temperature it retains its heat without materially warming the air above it. Water has far less absorbing and radiating power than dry land, and, there- fore, would have less effect on the air above it. R. H. Scott states that " as the specific heat of water is five times that of dry land, it takes five times as much heat to raise a given mass of water through a given range of 77 DEW-PONDS temperature as it does to raise an equal mass of dry land/' If mist be admitted as a valuable agent in recruiting the ponds, then it is a fit subject for inquiry as to what steps should be taken to encourage the deposition of the mist as water. White admitted that an overhanging beech or other tree was of importance in connexion with some of the ponds around Selborne. Clement Reid thinks that an over- hanging tree on the side nearest the source of the moisture-laden currents of air is of importance. The position of the pond now becomes of importance, and if the pond has a high southern or south-western bank it seems to act in a favourable way in causing fog to precipitate its moisture. The Sussex Downs are the home of the dew-pond, and many a time for the whole of a day I have walked through dense fogs which have rolled in from the sea, and have finally taken their flight, as from a jumping- THEORIES OF ACTION off ground, along the northern ridge of the downs between the Dyke and Plumpton. The trees, where there are any, such as the Holt, near Clayton, will then be seen and heard dropping water on to the leaf-soil below, whilst one's own garments become damp and clammy. The question as to whether the upland " dew-ponds " are aided to any extent by the atmosphere's electricity in obtaining con- densation of water- vapour opens up a number of interesting questions, and perhaps deserves discussion. Condensation of vapour gives rise to electric separation. So also does the splitting up of raindrops. A mist or cloud in the process of growth, from mere droplets to larger drops, preparatory to falling as rain, would accumulate an enormously high electric potential. Professor Tait says, " if vapour-particles were originally electrified to any finite potential, cloud-particles would be each at a potential enormously higher, and raindrops considerably higher still." 79 DEW-PONDS Within a highly electrified cloud, as Dr. G. C. Simpson said, there must be rapid combination of the water-drops, and from it considerable rain will fall. This rain will be positively or negatively charged according as previous existing conditions favour the one or the other. Messrs. McClelland and Nolan recently reported that during a period from October n to May 12, observation of the electric charge on rain showed that the positively-charged rain was 76.9 per cent, of the whole examined.* Fuchs has stated that when water-drops are electrified they condense and become larger, and there is a distinct connexion between the size of the drops and the electrical condition of the atmosphere. The growing of a mist on the high downland must give rise to considerable electrical separation, and it should be re- marked that it is recognized by even un- scientific persons living near the downs that the mists of the downs are of quite a different * Royal Irish Academy, June 24, 1912. 80 THEORIES OF ACTION nature to the fogs which arise in the Weald below the downs. As the vapour is blown up the valleys from the sea and becomes chilled after nightfall, condensation takes place. Mist is formed, and tension is greatest on the outside of each droplet. As the droplets coalesce, the tension becomes enormous. The potential of the rounded hills of the downs is probably always some- what high, there being no angular protuber- ances or rocks to carry off their charge into space. But owing to the higher potential of the atmosphere even at a few feet only above the surface, a drifting mist would be seized upon to convey to earth the charge necessary to bring about equilibrium, and as mist is almost always positively electrified, and the tension is increasing in the upper layers of the mist by union of the droplets, so the draught of the fog-particles in the lower layers towards the earth would also be increased. Downland mists drench cloth- ing and trees and grass and ponds with F 81 DEW-PONDS moisture, although their temperature may be well above dew-point. The high positive potential would have to be relieved sooner or later, and as a cloud obtains relief in the lightning flash, with a fall of rain or hail, there must be a corresponding relief to the potential of a mist. We do not see it, because its discharge takes place silently, but I think we are justified in believing that precipitation takes place from a mist, aided by its electrical condition just as well as from a cloud. We know it is so from personal experience as we walk through a downland mist, and I can see no reason why we should imagine it to be otherwise. Still we do not always realize that mist and cloud are the same, and that both must give rise to precipitation. Maxima of atmospheric electricity occur at the periods of most rapid change of temperature,* and these are the periods when the densest down- land mists are found. * R. H. Scott, "Meteorology," Int. Sci. Series, p, 171. CHAPTER III VARYING MODES OF DEW-POND CONSTRUCTION As regards the construction of a dew-pond, there are great differences of opinion. The simplest form nowadays is that in which the bottom is merely covered with puddled clay, which, however, is carefully tempered with lime, either with the object of preventing the working of worms, or to prevent the clay slipping, or in certain circumstances, cracking. There are three dew-ponds near Chanctonbury . Two of them have been made in recent years by the owner of the land and the occupier of the farm jointly. The clay which is found hereabouts in pockets in the chalk was used for puddling, and this was lime-tempered. About a barrel and a half of water was poured in, and then the ponds filled. They have 83 DEW-PONDS not since failed, not even in the long, hot summer of 1893, when the ponds in the Weald were dried up. The largest of these ponds, which may be of ancient origin, lies due south from the Ring, between it and Cissbury, and although I ascertained that it had been cleaned, I could gain no information as to its having been made within the memory of man. As regards the recently-made ponds, it will be noticed that nature is assisted by a small amount of water being poured in as a commencement, and, whatever the scientific action of the pond is, it is successful in accomplishing the desired result. J. C. Clutterbuck, in 1865, said that in making such ponds an excavation was made in the chalk on the tops of the hills, from 30 to 40 feet or more in diameter, and from 4 to 6 feet deep. The bottom was " covered with clay carefully tempered, mixed with a considerable quantity of lime. . . ." This was " protected from the action of the sun and atmosphere by a covering of straw." 84 CONSTRUCTION After this " efficient and impermeable coating or puddle " is completed, " a layer of broken chalk is placed upon it." The water must, he added, first be introduced by artificial means. In 1877 Mr. H. P. Slade discarded the term " dew-ponds " in favour of " artificial rain- ponds/ 1 and scouted the idea that dew had any part in filling ponds at all. His remarks dealt practically with one pond, the greatest diameter of which was 6gJ feet, which was constructed in 1836 at a cost of 40. It was bedded in the Thorpe Downs, near Lough- borough, on the Berkshire Hills, at a height of 450 feet above the level of the sea. Being " fed from the heavens," this fact probably gave rise to its being classed as a dew-pond. The basis of this pond was stated to be, first a layer of clay about 12 inches thick (mixed with lime to prevent the working of earth- worms), second, a coating of straw, " to prevent the sun cracking the clay," and, thirdly, a layer of loose rubble. During an 85 DEW-PONDS interval of forty years, till 1876, the pond had only once been dry. The exception was in 1854, and this resulted principally from the growth of rushes, whose roots stuck through the clay bottom, causing leakage in what was otherwise " a waterproof bed." The straw was not held to have any particular effect in causing dew-precipitation, and the rubble, which would, of course, by the way, allow of the straw becoming saturated, was merely to prevent the hoofs of cattle trampling upon and perforating the clay, or puddle. Mr. Slade showed the results of a large number of meteorological observations. It was noticeable that, during a twenty-two hours' observation, on July 21 and 22, the pond's temperature, 3 inches under the surface, varied between 67 F. and 87 F., the former being found at 5 A.M., and the latter at 3 P.M. But 3 feet under the water the temperature only varied between 68 F. and 70 F., the former being between 4 A.M. and 5 A.M., and the latter being between 10 P.M. 86 CONSTRUCTION and midnight. At 5 A.M. only the surface temperature was lower than that 3 feet below the surface. Thus the surface had a variation between 20 F., but 3 feet below it had one of ij F. only, during the two days quoted. Somewhat similar results were found during each summer month, when one temperature per day only was taken. Thus, in June, the surface varied 17 F., and 3 feet below it 10 F.; in July, the former varied 25 F., the latter 9 F. ; in August, 22 F. and 6 F. ; and in September, 19 F. and 8 F. From certain observations, taken at a depth of i foot below the surface, there appeared to be a time, viz., between 4.45 A.M. and 6.45 A.M., when the temperature at this depth was slightly greater than that at the surface. This is when the surface of both land and water has been chilled to the greatest extent by radiation, and I suggest that it is thus in the early morning that the greatest con- densation of fog occurs, as it is undoubtedly the time when the greatest amount of fog, 87 DEW-PONDS mist, or cloud, whichever one likes to call it, is seen rolling over the downs. Messrs. Hubbard have described the forma- tion of a dew-pond, although we are not informed where the practice quoted is in vogue. As a matter of fact, in their more recent experiments they departed from the simplicity of the theoretical arrangements which they had quoted. The space hollowed out for the purpose, they say, is first quickly covered with a coating of dry straw. The straw is in turn covered by well-chosen, finely-puddled clay, and the upper surface of the clay is then closely strewn with stones. The margin of the straw has to be effectually protected by the clay, since if it becomes wet it will cease to attract the dew, as it ceases to act as a non-conductor of heat and " becomes of the same temperature as the surrounding earth/ ' This would, of course, follow quickly if a runnel or spring were allowed to drain into the pond. The puddled clay is chilled CONSTRUCTION by the process of evaporation, and the dry straw prevents the heat of the earth after a hot day from warming the clay. It is stated that there is at least one wander- ing gang of men, who will construct for the modern farmer a dew-pond of this kind which will contain more water in the heat of summer than during the winter rains. It is very certain, however, that many alleged dew-ponds are not formed on this plan. I interviewed a well-known pond-maker at Alfriston in 1906 on the subject, but only found that the ponds which he made, whether on high or low ground, consisted of an excavated hollow, with a carefully concreted bottom. With thermodynamics he had nothing to do, nor did he show any inclination to advance the cause of science by building a scientific dew-pond. For 30 or 40 he would build one anywhere, but he would choose a site where runnels made their appearance in rainy weather. DEW-PONDS Acting on the principles quoted, Messrs. Hubbard constructed a pond which was founded on a concreted base, and on this was placed a coating of pitch, dry sand, mica blocks, and a coating of asphalt. The result was not a complete success, but the operation was cut short by untoward wilful damage to the foundations. Such a pond would be too complicated in its construction to be of practical use, and its cost would be as a rule prohibitive. On June 23, 1905, they lodged an application for a patent for making a " reservoir for dew/' and followed it up by a complete specification on January 22, 1906. The principles shown above were incor- porated in the application, but the details of the various linings of the pond were a concrete foundation, when not on chalk soil, a layer of asphalt, a layer of asbestos, a second layer of asphalt, joined to the first around the edges, and a layer of porous bricks. 90 CONSTRUCTION No patent was granted on this applica- tion. A question of importance now arises, and that is as to whether the old dew-pond builders did actually utilize straw, for the double purpose of encouraging radiation from the puddled chalk, and of preventing the uprise of heat from the earth below. So far as I know, there is no evidence that such was the case. There are numerous ponds around Lewes, which are now dried up, which show that they were based on concrete or on puddled clay, but there is no evidence in this neighbourhood of straw having been used as a foundation. In Dorsetshire, I am informed by Mr. A. J. Hogg, the formation of dew-ponds is as follows : " They are made with sloping sides, which, as well as the bottom of the pit or hollow, are covered with about 9 inches of clay, dry enough to allow of being rammed or beaten into a solid mass, upon this another 9 inches of clay is firmly rammed ; this 91 DEW-PONDS producing a solid lining impervious to water of some 1 8 inches in thickness, over the whole of the sides and bottom of the pond. "A final coating of stones or other hard material is added, to preserve the clay from injury by too deep penetration of the feet of the cattle using the pond, which is then com- plete, and ready to receive the water. Straw is not used in the formation of ponds in the Dorchester district. I am told that the ponds do not require filling with water in the first instance, the accumulation of rain and dews being sufficient to produce a supply. This, however, probably applies to ponds made in the spring and autumn. In the summer the surface of the clay-lining would be liable to become cracked under the heat of the sun. " One made about fifty years ago, on Maiden Castle, near Dorchester, which was seen by my informant immediately after completion, was filled with water brought in barrels from a brook about half a mile away. 92 CONSTRUCTION This pond was lined throughout with cobble- stones, of which there was an unlimited supply within a convenient distance, on the sea-beach near Weymouth ; and they doubt- less afforded effective protection to the surface of the clay. These dew-ponds are formed on the tops of hills, or on other elevated positions, in order to avoid the silting, by which, at a lower level, they would otherwise be filled up in process of time." Mr. Hogg also tells me that in his boyhood days he remembers his father telling him that the dew-ponds in Essex where the subsoil, the Boulder Clay, is of quite a different nature to that of Sussex are made by a layer of straw being first laid down, and this being covered by a layer of puddled clay. It was not, however, contended that the straw fulfilled any object in the way of causing precipitation of dew. An important piece of evidence in the direc- tion of showing that straw may be used with a view to causing precipitation reached me 93 DEW-PONDS from Pewsey, in Wiltshire. Here a farmer, Mr. Keevil, states that the Marquis of Ayles- bury makes a point of including in his leases a clause in regard to the construction of ponds, providing for their formation in the following manner : Three layers of straw, each 3 inches in thickness when pressed down, are placed alternately between three layers of puddled clay, the whole series giving a thickness of 2 feet 6 inches, no other substance being required. The best wheat-straw is used, not cut in any way. The lowest layer is straw and the topmost clay. The ponds are then railed round in order to prevent cattle straying into them. Mr. J. Keevil states that each pond costs about 300, but the ponds referred to are of large size, being about 60 feet across. The pond so made lasts roughly about one hundred years. He states also that he had used dew-ponds on Salisbury Plain that kept full in very dry summers. The dew-pond in Wiltshire to which he specially referred is that on St. Martin's 94 CONSTRUCTION Hill, either at Lovett's Farm or Clinch Farm. The straw was always to be provided by the farmer. For ponds in general, straw was cut into short lengths and mixed with the clay, but this operation was merely to prevent it cracking, and so letting the water through. The statement was made that the water in the ponds used to rise up in a single night, but no one knew how it came about. Some of these ponds have recently been attributed on very insufficient evidence to the work of neolithic man, but it is extremely difficult to say, as a matter of fact, which ponds are of ancient, and which are of modern con- struction. A recent writer in The Field, in discussing " the use and disuse of chalk/' refers to the former use of that substance as a foundation for puddling dew-ponds. He says that a Sussex farmer who was born just before the middle of the last century, had been telling him how the dew-ponds were made in his younger days, and in the days of his fore- 95 DEW-PONDS fathers. " The requisite hole having been excavated, the chalk was laid down layer by layer, while a team of oxen harnessed to a heavy broad-wheeled cart was drawn round and round the cup-shaped hole to grind the chalk to powder. Water was then thrown over the latter as the work progressed, and after nearly a day of this process, the resultant mass of puddled chalk, which had been reduced to the consistency of thick cream, was smoothed out with the back of a shovel from the centre, the surface being left at last as smooth and even as a sheet of glass. A few days later, in the absence of frost or very heavy rain, the chalk had become as hard as any cement, and would stand for years without letting the water through. This old method of making dew-ponds seems to have died out when the oxen disappeared from the Sussex hills, but it is evident that the older ponds, many of which have stood for scores of years practically without repair, are still more watertight than most modern CONSTRUCTION ones in which Portland cement has been employed " (The Field, December 14, 1907). Further evidence of the making of modern ponds in the Marlborough district reaches me from another source. The description given is as follows : " The dew-pond is constructed on the downs, on the high ground, and is filled by the moisture being attracted by the water already in the pond ; at first, it must depend upon rainfall to start it, the water falling upon the sloping sides of the pond being a considerable quantity. The pond is dug and shaped out. Then clay and fine lime are mixed into a tough condition, and well-beaten over the chalk in layers, say 12 inches in thickness. Straw (good, tough wheat-straw) is laid over the clay. When very firm, chalky rubble is put over the straw. Straw is laid over all until the rubble becomes hard. The straw between the clay and the rubble serves to prevent the rubble from getting mixed into the clay." Here again it will be seen that there is no suggestion G 97 DEW-PONDS that the straw is a factor tending towards the condensation of the atmospheric mois- ture. As regards the Yorkshire Wolds, the late Mr. J. R. Mortimer, F.R.S., wrote as follows : u Perhaps no district of the same area con- tains more ponds than the Midwolds of Yorkshire. These are partly ancient, partly modern. The latter can be numbered by hundreds, nearly all of which have been made during the last 150 years mostly after the inclosures of the parishes. Previous to that, the stock grazing on the open commons were driven to be watered at the ponds in the villages. These ponds are mostly of a circular form, and of all sizes, from ten yards to fifty yards in diameter. A few are oval, to adapt themselves to the ground on which they are constructed. Their chosen sites are generally in depressions on the surface of the land, in which the rain-water has a tendency to collect, or on sloping ground, and often near the side of a high road or track- 98 CONSTRUCTION way, where the running water from the roads during rain can be conveyed by a channel or gutter into the pond. This is the means by which the ponds are supplied on the Yorkshire Wolds. No one ever thinks of filling them by any other means, the con- densation of fog or mist being a very small factor. In this district they are constructed as follows : First, a dish-shaped excavation with a gentle slope to the centre, is made in the ground, to the depth of 4 to 6 feet, ac- cording to the diameter of the pond. This is then covered with quick-lime, next a layer of clay, which is wetted and beaten with wooden mallets into an impervious sheet, 3 to 4 inches thick. Again a covering of quick-lime is applied, then a coating of stiff wheat-straw, and on the top of this is spread broken chalk. The two coverings of lime are to prevent the earth-worms boring through the bed of clay. The bed of straw is to prevent the covering of broken chalk from being trodden, by cattle going to drink, 99 DEW-PONDS into the impervious bed of clay, which, if not protected by the straw and broken chalk, would be pierced through, and the pond would lose its water. I believe it is generally considered that, as soon as the pond is con- structed, the sooner it is filled with water the better, as, if without water for any length of time, the clay lining is liable to shrink and crack from the effect of dry weather." After criticizing Messrs. Hubbard's remarks as to the necessity of keeping the straw layer from getting wet, he remarked : " But such does not happen to the ponds on the Yorkshire Wolds, where the straw is liable to become wet quickly from always being placed on the top of the clay and only covered with the porous bed of broken chalk, while the flow of rain-water is always encouraged to enter the pond. Is it possible that Messrs. Hubbard have been wrongly informed as to placing the straw under the clay ? If so, their dew- pond hypothesis must come to an end. It would be interesting to interview one or more 100 CONST ; RUCT;JftN;, of the professional pond-makers in the neigh- bourhood of Messrs. Hubbard's researches, and so ascertain whether it be the custom to place the straw under the clay, and not upon it, as in Yorkshire. For if such an error has been made their means of condensing the fog does not exist. My belief is that these dew-ponds are no other than rain-water receptacles, the condensation of fog and mist being of the smallest quantity. " A few ancient ponds are to be found on the wolds where no settlement now exists near them. A typical example of one of this class is on the Birdsall Wolds, between Aldro and Vessey Pasture. Undoubtedly most of these ponds date from the beginning of the villages which surround them, and some probably before. These ancient ponds mostly owe their existence to the presence of local patches of clay, occupying depressions on the surface of the chalk rock. This is proved by the fact that the ponds in many of the villages are on such local beds of clay. Their 101 varying size is mainly due to the following causes : After rain, many of these patches of clay would gather and hold pools of water on the surface, while no pools would exist on the naked chalk ground round them. Hence these patches of clay would, at times, supply primitive man with one of the chief necessaries of life. They would also furnish him with the material to plaster the wattled sides of his rude dwellings, and also supply him with material to assist in erecting his burial mounds. He would consequently be induced to fix his rude dwelling near these patches of clay, and presently he would observe that the place from which he took this clay continued to hold more water as the place increased in size, and it would finally become the village pond. " That such was the original use at Fimber was proved when, during the dry summer of 1884, the long oval mere was cleaned out, and its bottom at the west end showed pit- like hollows, where it seemed that the clay 102 CONSTRUCTION had been taken from irregular depths, while the adjoining circular pond has a similar deep depression in its bottom, locally known as the deep hole, half-way between the centre of the pond and the west side adjoining the church. " That clay has been taken from the sites of these ponds we have proofs. At Fimber this clay had entered largely into the con- struction of a British barrow, on a slight eminence near to the southern margin of the western mere, on which barrow three suc- cessive churches have stood. Besides, it is known that the walled sides of the oldest houses in the village were originally plastered with this clay mixed with short straw. Later, these mud walls were gradually re- placed with walls of chalk-stone, as now seen in some of the old houses. Yet, even up to the last three or four years, some of the partition walls in the old house in which I was born consisted of lattice-work plastered with clay, mixed with short straw, obtained 103 DEW-PONDS from the site of the Fimber mere, there being no other similar clay nearer than the village of Fridaythorpe, two miles away." The sectional diagrams on pp. 106, 107 will show the foundations of ponds according to the various authorities to whom I have referred. It will thus be seen that, in a majority of the cases quoted, straw is used in one position or another, whilst its utility as a dew-producer is, as a rule, denied. In the majority of instances shown, chalk rubble is used as the topmost layer, to prevent the perforation of the puddled clay by the hoofs of animals. In Sussex I have examined a number of downland ponds, and do not find this to have been the case. It is true that small chalk is now strewn over the bottom of a pond when repairs are under- taken, but this is only preliminary to the grinding of it to powder and to the puddling which follows, by driving a team of oxen or horses harnessed to a broad-wheeled wagon, repeatedly round and round, and through 104 CONSTRUCTION the pond, for at least a whole day. This process grinds the chalk to a powder, and lays it down as a waterproof puddle, and when there is no clay near at hand it is, of course, a far more economical process than carting clay to the site. When there is clay at hand, clay undoubtedly is used, as being more reliable as an impervious bottom. An old Sussex resident who went to live at the top of Clayton Downs in 1875, informed me that when a boy, he knew a pond which sometimes held a little water in it. When it had some water he obtained permission to drive his horses into it to clean their hoofs on his way home. After having done this repeatedly the pond began to hold more water. In fact, he had involuntarily puddled it, and the pond ceased to leak. In these cases there was, at any rate, no fear that the cattle would perforate the bottom. He also says that when a pond on Dale Hill was getting low his master used to set him to drive the oxen round and round the pond. 105 /- "\ Tb&h - of DEW-PONDS He got very tired of his job, not seeing any reason for it, but it was no doubt with the object of well puddling the bottom, and so making it waterproof. When this process has been completed, and the cattle or the broad-wheeled cart have done their work, men would take broad spades and flatten down the surface or smooth out the puddle until it was smooth as glass, but the centre was not generally touched, and if there were weeds and grasses growing there these were left, unless it was thought that their roots had gone too far down. The puddle on the banks of the pond, which increased in area as the water diminished in area during drought, set as hard as cement . The absence, then, of rubble, and the use of oxen, etc., for puddling, make me doubt whether the Southdown ponds have straw, at any rate immediately under the puddle. I am referring, of course, to those ponds in which one can see that chalk- puddle and not clay-puddle has been used. 108 CONSTRUCTION Mr. Turner, of Ditchling, who is familiar with the ponds on the hills of his neighbour- hood, has never heard of straw having been used in any of them. On the other hand, no one hereabouts seems to know of any ponds having actually been made within recent years. There seems to be a general agreement that lime should be mixed with clay-puddle to prevent the working of worms. It is a question whether earth-worms could be a danger under a pond, but there is the little red- worm, which must be guarded against. I am not familiar with many clay-puddled ponds, but there are two on the down over Pyecombe, which I have examined, in which not a vestige of chalk or white puddle can be seen. One (No. 9) is excavated on a steep slope, so that one bank is very much higher than the other. On the side nearest the lower bank I was struck, on approaching it, by the red colour of the ooze where it was just covered by the water. Immediately I 109 DEW-PONDS touched it, in order to take up a handful of the muck, the colour suddenly disappeared, only to reappear in a few seconds. On re- peating it, I saw that the colour was caused by thousands of little red worms, and these, on the slightest movement reaching them, disappeared as if by magic. I think that the working of these red-worms would be a thing to guard against, in the case of ponds bottomed with clay-puddle. I have kept some of these worms in captivity, so to speak, and find that, when the mud is sufficiently tenacious, they not only honeycomb it with their borings but form tubes of mud, which may project as much as a quarter of an inch from its surface. When in the water, they travel far and wide by rapidly-made twists and jerks, looping themselves like the letter S, until they finally come to rest. The pond I refer to received very much drainage, and could not be regarded as a dew-pond. The second clay-puddle pond (No. 10) was about a mile farther north, and contained clearer water, no CONSTRUCTION although it was also on the slope of the hill, but there was a gorse-bush growing down to the water's edge, whilst on the opposite side there was, close to it, the hedge which formed the boundary of the field, with a few fair- sized trees in it, and these, no doubt, gave a good deal of water to the pond by means of a high-formed dew and the moisture from the mists. There was also a considerable quantity of. drainage from one side of the pond. These two ponds, although not dew- ponds, bear upon the subject, and may serve to throw light upon their replenishment. in CHAPTER IV EXPERIMENT AND OBSERVATION FOR some considerable time I had under my special observation a tract of downland in Sussex, bounded on the north by the escarp- ment of the South Downs from the Devil's Dyke to Plumpton, and in this area are in- cluded some excellent ponds. A list will be found in the Appendix. My headquarters were one of the disused mills on the Clayton Downs, close to a shed bearing the Ordnance Survey bench-mark of 581 feet. A long spell of dry weather happened while I was there, and, being on the spot from dewy eve till early morn, I was able to make some useful and interesting observations . There are many downland ponds in the area to which I have referred ; but two ponds stand out con- 112 EXPERIMENT spicuously amongst the rest as excellent examples of dew-ponds. The first is one which stands high up above the waterworks in the Waterhall Valley, near Patcham, and is about 700 feet above sea-level. When I visited it, on September 13, 1908, the water- edge had a circumference of 260 feet . Between the edge of the water and the top of the ridge surrounding it, there was a shelving bank, measuring 15 feet across, the ridge itself being 347! feet round. The centre had a quantity of green growth below the surface, but this did not approach within 3 feet of the edge of the pond, nor did it grow above the surface of the water. All around this centre of green was a band of clean white chalky puddle. Around the edge of the water there was, on the northern side, a peculiar little ridge, like cliffs in miniature, marking the boundary of the pond. On the shelving bank thistles and hawkweeds and other plants were growing, and the pond would not appear to have ex- ceeded for a long time its present limits. The H 113 DEW-PONDS pond is but slightly below the summit of the down, and had no drainage except such as it received from the banks. The second dew- pond I specially refer to is near High Park Corner, 700 feet above sea-level, on the road between Ditchling Bostel and Brighton. On August 12 it measured 227^ feet around its edge, and between the edge and the top of the ridge there was a gradually shelving bank, 20 feet across. There was thus consider- able collecting-ground, but with this exception there was no drainage whatever. A few yards away a large pit has been excavated to a depth considerably below the level of the water, and if the bottom of the pond were not completely watertight, the water would rapidly drain away. There is but little weed in the pond. The water is clear, and the puddle is a chalky white, hardening completely where exposed to the sun. I could gain no information as to when any of the ponds about here were made, and therefore no information as to the manner 114 SHAMBLEDEAN BOTTOM PONU A drainage pond ; wide rain-collecting margin WATERHALL POND Showing encroachment of weeds on margin of pond EXPERIMENT of their formation. There is some clay in large pockets in the chalk near at hand ; but so far as could be seen, no clay had been used. A little distance below the highest ground on the west is another pond (No. 20) which, I was informed, had been repaired about fifty years ago, and the clay had been carted there for the foundations. Many of the modern ponds are clay-puddled, being more easily rendered waterproof by this means. All ponds receive a quantity of rain ; but to be a dew-pond it should in some way or other attract dew from the atmosphere, deposited either upon its shelving banks or upon the water itself, or it must receive and arrest thick banks of mist, fog, or cloud, and assimilate the moisture. The shelving bank must in any case be a considerable factor in the matter, as I have poured water on the banks of well-puddled ponds and have found that the water had run down to the pond with no apparent percolation into DEW-PONDS the bank. For dew to be deposited on the surface of water it is generally considered that the temperature of the water must be below dew-point. We can leave aside the question of dew in the winter and spring months. The ponds to be useful must be receiving dew in the hot summer months, when drought may be prevalent, and when ponds on lower grounds are drying up. Ponds exceeding 3 feet in depth are rare, so that we have these ponds in shallow basins of perhaps 30 to 70 feet across, and the broad surface is being heated all day long by the sun. Water takes much longer than the soil or the atmosphere to become heated, but when heated it retains its heat correspondingly longer, and this is the greater the greater the depth of the pond. Whereas the tem- perature on the grass went down from 52 to 30 F. on the night of September 21, that on the surface of the water in my experimental pond sank merely from 50 to 42.5. Dew was forming on the grass at 9 P.M., the dew- 116 EXPERIMENT point being 45.5, as determined by the wet- and dry-bulb thermometers. Of course it does not follow that when dew-point was reached on the grass, the pond-water com- menced to receive dew. The deposition of dew would be at first upon the most powerful radiators of heat, and amongst these grass stands high. Water is a very bad radiator, and would be the last to receive dew, if the acquisition of dew on its surface were merely the result of radiation of its own heat and the chilling of the air above it by contact with the cooled water surface. The water on this occasion cooled down to below dew-point, and soon after 10 P.M. should, according to current theory, have commenced to receive dew. The water (rain- water received during the few days since the completion of the pond) had already, at 8.30 P.M., reached a comparatively low tem- perature. This may be attributed to the fact that there were but 3 or 4 inches of water and to the special construction of its bed. 117 DEW-PONDS It has been stated by Professor Miall that a dew-pond must not be greater than about 4 feet in depth, this being alleged to be the depth which will allow the water to cool as a whole during the course of a summer's night. But a downland pond 4 feet deep is rare, and if a pond over 2 feet deep in the centre is heated by a summer's sun for a few days, it is certain that evaporation will be such that the pond will gradually diminish until it is about that depth. During the nights that follow there will be radiation from the water as from the grass bank around, but the grass will be the first to receive dew, and the period of dew-deposit will be much longer there than in such a pond, since even sup- posing a large pond does, as a whole, actually reach a temperature below the dew-point, this will only take place late at night or in early morning. As will be seen later, it is only on rare occasions that the water of a dew-pond passes below dew-point. Much depends on whether or no a pond starts 118 EXPERIMENT working in a cool state. If made during a long spell of hot weather it cannot be ex- pected to become cool in the short summer nights. One can now understand why Clutterbuck and other authorities insisted on the pond being first filled, after being made, by snow being piled around the edges of the pond and being allowed to drain into it. There was a long fogless period in August and part of September, 1908, and this may have accounted for the drying up of the pond called One Tree pond. On July 25 there were but 9 inches of water, although the circumference of the pond was 102 feet. Owing to the shallowness it became rapidly heated each sunny day. The temperatures at i inch, 6 inches, and 9 inches were all 71 F. at 1.30 P.M., that on the bank exposed to a south-west wind being 67. On July 24, at 8.20 P.M., the temperature at the same three depths was 67.5, the cooling to this extent in the shallow basin having apparently 119 DEW-PONDS been uniform. The temperature on the bank was at the same time 58.5, this again showing the less rapid cooling of the water than on the dry bank. A shallow pond seems to have no chance whatever in hot, settled weather, since the evaporation of the day would be continued well into the night, and possibly throughout it. On July 28, at 8.30 P.M., the thermometer at i inch showed 64, and at 6 inches 63.5. These tempera- tures were soon reversed, and at 9.10 P.M., whereas I inch showed 62, 6 inches showed 63. Then, at 9.50 P.M., both showed 61. On each occasion the temperature of the air immediately above the water, as shown on floating corks, the bulbs of the chemical thermometers used overhanging the water, was less than the i-inch water temperature. At 8.30 P.M. it was 57.5, and at 9.50 P.M. it was 55. The dew-point some distance away was 50.8, but the wind was too high for dew-deposition. It will thus be observed that the atmosphere was on each occasion 1 20 EXPERIMENT colder than the water, and hence there would be no possibility of the water chilling the air sufficient to compel the latter to disgorge its aqueous vapour. If vapour were actually given up by the air there must have been some other cause for it. The rain began at 10 P.M. Had it proved a clear, starry night, the atmosphere over the pond might have become still further reduced, but the pond would have lost its heat still more slowly, and the difference between the two would have become accentuated. The effect of radiation of heat from a pond upon the air above it must be regarded, I think, in a different light to the effects of radiation of heat from the earth's surface, e.g. the downland. In the latter case, the heat rises with vapour out of the interstices of the soil, and is dissipated into space. The soil being cooled, brings about the cooling of the stratum of air in contact with it. When the cooling reaches the dew-point, the best radiators receive a deposit of dew. But as 121 DEW-PONDS regards radiation from the surface of the pond, we have seen that the surface tempera- ture is such that it can seldom have any effect in cooling the air above it, since the air is already at a lower temperature. The cooling of the air must be due to other causes, and I doubt if equilibrium between the two is ever obtained. Take one of the cases mentioned. At 9.50 P.M. the water was, everywhere in the pond, at 61. The air immediately above it was 56. The water was no doubt losing heat, but the air was losing it more rapidly, and if the temperature of the air fell below dew-point, the point of saturation being reached, the stratum of air must give out some of its moisture, and this would be rapidly assimilated into the pond. If evaporation of the pond were proceeding at the same time, this would tend to cause the point of saturation of the air to be reached the earlier, and this might take place even while the pond-water itself had not yet fallen to dew-point. On the other hand, if it can only receive dew by 122 EXPERIMENT chilling, as it is often said to do, the air above it, this can scarcely ever happen, as pond- water can rarely reach a temperature at night-time lower than the air above it. It could only happen after some hours of freest radiation, and by this time probably the early sun will have begun to warm the air, the effect being to increase the amount of vapour which the atmosphere could hold without being saturated, and hence removing all possibility of precipitation. Even at 8 A.M. on August 9 the air at the surface of the water was but 58, whilst i inch under the surface it was 60. This was in light fog blowing from south. It is not, perhaps, very important as to whether the dew "falls/' as Dr. Wells demonstrated in 1814, or whether it rises, as has since been shown by Edward Blythe, and later by Dr. Aitken. Both theories have been maintained in turn by various meteorologists. It is, however, very easily demonstrated that both are correct. Two 123 DEW-PONDS illustrations will suffice. On the night of August 13, two opaline saucers were placed on the grass. On one was placed a portion of cotton- wool. The other saucer was inverted, and a portion of cotton- wool placed beneath it. The former portion was covered when examined the following morning with medium- sized beads of dew, the latter was dry. The dew, therefore, fell in this case. Again, a thermometer with a broad opal scale was supported between two double earthenware pans. Where protected from the rising heat of the earth, i.e. over the pans, there were large beads of dew both on and under the opal scale, but the centre portion of the opal was dry top and bottom. There was a clear line of demarcation. Here again the dew apparently " fell." But on October 30, 1907, I placed two square polished tins to catch and condense the vapour radiated from the earth. One was inverted. In both cases no dew was 124 EXPERIMENT found on the uppermost sides of the tins in the morning, whilst in both cases the undersides were well covered with dew. Here, therefore, the dew rose from the earth after nightfall. Thus, on different nights it is not difficult to prove that both theories may be correct. It must be confessed, however, that out of the numerous occasions on which I placed articles out at night to detect the dew, by far the majority receive dew on the uppermost side, and so the older-fashioned theory of the two was well sustained. And it is well for dew-ponds that it was so. If a pond were recruited from dew only which rose as vapour from itself, it would gain nothing, whereas it would lose by evaporation which proceeded in the day- time, and the balance would be on the wrong side. It has been estimated that the average annual dew-fall does not exceed 1.5 inches. This, of course, is founded on the amount 125 DEW-PONDS of dew which is deposited upon surfaces other than water. A heavy deposit occurred on the night of September 15. The night was clear, and the moon was bright. I had excavated a trough about 4 feet deep, and had placed a quantity of wood-wool over an area 6 feet square at the bottom, and this was covered by planks of wood. In order to keep it dry, and anticipating a dewy night, I spread over the whole a piece of waterproof sheeting 6 feet square. In the morning I poured the dew from it into the rain-measure, and found that it showed 1.31 inches. Spread over the whole area of 36 square feet, this would be a deposit of about Y^ inch (.00496 inch). If this amount of dew fell each night throughout the year, we should have a total annual dew-fall of 1.8104 inches ; but I do not think it would be right to regard this one night's dew as a fair average. Probably not more than three times the amount would be deposited in any one week, and, estimated on these lines, the 126 EXPERIMENT annual dew-fall would be .77376 inch, or somewhat over three-quarters of an inch. I have elsewhere made the suggestion that if we seek a non-conducting material for the bottom of the pond, which shall at the same time be comparatively inexpensive, we shall find it in logs or planks of wood. Vegeta- tion of all kinds is, of course, an excellent radiator and non-conductor, and when we see the bushes and grass covered with a deposit of hoar-frost, we can also notice how thickly the frost stands upon the wood of a fence. Evidence of its non-conducting property is also seen in the appearance presented by the ballast over the sleepers of a railway line, where frost will remain after it has disappeared from the material which is not underlain by wooden sleepers. In order to test the extent to which the temperature was affected above and below wood resting on the downland, I placed two thermometers on a plank which rested on the grass of the downland, and one under- 127 DEW-PONDS neath the plank (August 16). Next morning the one beneath showed that a minimum of 51 had been reached. One that was placed completely over the wood went down to 48.5. The third was placed so that the bulb and half of the tube were resting over the grass. This went down to 42.5. The night was a clear one. The first one, placed beneath the board, showed the effect of the absence of radiation from the grass, and that the board had to some extent prevented the escape of the rising heat of the earth. The second showed the effect of free radiation from the wood, but this was not so powerful a radiation as the grass, since the bulb of the third, resting directly over the grass, showed a reduction in temperature to 42.5. It should be men- tioned that the bulbs of the thermometers were in all cases open both back and front of the opal scales. Similar results were obtained on the night of September n, when the thermometers 128 EXPERIMENT were respectively 42.5, 37, and 38, with heavy dew everywhere, and on September 12, when similarly placed, they showed 41.5, 38, and 38 respectively, and on September 8, where they showed 49, 41, and 42.5. On the night of August 16, the minimum tempera- ture of 48.5 was reached also in the following instances : in an opaline saucer, lying over a small red earthenware pan, and in a similar saucer over a tray containing chalk-puddle, both upon grass. The chalk-puddle had no effect in reducing the temperature over it. On another occasion (August 21) I placed three thermometers as follows : one beneath the wood, which showed a minimum of 51; another resting completely on the wood and a third resting with its bulb and a portion of the tube over the grass, with the remaining portion on the wood. In the morning the latter two showed a temperature of 55, a higher temperature than the first mentioned ; thus the relative temperatures were reversed when compared with the results of the night I 129 DEW-PONDS of August 16. The night was, however, a very cloudy one, and the deposition of dew at all was somewhat remarkable. There was no dew on the opal scale of the thermo- meter beneath the wood, but there were large beads on that portion which projected. The second thermometer was covered with large beads. The third had no dew at all where it overhung the grass, but where it rested on wood there were large beads in the morning. Thus, though it was a night when but little dew, if any, might be expected, yet the wood had so lowered the temperature of the air above it, by cutting off the earth's heat, that dew was deposited. The period of my stay on the downs in 1908 was, on the whole, remarkably free from fog. I have lived, however, many years near the downs, and have constantly rambled over them, and so can speak with a good deal of experience of them. I would add that, though in July, August, and September the downs were very fogless, morning after 130 EXPERIMENT morning, as I looked out from the hill, the Weald was shrouded in mist, with only here and there a higher eminence showing, or the tops of trees or the steeples of churches. In a rain-gauge which was kept at Ditchling, the fog was a measurable quantity when no rain had fallen. This fog was lowland fog and of quite a different nature to downland fog. It rose from the low-lying meadows and ditches, whereas the fogs of the downs are of sea origin and are wafted in from the sea. There are two ponds between Keymer Slipe and Ditchling Beacon which I will refer to as No. 3 and No. 4 ponds. The two ponds are not more than half a mile apart, No. 3 being on the north and No. 4 on the south of the horse-track passing along near the brow of the downs. During the drought of July there was a small shrinkage in the area of both ponds, but this was only noticeable by observing at different dates the clumps of reeds which here and there dotted the edges DEW-PONDS of the ponds, and also marked stones which I placed in certain positions. In wet weather there would be some drainage into No. 4 pond from the horse-track, and into No. 3 from a clump of gorse at slightly higher ground, but in a drought these would be negligible. There must have been something which kept these ponds, as well as that at Ditchling Beacon, fairly filled, although No. i had dried up. In the cases of Nos. 3 and 4, I attribute this to (i) dew deposited on the vegetation growing out of the pond and running down the stalks into the water, and (2) to the depth of the water. It is noticeable that these and other ponds have high banks around, but the growth of common grass low down on the inner side of the bank shows that not for a very long time, perhaps never, has the pond been really full. There is a limit to the depth of a perennial downland pond, and probably about 3 feet is the greatest depth, although those that I have met with do not greatly ex- ceed 2 feet, whilst some are of even less depth. EXPERIMENT As regards the sufficiency of rainfall alone to fill the ponds, if we admit the fact, as White informs us, that these strange little ponds on the tops of the hills are full when those in the bottoms are dried up, that is, in times when there has been a dearth of rain- fall, it must be evident that rain cannot be the sole means of the replenishment of the ponds. Most of the evidence goes to show that the ponds on the lower ground are the first to suffer in times of drought. This fact alone proves the necessity of something acting in addition to rain. The ponds on the South Downs, which I have numbered as i and 3, show this strikingly. No. i (One Tree pond), although 34 feet across (July 22), was very shallow, nowhere being more than a foot deep during the time that I watched it. In the drought, from the middle of July to the end of August, it dried up. Heated day after day by a hot sun, it parted but slowly at night with some of its heat, and the whole of the water was practically of one temperature. 133 DEW-PONDS On July 24, the thermometer showed 67.5 at 8.20 P.M., at i inch, 6 inches, and 9 inches, whilst that on the bank showed 58.5. On July 25, at 1.30 P.M., it showed 71 at i inch, 6 inches, and 9 inches, while the temperature on the bank was 67. Evaporation of the pond's water rapidly proceeded, and on August 10 a few pools alone remained, On August 13 one pool remained, and the tracks of a traction engine on the bank showed that nature had receive assistance in the drying-up process. No. 3 pond, on the other hand, was a fairly deep pond, about 2 J feet deep in the centre. The drought of about four weeks certainly diminished the area of the pond to some extent, but not noticeably, and thermometrical readings showed differences of temperature at various depths. On July 26, at 6 P.M., it was 76 at i inch., 74 at 6 inches, 71 at 9 inches. At 8 P.M. the surface (i inch) had gone down to 70.75, at 6 inches to 71, while at 9 inches it remained stationary at 71. The loss of surface temperature 134 EXPERIMENT through the night would result in a great saving of loss by evaporation. Thus the shallow pond without weed dried up, and the deeper one with weed lost but little, and in the latter case it could not have been rain that maintained the pond, as the fall had been slight (.004 on August 5), and both ponds had equally benefited by it. It is quite clear that something besides rain is required. As regards rainfall, however, an interesting fact came to light. When the hollow had been dug for an experimental pond, I placed a rain-gauge at the bottom, and another on the bank above it. The results were as shown on p. 136. This is to say, that during the period quoted the rainfall into the hollow was nearly f inch greater than on the down. When the rain- fall on the down was y 1 ^ inch or over, that in the hollow exceeded it ; if less than T^inch, then, as a rule, that in the hollow was less. DEW-PONDS In hollow, with igo8 On bank In hollow straw and grass in rim Aug. 24 0-37 0.4! 0-54 Strong S.W. wind 25 O.OI O.OI 26 0.32 0.53 0.69 Gale. 27 0.15 O.2I O.23 Strong wind 28 0.34 o-43 0.48 29 0.09 O.IO 0.08 >> 30 0.05 0.04 O.O2 Hurricane Sept. i 0.32 -45 (blown over) 2 0,05 0.07 Do. Strong wind 3 0.46 0.60 0.80 4 0.08 0.08 0.09 5 O.OI O.OI 9 0.09 O.II >, ii 0.08 0.08 /> 15 O.II O.II ,; 16 0.03 O.O2 20 O.OI O.OI ,; 21 0.04 0.05 22 0.09 O.OI ,; 24 0.08 0.06 ,; 25 0.09 0,12 m 2.87 3-51 The third column shows the rain collected in a gauge in the hollow, in which some grasses and straws were arranged in a cluster in the collecting gauge. The hollow was of course empty, so that the top of the gauge 136 EXPERIMENT may be regarded as being at the level of a pond a foot deep, the grasses representing weeds growing in such a pond. Thus it will be seen that during days of fairly heavy rain- fall this third gauge collected more rain than in the case of either of the other two gauges. See particularly August 24, 26, and September 3. So far as these experiments go, it shows that more rain falls into a hollow than on the bank, being diverted by eddies in the hollow. From this it would seem to follow that where a pond is dependent on rain for its replenishment, the banks should be high, and grasses should be encouraged to grow in a pond, so long as their roots are not allowed to perforate the bottom. There is one fact that is sometimes not completely realized, and that is the large area for drainage which some ponds possess by reason of their broad margins. Three examples will suffice. No. 3 pond had a margin 28 feet across, although part of this was grass-covered. 137 DEW-PONDS The pond was about 67 feet across. The pond area = 3333 square feet, and the area of the margin = 8197 square feet. No. 6 pond (High Park Corner) had an un- covered margin 20 feet across, from the edge of the pond to the grass bank. The pond was 76 feet across. Thus the pond = 4120 square feet, and the margin = 5795 square feet. No. 9 pond (Waterhall) had a margin 15 feet across. The pond was 87 feet across. Thus the pond = 5380 square feet, and the margin = 4618 square feet. The gathering-ground will thus be seen to be considerable, and this, coupled with the fact that the rainfall is considerably greater in a hollow than around its ridge, may be held to show that rainfall may have a great deal to do with the filling of such ponds. Yet not everything, as has been held by some, and this has been, I consider, sufficiently shown by these ponds which have but slightly suffered during drought , although drawn on by hundreds of sheep day after day during the same period. 138 EXPERIMENT As regards the experimental pond which I constructed on the Sussex Downs, at Clayton, and, acting on the principle which had so far been generally accepted, that the pond and puddle must themselves be cooled in order to condense the vapour in the at- mosphere, or at any rate to reduce evapora- tion, I proceeded to select the material which seemed, from numerous experiments, to be the most suitable, and yet was inexpensive. Pond-making is an expensive hobby, and if results are to be in any way of economic value to the farmer and grazier, we must not expect an outlay of 100 on a single pond, which is not to serve his purpose only, but that of at least two or three future generations on the same farm. Straw is fairly cheap, and would serve to some extent, but if puddled clay were laid down upon it, the consequences would be that the straw would become in the first place moist, and therefore act as a good conductor of the heat of the earth to the water and puddle ; and, secondly, 139 DEW-PONDS the straw, by the crushing weight of the clay, would become so flattened out that much of its non-conducting qualities would become lost. Wood has been seen to be a good radiator, and it seemed to me that if wooden planks were made to form a slightly raised platform over the straw, excessive crushing would be prevented, and at the same time the wood would act as a buffer between the moist clay and the dry straw. Also, since wood acts as a good radiator and therefore a bad conductor, I essayed an experiment with what is known as wood-wool, so that beneath the straw was thickly strewn a quantity of this non-conducting material. Thus we had from the bottom upwards, a chalk base, wood-wool, straw, wooden planks, and puddled clay. Next was placed a thin layer of powdered chalk, mixed with cement to prevent any untoward trampling through the clay whilst the pond was under construc- tion, and, over this, powdered chalk was thickly strewn, in order to make a rough 140 EXPERIMENT surface on which the dew would settle, pending the coming of the rains. The layer of clay projected up the sides beyond the limit of the cement, and also beyond the subjacent straw. The whole was surrounded by a wire-netting fence. I have said that it is important that inexpensive materials be used, and I have endeavoured to sug- gest how this may be carried out. A pond of the usual proportions would, however, even with these economical materials, cost between 30 and 40 ; whilst if carried out as done in Wiltshire, with three layers, each of straw and puddle, and the digging out of the hollow in the first place, the expense would be at least 100. During the late autumn of 1908 the pond had filled rapidly, and with the heavy falls of snow which occurred during the winter, which was artificially piled around the edge in accordance with Clutterbuck's suggestion, the result was satisfactory, in that the pond became practically full, although not with 141 DEW-PONDS dew. With the break up of the frost, the pond showed signs of serious leakage. This may possibly have had its origin in the disintegrating power of frost, the damage as usual, appearing when the thaw set in, or it may have been owing to the fact that a deep hole showed itself in the clay bottom where a thick stick had almost, if not quite, pierced it. Visiting the pond in December when the temperature of the air was several degrees below freezing-point, nearly the whole of the water was a mass of ice, and it was quite easy to realize that the sharp edges of the thickening ice may have found a weak spot in the bottom and caused a leakage. Taking a lesson from the " pinching " of an arctic vessel when caught in the ice, it should be noted that it is well to construct the sides of a pond with very gradually sloping banks, so that the ice, instead of piercing the banks, will be lifted as expansion takes place. I took opportunities during the winter to visit the various ponds referred to previously under 142 EXPERIMENT arctic conditions, and found that their margins were well adapted to this end. I may remark that a journey over the downs in mid-winter at night-time, with only the stars to guide one, with the thermometer considerably below freezing-point, and an icy blast blowing from the north-east, is an experience to remember. There was a great shortness of rainfall in February 1909 (.20) at Clayton Mills (584 feet) and the pond completely dried up, with much desiccation of the bed. Subsequent rains showed the centre as holding water ; and on June 5, when I again commenced to reside there, the water surface was about 4 feet across and about 6 inches deep. The bottom and edges I then smoothed down until they were as smooth as glass, and there seemed a prospect of the pond filling, but in a few days I found, on returning home one evening, that a flock of sheep, having trampled down the fence, had drunk the greater part of the water, and had made many holes in the clay with their pointed cloven fore hoofs. Again 143 DEW-PONDS we trampled and smoothed the clay out, and then to encourage the dew, or arrest the rain should that come first, we laid an artificial roof of faggots over the pond, on the principle of the neolithic hut-dwelling. I must confess that it did not have much effect in either direction, but perhaps the experiment was not continued long enough. I should mention that on one occasion on which I placed in position some faggots, supported over the pond, I covered them with a waterproof sheet 6 feet square. On the following morning (June 12) I found considerably more dew on the under side of the sheet than on the out- side. Here there was rising dew moistening the sheet, and this had probably its origin from evaporation of the pond itself. On the following morning, however, there was dew only on the upper side of the sheet. There was no rain from June 10 to June 21 inclusive, at the end of which time the pond was dry, and large cracks had appeared at the sides. A storm on June 24 and following night 144 EXPERIMENT gave 4 inches of water in the pond, and with the swelling of the clay the cracks disappeared. I have referred more particularly to rainfall as affecting the pond, as this was the more noticeable in its effects. Hitherto there had been but little dew, and the only fog on the downs (June 17) had been instrumental in preserving the clay in a damp condition. On July i I determined to excavate the pond to a greater depth, and to make some altera- tions in the bed of the pond. We first re- moved the chalk-puddle, keeping that for future use, and next the clay-puddle and basement of planks and straw. The clay was moist all the way through, although when thrown up on the bank it became in a day or two almost as hard as brick. The moistness of the clay seemed to show that there may have been some slight percolation through the bottom. Next the planks on which the clay rested were removed. These had retained their colour well, and were not decayed. They were moist on the upper K 145 DEW-PONDS side on which the clay rested, and beneath to a slighter extent where they rested on the straw. The straw and the wood-wool, though much crushed, were moist, and well preserved. In laying down the new founda- tions, the planks were laid down immediately on the chalk. On these were laid straw to a much greater thickness than formerly, and on the straw was laid a fresh supply of well-puddled clay. The clay resting on the straw, a greater thickness of clay was neces- sary, as, of course, the straw yielded to each mass of puddled clay as it was put down, and the clay was made to overlap like tiles on a house. It was no easy task to make such a bottom watertight, but it was necessary that it should be done in order to test the truth or otherwise of modern theory. The dew-pond was finished on July 13. This was followed by a few dry and hot days, when artificial watering had to be resorted to to prevent cracking of the clay. However, by July 20 there had accumulated sufficient 146 EXPERIMENT water to enable useful observations to be taken. Ten thermometers were placed in and about the pond as follows : On the grass above the pond. Air 3 feet above the grass. Dry puddle on the bank. Wet puddle at edge of water. Puddle under water. Air immediately over water. Surface water. Bottom water. Six inches above surface. Wet bulb over surface. The Tables showing the results of these observations in detail are published in the Geographical Journal for October 1910, and a summary of the results is shown in the ensuing chapter. The dewy night of July 19 (the first for a long time) made no measurable alteration in the depth of the pond ; and this was so 147 DEW-PONDS on many subsequent occasions. Repeatedly when there had been heavy dews on the grass, I looked to see if there had been any rise in the pond, but without finding it. Not only so, but the dry puddled bank showed no signs of dampness. In the twenty-four hours ended July 28 (7 A.M.) i. 21 inches of rain fell. A rain-gauge in the water of the pond received 1.41 inches. A -rain-gauge at Ditchling (below the hills) measured .88 inch only. In the case of heavy rains, my gauge on the hill generally exceeded that in the valley. In the case of light rains, more water is usually measured in the valley than on the hill. The result of the rain was considerably to increase the water surface of the pond, the depth in the centre being 4j inches. The nights since my arrival on the downs on June 6 had been very free from dew. On the night of August 4 there was a heavy deposit of dew, the fourth night of this kind since the date mentioned. Dew im- mediately appeared on the thermometers 148 EXPERIMENT which were put out at 10 P.M. There was no rain from July 31 to August 15 inclusive, and the water almost disappeared. On the morning of August n there was a thick fog everywhere, but it had not moistened the clay-puddle above the pond, and this re- mained quite hard. This was the first good downland fog for two months. What it did do was to prevent evaporation. On the morning of August 12 there was i inch of water left in the pond. The grass was wringing wet with dew at 7 A.M., but there was none on the hard dry puddle. The wet puddle around the edge seemed to be more wet than usual, which may have been due to capillarity. The pond and mill had no fog around them, but seen from Hassocks station below the hill, two miles away, the mills were invisible, being hidden by a valley fog. It seemed advisable, in order to ascertain the variations of temperature and to test the theory of dew-deposition in a pond, to 149 DEW-PONDS take hourly temperatures for twenty-four hours in a selected dew-pond, and at various positions around it. I selected Upper Stan- dean pond (No. 17) for the purpose. This pond is a chalk-puddle one, and contained very clean water, and had been in existence many years. It is situated at a height of 560 feet, and was 160 feet around the water's edge, with a margin, partly grass-grown, measuring 16 feet from the water's edge to the top of the bank. The experiment commenced at 7.30 A.M. on August 14, a hot sunny day. During the night following, a clear star-lit night, there was a heavy dew, and the cir- cumstances were all that could be desired for the purpose. Fourteen thermometers were placed out in positions in and around the pond. An examination of the detailed Table of results * showed that, during the hours of possible dewf all the temperature of the surface air was, with but one exception (12.30 A.M.), * Vide Geographical Journal, October 1910. 150 EXPERIMENT always lower than that of the surface water, and that on no occasion did that of the water approach dew-point. During the same hours the dry puddle was always colder than the wet puddle. The dew-point over the water was always higher than that over the grass ; and the surface air was never below dew- point. A stick marked with inches and half- inches was sunk in the middle of the pond, and the depth shown was I foot 8J- inches. On September 5 this had increased to i foot 9^ inches. The pond is drawn on largely for sheep and cattle. In one morning three flocks of sheep, numbering three hundred each, and cows to the number of twenty visited the pond. So far as the experimental pond was con- cerned, the rainfall of August 17 (.43 inch) saved it from complete desiccation, and with .78 inch on August 19 and 20 experi- ments with it were again possible. By August 24 there was a depth of 8 inches in the pond, and with .76 inch on August 24, DEW-PONDS the pond was now 10 inches deep. Watching it in a heavy rain, it was observable how that the impermeable clay caused the water to race down runnels which it made into the pond. It must not be supposed that all the water which falls on the bank of a chalk-puddle pond runs down into the pond. Although many ponds have enormous margins, these are proved by testing to soak up almost all of the moderate rains, except that which falls a foot or two above the actual edge of the pond. But this must be modified in the case of one or two ponds, notably that at Ditchling Beacon (No. 5), and concreted ponds when the cracks are kept in a state of repair, such as that on Stanmer Down (No. 30). In heavy rains, however, the area available for drainage is undoubtedly much larger, and in such times when there are artificial runnels emptying through the bank itself, these undoubtedly feed the ponds enormously. Heavy rains quickly saturate the puddle. There is a winding lane leading off the down at Clayton 152 POND AT TOP OF DITCHLING ROSTEL STANMER DOWN POND A concreted pond, kept in repair,- heaps of pebbles shown for use in repairing concrete EXPERIMENT to the farm-pond below. The lower portion is over chalk-marl, the impure chalk at the base of the lower chalk. In ordinary rains the water quickly soaks through. The rain of August 23 and 24 was persistent, and the percolating power of the chalk-marl had almost ceased. Miniature rivers then ran down the lane, and finally made their exit in the pond. The pond is clay puddled, but in consequence of the unusual entrance of numerous small chalky streams into one corner of it, the green water showed their tracks through it in the shape of sinuous white courses, preliminary to the sinking of the finely divided chalk to the bottom. This ceased with a marked diminution of the rainfall. If dew is deposited on a pond, it might be possible to ascertain the amount of accession of such water by floating suitable receptacles on the pond. In order to satisfy myself that such would be a true test, it was necessary first of all to ascertain if the temperature in a confined receptacle floating on the water 153 DEW-PONDS would differ from that of the surrounding water. A series of careful experiments showed that the temperatures were with slight exception identical. On the evening of August 26 I placed a small shallow glass receptacle so that it floated on the water of the experimental pond, and in this I placed a measured quantity of water. The night was a very dewy one, but in the morning exactly the same quantity of water remained. The grass around the pond was drenched with dew. In the rain- gauge .003 inch was measured. There had been no rain or fog. The dew-point at the surface of the water was well below the air- temperature. Similarly, no accession of water took place on the night of August 27-28, although again there was a heavy dew on the downs. The minimum reached on the grass was 43. On the night of August 31 Sep- tember i, a measured quantity of water was reduced during the night by two-hundredths, the night being clear and cold, and the mini- 154 EXPERIMENT mum on the grass being 37^. Further ex- periments of a similar nature were constantly interfered with by the untoward nature of weather ; indeed, throughout the summer of 1909 useful work was frequently at a stand- still from the same cause. On this night also opal-scale thermometers were placed on the grass, the dry puddle, and the wet puddle. On the first there was dew-deposition on both upper and under surfaces of the opal, but in the cases of the last two the dew appeared only on the under surfaces. In the case of that on the dry puddle, this shows that the dew was formed by the rise of vapour from the puddle itself, and thus, so far from puddled clay causing dew to be drawn out of the at- mosphere, it actually parted with its own moisture without in any way feeding the pond. The puddle itself had not become moist. On the night of September i, the same opal- scale on the dry puddle received dew on both upper and lower sides, but the puddle it- self was hard and dry. On the night of 155 DEW-PONDS September 7, a measured quantity of water from the pond was placed in a square tea-tray, so that the level of the water inside the tray was the same as that of the water of the pond on which it floated. On the morning of the 8th, after a night of heavy dew, with a thick fog enveloping the downs till late in the morning, the measured water had increased by three-hundredths of its volume. There had been no rain during the period of the experiment. Two measured quantities of water, placed similarly at 9 A.M. on the morning of September 9, were found to have lost by 6 P.M. ten-hundredths and twelve-hundredths of their volumes respectively. The day was cloudless and sunny, the difference between the two being probably due to the shade of some high-growing weeds, which may have hidden the direct rays of the sun from one of the pans during a portion of the day. It has been argued that the difference between evaporation and precipitation on the 156 EXPERIMENT downs would be sufficient to account for the filling of the ponds by rain alone, on the ground that the rainfall on the downs is greater than in the valley below. It is a fact that the rainfall on the downs is greater than in the vales below. During the years 1909 to 1912 inclusive, a rain-gauge on Clayton Down (584 feet) gave 39.18, 45.73, 45.72 and 43.80 inches. Another rain-gauge beneath the downs, at Ditchling (220 feet), gave 36.73, 38.41, 36.32, and 40.97 inches during the same years. In each case there was an excess in favour of the higher ground. But whilst it must be granted that rain is the chief source of pond-water, it is so irregularly spread over the year that ponds would most frequently dry up, if this were their sole source. The average annual rainfall on downland has been shown by Dr. H. R. Mill to be 32.5 to 35 inches, the result of observations extending from 1868 to 1908. I know of no systematic observations in similar localities in regard to evaporation. Mr. Baldwin Latham has 157 DEW-PONDS conducted experiments at Croydon on an evaporator consisting of a floating copper vessel, i foot in diameter, in a tank of 4 feet internal diameter. We learn from a paper by Mr. P. R. Lowcock * that, owing to the construction of the various evaporators used by observers, many of the evaporation records are unreliable. That used by Mr. Latham seems to be accepted as the most reliable form used, although it may be suggested that, as the amount of evaporation depends chiefly on solar radiation and duration of sunshine, it is possible that the use of a copper vessel may affect the evaporation, especially when we compare the results with those which might be obtained from evaporation from the surface of a chalk or clay-puddled pond. However, the results obtained from Croydon, which, it may be added, have extended over thirty years, are the only ones comparable with evaporation from chalk * " Percolation, Evaporation and Condensation," Q. J. Met. Socy., No. 151, July 1909. 158 EXPERIMENT downs, Croydon being on the northern verge of the Surrey chalk hills, although, I believe, the position of the evaporator was not actually on the chalk, but on tertiaries covering the chalk. The annual average amount of evaporation was here found to be 18.14 inches. There was also found to be an average annual amount of condensation of .36 inch. Taking the difference, namely, 17.78 inches, as the possible net amount of evaporation on the chalk downs, and the rainfall as 35 inches, we have a balance of 17.22 inches of rainfall which will go to feed the pond. From my experience of the downs, I should think that we may safely say that evaporation on the higher parts of the chalk downs would exceed Mr. Latham's figure, especially when we remember that the broad shallow areas of water are com- pletely exposed to solar radiation, and to very high winds. At the same time it would not greatly exceed the value given since evaporation would be restrained by the fact 159 DEW-PONDS that the banks around a pond would tend to retain the cold and more or less saturated air, and to prevent it being blown away by the wind. It is well known that an evapo- rator, to be most efficient, must have its water surface kept flush with the rim, and this is a state of affairs which never occurs in ponds. Possibly this is indeed one of the reasons why a dew-pond never overflows, for the evaporation is thus greatest when the level is highest, but its level is lowered at a rapid rate until it has gone down to such a level as to leave above it and within the banks a certain amount of moist air. As, too, the limit of diurnal variation of tempera- ture passes about 4 feet below the earth's surface, this may have its bearing in limiting the depth of water to be found in a downland pond. It would be safe perhaps to put the evaporation at 20 inches, on the average, although possibly it may be proved to be more. This would give a net annual accretion of water into the pond of 15 inches. But the 160 EXPERIMENT margin of a pond, that is, the area between the water and the top of the bank, is some- times of very great width. As a rule, taken on the flat, even after a wet season, it is not less than twice the area of the water surface. Not all the water that falls on the margin passes down into the pond. Light rains percolate into the soil. I estimate that not more than one-half of the rainfall on the margin runs down into the pond. This would give us another 35 inches, to be added to the difference between 35 inches and 20 inches (the net evaporation), or 50 inches in all. But there is yet the other factor already mentioned, which goes to increase the amount of rain falling into a pond. In one month from August 25 to September 24, 1908, there fell into a rain-gauge placed outside the bank of my experimental pond 2.11 inches. In a rain-gauge placed in the centre, on the bottom of the depression, there fell in the same period 2.98 inches, i.e. .87 inch more than outside, or more than one-third as much L 161 DEW-PONDS again. The deeper the depression and the steeper the banks, the more would this additional rain be drawn into the pond. The proportion would not be so great with a very shallow pond-area, and very low banks, such as, indeed, some ponds have. It would, however, make a substantial addition to the 50 inches already estimated as falling into a pond. It would seem that from rainfall alone, these ponds ought never to dry up, and perhaps they never would, if the rain fell regularly throughout the year, and the evaporation were also regularly distributed. But such is not the case. Mr. P. R. Lowcock* has shown from the records of eighteen British evaporator stations that at fifteen of them mean evaporation steadily increases from January to July, the maximum of the remaining three being in June ; they then steadily decrease to December. He also * "Evaporation from Water Surfaces," Association of Water Engineers. l62 EXPERIMENT showed that nine-tenths of the total evapora- tion occurs in the six summer months, but of the annual rainfall on the downs, I find that in the years 1909, 1910, and 1911, only five-thirteenths, two-fifths, and one-fifth re- spectively of the totals fell in the same six months. If it could be arranged that in these six months the greater proportion of the annual rain fell, we might expect that ponds would actually never dry up. Unfortunately, this is not the arrangement in nature, and yet it is rarely that a well-made pond, free of leakage, does dry up. And for this reason we must, I think, find a second recruit other than rain. This can only be mist and fog, or low-lying clouds, which come to the rescue and make good the irregularities and inequalities of combined rainfall and evaporation. That there are occasions when a rise in a pond cannot be explained by rainfall I am convinced, although I am unable to credit the phenomenal rises which have now and DEW-PONDS again been recorded. In a fortnight's ob- servation of my pond in 1910, paying parti- cular notice to the rises and falls of the surface, there were three nights when the rise, slight though it was, could not be ex- plained by the rainfall. Between 9 A.M. on August 6, and 10 A.M. on August 7, 1910, the pond rose a little over a quarter of an inch. In the same time .11 inch of rain fell. The day had been windy and sunny in turns, with cloudy sky occasionally. There must have been a good deal of evaporation, although I was not in a position to measure it ac- curately. From 10 A.M. on the 8th to 9 A.M. on the gth, the pond rose three thirty-seconds of an inch ; there was no rainfall, but an overcast and cloudy night. Here was pre- cipitation without rain. From 10 A.M. on on the nth to 12 noon on the I2th, there was a rise of three-sixteenths of an inch, with .03 inch of rain. There was a mist all through the night, with some wind, and in the morning the downs were enveloped in mist. 164 EXPERIMENT It might be possible, perhaps, to ascertain to what extent the ponds are indebted to sea- mists for their supply, by taking careful analysis of water from a number of ponds with a view to detect the quantity of saline matter in the water. Sea-salt is sometimes conveyed great distances inland in finely- divided spray, and it is stated that saline particles have been known to traverse half the breadth of our island. When the winds are strong, the spray is carried far inland, and, as the moisture is gradually evaporated, the salt is left in a finely-divided condition in suspension in the atmosphere, and carried in- land. The salt has been detected in almost all soils, and it is thought that a greater quantity will be found along the downs, both in the air and in the ponds than farther inland. There is evidence, also, that even with the process of evaporation of sea-water, there evaporates some proportion of salt. That this is possible was shown by an ex- periment made at the Nauheim salt-works, 165 DEW-PONDS where a plate of glass was placed upon a tall pole between two evaporating houses, which were about 1200 paces from each other. In the morning crystals of salt were found on the glass when the precipitated moisture was dried up, and. these crystals of salt appeared on one side or other of the glass, according to the direction of the wind. Now this deposition of moisture upon the glass can scarcely have been of the nature of true dewfall, so that if sea-salt may be found in our ponds, it may safely be attri- butable to mist. Pallas, it may be mentioned, stated, on the other hand, in 1770, that the dew was salt to the taste in the salt-lakes of Russia in Asia, and this was true as regards dew deposited on plants, collected on smooth surfaces, or on the dress of people who were exposed to it. Here again it seems probable that the term " dew " has not been used in the strict sense of the word. One cannot help thinking that the salt found in the ponds must be attributable to 166 EXPERIMENT mist. The saline particles, flung into the atmosphere by the spray of the waves, as well as drawn up with evaporation, would be seized upon by the vapour of an almost saturated atmosphere, and on a slight lowering of temperature at nightfall would be contained in the resulting mist. It may be of interest to note the small quantity of the saline constituents in sea- water. In water from the Dead Sea, they are 24.58 per cent, by weight. In the Mediterranean there was found to be but 4.09 per cent, by weight. Near Brighton there was 3.52. Bischof gave the saline constituents as 3.52 on the average. Of these constituents about three-fourths consist of common salt (chloride of sodium). With the object of ascertaining what light, if any, would be thrown on the subject by analysis of the water of dew-ponds, I collected specimens of waters, and, by the kindness of Mr. Sidney Skinner, these have been analyzed by Mr. C. S. Grace at the South- DEW-PONDS Western Polytechnic, Chelsea. The results are as follows : ANALYSES OF POND WATERS* IN PARTS PER IOO,OOO Total Tem- Per- Name and number of pond Height Chlorine (by exp.) hardness (by exp.) porary hardness (diff.) manent tiardness (by exp.) No. i. One Tree pond 700 2.12 8.71 3-42 5-29 , 5. Ditchling Beacon . 800 3-475 I3-9I 8.37 5-54 6. High Park Corner . 700 3- I 75 12.41 5-98 6-43 7. Red Worm pond , 9. Over Waterhall Valley 500 700 12.15 4-50 30.20 14.21 23-06 9.78 7.14 4-43 , 10. Oblong pond, over\ Plumpton / 700 3-15 6.14 0.85 5-29 II. Round pond, 50 yds. \ east of No. 10 / 700 2.70 5-20 0.06 5-14 Ramsbury, Wiltshire . / 693 \ \ (inland) / o-75 3-6 f 787 1 St. Boniface pond \ (near I 3-5 4-3 I sea) / Shanklin pond . (near sea) 3-4 7-7 The last three specimens were obtained by Mr. Skinner from the ponds shown. The quantity of chlorine found in these is noticeable. In analyses of waters it is usual to regard the chlorine as derived from sodic chloride, although occasionally it may be present as a calcic salt. The specimen from * The ponds are numbered in accordance with the Table shown in the Appendix, 168 EXPERIMENT No. i pond was obtained after the refilling of the pond, subsequent to the drying-up already referred to. The refilling may have been due in part to mist, but certainly the greater part was rain. It contains the lowest proportion of chlorine of any in the series. It was situated about five miles from the sea. Sutton, in his " Handbook of Volumetric Analysis/' states that unpolluted rain-water usually contains less than one part, but average town sewage about eleven parts. It is scarcely affected by filtration through soil. It is safe to say that there was no question of pollution by animal excreta, and the chlorine came in with the condensed vapour. The rain was brought in with south-west winds and gales, and probably washed down from the atmosphere into the pond a good deal of sea-salt, which had been blown inland. But it is likely also that the mists which blew in from the sea were condensed around finely divided salt nuclei, and this may to some extent have caused 169 DEW-PONDS the presence of chlorine. Ponds which have been existing for a longer time would naturally, by process of evaporation, have a greater proportion of sodic chloride, and Nos. 5, 6, and 9, which have never been known to empty, bear this out. There is no evidence to show that dew will cause salt to be preci- pitated with it when the atmosphere contains a normal amount of the latter, although it has been stated that in the neighbourhood of salt-works, dew is salt to the taste. I think that real dew, being drawn from a thin stratum of air resting on the soil, would scarcely be likely to contain an appreciable amount of salt, although if we extend the term " dew " to include condensation on the branches and twigs of tall trees, from which it may drip into a receptacle or pond beneath, such a form of condensation is as likely as rain to contain chloride of sodium. The word " dew " should, I think, be confined strictly to that deposited in the lowest stratum of the atmosphere. 170 EXPERIMENT No/7 pond is inserted in the Table as a contrast to the others. It is situated at the side of a hill, and there was a good deal of drainage into it. On the face of it, it seems to be a very polluted pond. It was puddled with clay, which lies thickly on the sides of the hills overlooking Pyecombe. The amount of chlorine (12.15) is comparable with that contained in sewage, and probably the extreme hardness was due to the same cause as the excessive chlorine. The tem- porary hardness of Nos. 5, 6, and 9, CaCO 3 for the most part, is no doubt due to the chalk-puddling, although that of No. 6 is less than might have been expected. Nos. 10 and n, bedded in flinty clay, although upon the chalk, show a remarkably small amount of temporary hardness. I am afraid these analyses do not lead us very far in the solution of the problem as to whether dew or mist has much to do with the filling of the ponds. But they certainly show that whether brought as rain, 171 DEW-PONDS mist, or dew, the aqueous vapour comes from the sea. Some analyses of the water contained in certain other ponds have been made by Mr. Wm. T. Burgess, F.I.C., from specimens collected by myself, and the results are appended below. It will be useful to consider these together with those recorded above. Mr. Burgess remarks that " the analyses, taking them on the whole, are what one might expect from rain and surface waters after storage in open ponds on the chalk hills. You will note that the lowest chlorine figure, i.o, was from the pond farthest from the sea. The point which has most interest to me is the fact that several of the waters contained carbonate of soda. In some cases the amount was small, but in two Lewes Racecourse and No. 21 the proportion was quite marked. As far as I know, this is a disco very. " It is interesting to notice the behaviour of the upland ponds during the severe drought 172 EXPERIMENT POND WATERS RESULTS OF ANALYSIS EXPRESSED IN PARTS PER IOO,OOO No. 3 pond, near No. 4 No. 4 pond, near Burnt House Bostel No. 13 pond, over North Bottom, near Heathy Brow No. 17 pond, Upper Stan- dean (received Aug. n, 1909) Total solid residue . Oxygen consumed \ ( 4 hrs.8oF.) / Nitrogen as nitrates Nitrogen as nitrites Chlorine Total hardness Lime (CaO) . 16.8 0.276 0.037 0.005 1.8 *7-3 4.1 19.3 0-542 trace 0.02 2.8 *8.7 5-1 16.6 0.304 trace O.OI 1.75 *9-4 5-4 18.5 0.58 o o 2.0 IO.O 5-9 Very turbid Turbid Slightly turbid Turbid. Hardness all due to carbonate * Hardness practically all due to carbonate of lime No. 1 8 pond, Ewebottom hill, Patcham (received Aug. xi, 1909) No. 21 pond, Piddingworth 580 feet (received Aug. 1 6, 1909) No. 22 pond, Greystones, 620 feet (received Aug. 16, 1909) No. 29 pond, Shambledean Bottom, 350 feet (Aug. 22, 1909) Total solid residue . Oxygen consumed \ ( 4 hrs.8oF.) / Nitrogen as nitrates Nitrogen as nitrites Chlorine Total hardness Lime (CaO) . 30.90 0.768 o o ti 4-9 45-2 o 8.7 16.3 36.7 o o 5-55 16.6 36.8 O o 2.8 14.0 Turbid (con- tained small amount of carbonate of soda). Hardness all due to carbonates Very turbid and dirty water. Car- bonate of soda, 13.0. Hardness all due to carbonates Very turbid and dirty water. Car- bonate of soda, 3.7. Hardness all due to carbonates Very turbid and dirty water. Car- bonate of soda, 5.0. Hardness all due to carbonates 173 DEW-PONDS POND WATERS continued. No. 12, No. 30 pond, Stanmerdown (concrete), Sept. 7, 1909 Cemented pond, south-west from Lewes Racecourse, Riddlesdown pond, Surrey (Aug. 25, 1909) 390 feet Total solid residue 41.4 43-9 25-7 Oxygen consumed (4 hrs. \ 80 F.) J 2.4 0.511 0.36 Nitrogen as nitrates . o trace traces Nitrogen as nitrites . o Chlorine .... 4-5 i'7 I.O Total hardness . 12.7 2.7 15.6 Lime(CaO) 1.6 Very turbid and dirty water. Car- bonate of soda, 2.4. Hardness all Turbid. Carbonate of soda = 14.3 Slightly turbid. Two-thirds of hardness due to carbonates due to carbonate of 1911. A very large proportion of the ponds dried up, but the fact remains that here and there was a pond which stood the severe test very well, and maintained a good supply all through the summer. Even in these the end of the drought found them considerably lower than in the spring. This was, of course, to be expected, but the point is that there was water actually remaining when most of the ponds near by were dry. It is unlikely that they would ever be put 174 EXPERIMENT to a severer test than in 1911. Three of those which I had watched during my three summers' stay on the downs at Clayton, stood the drought well. One was at Ewe- bottom Hill, Patcham ; the second was at Upper Standean, this being the pond on which I conducted experiments in 1909 ; and the third was at the head of Ditchling Bostel. In addition there was one at Pidding- worth, which showed practically no diminu- tion, but this is partly overhung on one side by trees. The pond at Ditchling Bostel deserves special mention. It is close to the top of the road which leads up the Bostel from Ditchling on the north, and it is also at the head of a long winding chalk valley which leads up from the south-east. It is not a little remarkable that this should have retained so much water throughout the dry summer. It was certainly decreased in area, but I think that we have evidence here that a position at the head of a mist-laden valley leading up from the south is advantageous 175 DEW-PONDS to the maintenance of a pond. I think that it is clear that moisture-laden winds from the south, condensing into mists during the early morning have had a good deal to do with the maintenance of the supply. There was an almost continuous absence of dew well into August, and reports of this reached me from different parts of the country. The Duke of Northumberland noticed it at Brentford. Mr. H. C. Brentnall reported similarly from Marlborough, and Mr. Heywood Sumner from Fordingbridge, Hants. I can speak for the South Downs (central) and for south London. If we do not allow that mist has been responsible for the maintenance of the ponds, then we must fall back on rain. To the end of August there was, however, only 13.74 inches of rain on the downs, as compared with 28.63 inches for the corresponding period in 1910. It seems to me to be clear that loss by evapora- tion was in part overcome by the aid received from condensation from mists. 176 EWEBOTTOM HILL POND Showing great width of margin, also ridges at successive levels of pond UPPER STANDEAN POND, SUSSEX Banked up on E. and S. E. sides EXPERIMENT Mist-particles would fall by gravity alone, as has been seen, into the pond-depressions. Of course they would also fall elsewhere, but the porous chalk would not hold them up as would the puddled bottom of a pond. With a minimum of dust-particles in the air, would not the vapour seize upon the salt- dust in the air with avidity, and weighing them down by condensation around them, make them heavy and liable to fall fairly rapidly ? The analysis of pond - waters showed how great was the quantity of sodium chloride contained in them. I think that certain ponds maintained a fair supply of water by the assistance of mists. Still, I think that even these would not have be- haved so well had it not been that they were fairly deep, and so started with a good supply of water when the drought caught them. Almost all the ponds on Salisbury Plain suffered desiccation during the drought. On Martinsell Hill, the well-known dew-pond M 177 DEW-PONDS dried up, although it had not done so for seventeen years. The loss of water and of " feed " for the sheep was disastrous to the upkeep of the flocks. If science can do anything to prevent similar economic failures as these, it will have given a satisfactory raison d'etre, and the scientific study of dew- ponds will have been justified. In the first place, farmers do not dig their ponds deep enough. Then they neglect them until they dry up. Then when they begin to repair them, they remove the puddle in the process of cleaning, and when it rains they wonder why the ponds fail to hold water. Near the foot of Martinsell Hill, on the road to Wootton Rivers, there was a pond which on September 18, 1911, was full of water, although the ponds on the hill were dry. But this pond was almost surrounded by trees, some of which overhung the pond. Evaporation was here reduced to a minimum, and there was no doubt considerable con- densation from the trees. Close at hand, EXPERIMENT on open ground, there was a square pond basin, but this was quite dry. At a meeting of the Royal Society of Arts on March 3, 1909, Mr. W. R. Stratton men- tioned that in the Colesburg District of Cape Colony he had noticed certain ponds on farms where the water in the dry season was higher in the morning than in the evening. There were no springs, and yet the ponds were always full of water in the driest weather. Mr. J. P. Johnson refers in his " Notes on Orangia " to the speculations which have been made as to the origin of the " pans " of that country. He says that they have flat and level bottoms, and the rain rarely fills them to more than 2 or 3 feet, and they appear to be found only where the water is held up on one side or other by one or more volcanic dykes, and along such a dyke there may be a succession of pans. Their formation is thought to have been caused by the constant trampling of cattle after a rainfall, the shale being disintegrated 179 DEW-PONDS by the feet of the animals, and the material afterwards being removed by the wind. It seems to have been overlooked that the trampling would have resulted in just that puddling which is a necessity in dew-ponds, and possibly they have persisted until now, for this very reason, although the rainfall is, as he suggests, much less than formerly. It might be of use to adopt the places where these pans persist, and make them into dew-ponds, by artificial trampling them with the usual farm cattle. As the silt in some of the larger pans is highly impregnated with salt, there may possibly be reason to think that the saline particles may have acted as carriers of vapour into the pan, and that the rain is not the only recruit. The pans, it should be added, are not confined to Orangia. Next to rain as the chief means by which the ponds are filled we may safely place mist and fog, especially as regards those ponds which are at the very summit of the 180 EXPERIMENT hills, where the condensation resulting from altitude is greatest. We have seen that the quantity of sodium chloride in the ponds is large. I think much of this may be attri- butable to its precipitation therein as the nuclei of mist-particles. In the early morning I have seen the mist gradually forming, not only in the valleys, but on the hills. Not always do they creep up the valleys from the sea, as has been alleged. If that were the case, I am afraid many of them would dis- appear in the rising sun before they reached the escarpment. Charged as the atmosphere always is with particles of common salt and other salts, the chills of early morning requisi- tion these particles as the carriers of condensed aqueous vapour, by reason of their affinity for moisture, and when the mist is on the move, as it almost always is, much of the mist and salt-particles are mechanically deposited by gravitation in the ponds as the mist passes over. 181 CHAPTER V SUMMARY AND CONCLUSIONS A DEW-POND is one situated on the higher grounds, generally on the chalk downs of the south of England, which retains by some means or other a supply of water throughout all but the most prolonged droughts, whilst those ponds situated on the lower lands have consistently dried up. It has been held that, since on clear summer nights the grass of the downland heights has been observed to become wringing wet with dew, it has followed that in some mysterious way the dew has also fallen into the pond, and has made up in summer droughts for the loss the pond has sustained by evaporation and the watering of cattle. How dew could be deposited on water no one had attempted to show, until 182 SUMMARY CONCLUSIONS the theory was broached in recent years that such a pond was specially constructed, by having a quantity of straw spread out on the basin excavated for such a pond, before the puddled chalk or clay was laid down as the waterproof bottom. It was stated that by this means the water was kept cool, the rising heat of the earth after nightfall being kept away from the water by the badly- conducting straw, whilst at the same time the water lost its heat by radiation. It was a pretty theory, but no explanation was given as to how the clay-puddle was to be made to behave properly, upon a loose, moving bed of straw ; nor how the straw was to be prevented from being pressed down into a hard layer under the heavy puddle, and becoming practically of no use as a non- conductor of heat. According to this theory, the water having been cooled to a tempera- ture below dew-point, dew was deposited on its surface, and thus the pond was replenished. It was entirely overlooked that when the 183 DEW-PONDS surface layer of a pond commenced to cool it immediately sank, and that during the process of cooling convection currents must be set up before and until the pond had become cooled as a whole, that the short nights of the summer months were all that were available for the process, and that very rarely could the whole of the pond be reduced in a few hours to a temperature below dew-point. The point to be decided was whether it was possible for water to accumulate in a prepared upland basin, where it was exposed to the influences of the weather, and where no trees or verdure existed other than the grass of the downs around. The banks of the ponds, whose behaviour was to be under consideration must, with the exception of grass, be absolutely bare of vegetation. It was not denied that a tree overhanging a pond would condense a large amount of moisture out of the atmosphere, and that this would drip into the basin and so feed the pond. This might happen during nights 184 SUMMARY CONCLUSIONS of heavy dew-deposition, as well as when the downs around were covered with fog. But in neither case, as I hold, are we correct in regarding this replenishment as due to dew, although, at any rate as regards the former, it is merely a matter of nomenclature as to whether we regard our condensed vapour on overhanging trees as dew. In both cases a pond would probably be kept " alive " by these means, and they would make good the irregularities in the supply of rain. But what it was necessary to establish or refute was whether an open pond, with no over- hanging trees whatever, was replenished by the dew which is seen so plentifully deposited on the grass around. One observer remarked that when there was dew on the grass, every pond must be receiving moisture, but no proof was given of so sweeping a statement. I have found no evidence in its favour. Another went so far as to say that the dew which was deposited on the grass must of necessity run down into the pond, oblivious 185 DEW-PONDS of the fact that the margins of most ponds are always more or less permeable, and remain so until saturated by a heavy rain ; and this being the case the small quantity of water resulting from dewfall on a very dewy night would but moisten the ground around the roots of the grass, leaving none to run away down the slope of a pond. It is a very rare thing to come across a puddle of water, or even a muddy spot, amongst grass that has not received any other replenishment than dew ; if dewfall were so heavy as has been assumed, I fear many of our well-kept lawns would frequently suffer. As a matter of fact, the total annual deposit does not exceed 1.5 inches as estimated by Mr. Dines, and the result of an experiment which I made on a night of very heavy dew, gave an annual dew-deposit of .77376 inch, or a little more than three-fourths of an inch. On the face of it, it seemed absolutely ridiculous to imagine that dew, such as we know it, could be in any way responsible for filling and maintaining 186 SUMMARY CONCLUSIONS a pond. It was, therefore, necessary to ascertain if it were likely that a pond by some speciality in its construction could be made to induce a far greater deposition on its own surface or on the margin around it, than on grass. It should, perhaps, be stated in passing, that the margins of all the best ponds are kept scrupulously free from grass, so that even if dew which fell on the grass around did not percolate into the ground but ran down into the pond, the only grass which would collect this useful dew would be that which grew on the inner side of the bank. But occasionally there is no bank at all, especially when there are overhanging trees, When the pond is partly on a slope it has to be banked up, and often a bank has been formed merely as a convenient place in which to deposit the excavated material of the basins. Some downland ponds are chalk-puddled, some are clay-puddled. Some have straw DEW-PONDS used as layers in some portion or other of their construction. Some have not. It has been suggested that after nightfall there is a substantial lowering of the temperature of the puddled margin around a pond, and that this reduction frequently brings the temperature below dew-point. At the same time the water parts with its heat, it is held, till dew-point is reached in this case also, and the pond receives an accession of moisture into itself. Again, it is held, straw is placed under the foundation of a pond, in order that the rising heat of the earth may be cut off from the pond, and thus may not interfere with the radiation of heat from the pond which is taking place. In order to test this theory I constructed a pond, selecting materials for its foundations which were comparatively inexpensive, and such as would be at the command of a farmer who might wish to construct a pond, at a cost, say, of 30 or 40. Straw is, of course, a bad conductor of heat, but as this does not 188 SUMMARY CONCLUSIONS give a firm basis on which to place a bottom of puddled clay, I placed planks of wood between it and the clay. Immediately be- tween the chalk and the straw I laid down a thick layer of wood-wool. I thus had three layers of inexpensive non-conducting material beneath the puddle. For the puddle I selected a fine red clay, this being more quickly worked into a fine puddle than chalk. I may say that these materials were selected after hundreds of experiments on their various heat-conducting qualities, and my note-books record these in full. Thermometers were placed upon, or in, or under each in turn alone, and then upon, or in, or under various combinations of the different materials. On nights of free radiation, there was no doubt that the straw, wood-wool, and wood inter- fered considerably with the rising heat of the earth. They were, however, dried in each case before being used, as, of course, if wet they become fairly good conductors. The straw and wood-wool were kept loose in 189 DEW-PONDS texture, so as to increase their non-conducting powers. Wood proved to be very useful on clear nights in keeping away from thermo- meters laid above it, heat derived from terrestrial radiation. But the clear nights, when rapid radiation took place, were in the minority. Taking the nights of the three months from June 6 to September 8, when a pond required all the help it could receive from sources other than rain, on not more than twenty-four nights were there suitable atmospheric conditions to give any dew on the grass. The whole question is one which depends on the replenishment of a pond in those months. In other months its level varies only to a small extent, and there is little possibility of it becoming dried up. It is in the months when there are spells of dry weather with much evaporation that the qualities of a downland pond are tested. At the commencement of a drought, unless there is at least a foot of water in the centre of a pond, a few weeks with little or no rain 190 SUMMARY CONCLUSIONS will see the pond completely dried up. Nothing is more striking amongst the numerous ponds I have examined than their shallowness. A pond measuring 40 feet across will not be more than 2 feet deep in the centre. This means that an enormous area is exposed to evaporation, and under a summer sun its temperature rises. Water is not so rapidly subject to extremes of temperature as the atmosphere, but the heat that it does receive it parts with more slowly. The rapid radiation which takes place from blades of grass so reduces the grass in temperature that it chills the air in contact with it to a degree below dew-point. Hence the grass quickly receives dew. If water in a pond is to receive dew in a similar manner, it must be also chilled by the radiation of its heat. Dew which forms on the grass around certainly never runs down into a pond. If the chalk or clay margin had similar radiating powers to grass, dew might form thereon, and, if the puddle did not allow of percolation, the water 191 DEW-PONDS might run into the pond. But, firstly, the puddle scarcely ever gets chilled sufficiently to form dew ; secondly, the puddled margins have twenty-nine times out of thirty been sufficiently porous to soak up all but heavy rains, and, thirdly, if the whole of the margin did receive dew, the quantity would be so small as to form no appreciable addition to the pond. Once grasp the principles on which dew is formed, and it becomes more and more difficult to see that dew has any share in the replenishment of a pond. Dew on the downs, although so apparent, must be put on one side. The filling of a pond must take place from causes acting within its own area, and on this area alone must atten- tion be concentrated. Dew on the pond- margin is an impossibility so far as my observations go, and our inquiry practically concerns the water itself. The various temperatures which I observed in connexion with the pond which I made on Clayton Down are shown in the Tables 192 SUMMARY CONCLUSIONS which are printed elsewhere. Together with those observed at various stations in and about a larger pond on Stan dean Down,* these showed that out of a total of sixty-five observations at all times of the night, when the dew-point was ascertained, only on five occasions did the water surface go down below dew-point. Two of these were at 7 A.M., one was at 7.30 A.M., one at 8 A.M., and one at 8 P.M. Thus : Aug. 7 7 A.M. Water 56-5 Dew-point 57-7 Aug. 12 7 A.M. 60 61-4 Aug. 8 8 A.M. 62 63-2 July 29 8 P.M. 58 59 Sept. 6 7.30 A.M. 57 59 It will be seen from this Table that the temperature was on no one of these five occasions greatly below dew-point. As regards the air resting on the water, this was on fifteen occasions below dew-point, but only on one of these were both the air and the surface water together below dew- * Vide, Chap, iv., p. 150. N I 9 3 DEW-PONDS point. This was at 8 P.M. It seems probable, therefore, that only on this one occasion was any dewfall possible. There were four occasions only when the surface air was at dew-point, but if any dew were then given out, the latent heat which was manifested in the process would at once have sent the temperature above dew-point. On four occasions the surface air and the surface water were equal, but two of these occurred after sunrise. There were fifteen times only when the surface air was higher than the surface water, but thirteen of these occurred also after sunrise. The remaining two were at 9 P.M. on August n, and at 12.30 A.M. on August 15. I mention these particularly, since they bear out the well- known fact that air loses its heat much more rapidly than water. If these facts be considered, it will be seen that they lend but little countenance to the theory of dew-deposition in a pond sufficient to act as an efficient recruit to its supply. 194 ROTTINGDEAN POND, EMPTY Top of Down, high bank all round SHEEP WATERING AT UPPER STANDEAN POND SUMMARY CONCLUSIONS The five occasions when the water was found to be below dew-point may show that these conditions prevailed during a total of twelve hours in all, this being a very liberal estimate. The sixty-five occasions when tests were made represent the twenty-four dewy nights during a period from June 6 to September 8, that is three months. The number of dewy nights throughout the year would not be more than twice that number, and if the total number of hours during which surface water was below dew-point be doubled similarly, we should have twenty-four hours for the whole year when there may have been deposition of dew in the pond. Whether this method of calculation be correct or no, the fact re- mains that on five occasions only in three months was the water found to be below dew-point. From this it can, of course, be easily judged how infinitesimal is the quantity of dew which thus reaches the pond. So far as the deposition of dew on the wet or the dry puddle is concerned, the proportion 195 DEW-PONDS of dew-producing occasions is slightly greater, but it would be idle to suggest that the twelve or nineteen occasions respectively, when thermometers placed closely touching them went below dew-point, would allow of any appreciable quantity of dew being de- posited. In the case of the dry puddle I observed one occasion when this was appreciably moistened by dew. The experimental pond was completed in September 1908, at the close of a number of experiments made with the object of choosing the best plan for the foundations. Owing to the lateness in the year, it rapidly filled with the coming of the autumn rains. The appearance of a new pond on the downs gave rise to a good deal of curiosity on the part of pedestrians so much so that it was damaged in the early spring of 1909 by some individual who endeavoured and succeeded in pushing a stick through the foundation and draining the water away. The severe spells of weather in the winter 196 SUMMARY CONCLUSIONS of 1908-9 may, too, have injured the founda- tions. On two occasions the pond was a mass of ice, and the ice may have pierced its way through the edges of the pond, and so damaged the puddle. But herein is shown that the shallower the bed, and the more gradual the rise of the bank, the more likely would the pond be able to resist the thrust of the ice as it expanded in passing from the liquid to the solid form. As a matter of fact, all the ponds I have examined have a very gradual rise from the pond-area to the banks. I have never been able to see how straw used in the foundation of a pond, and placed under puddled clay, could for any long period remain dry ; as a matter of fact, it must be considerably moistened immediately the wet puddled clay is placed upon it. And owing to percolation in the margin of a pond, the straw, if carried right up to and under the bank, must become more moist as time elapses. It is not as though the whole area was at any time full of water. A margin 197 DEW-PONDS always remains, and this margin either cracks under a hot sun, or becomes so dry as to allow rain to percolate into and through it when heavy autumn rains commence. When I decided in the spring to dig up the pond again, and remake it on slightly different lines, I particularly examined the straw as it was taken out, and found it was moist, completely crushed, and very brittle. On the first occasion the foundations of the pond were (i) chalk bed, (2) wood-wool, (3) straw, (4) planks of wood, (5) clay-puddle. When I relaid the pond, they were as follows : (i) chalk bed, (2) planks, (3) straw, (4) clay- puddle. In laying clay on the straw, the straw was, of course, immediately crushed down, and of necessity moistened to some extent. The thickness of the straw was such, however, as to allow of it being a fair test as to whether straw so laid would prevent the warming of the water at night by the upward radiation of the earth's heat. The results have been referred to, and I am bound 198 SUMMARY CONCLUSIONS to say that the theories which have received some amount of popularity in regard to the action and construction of dew-ponds, receive but scanty support from these results. When we turn to consider the matter of rainfall, we find that the total annual fall would be sufficient to fill the pond, if the rain fell for the most part in the summer months. This of course is not the case. Rain is fairly regularly distributed over the year. But nine-tenths of the total annual evaporation take place in the summer six months. Small wonder then that they do occasionally dry up. That they do not more frequently can only be attributable, I think, to the fact that downland fogs come to the rescue, and, in the summer months, make good the unequal relations existing between rainfall and evaporation. My intimate experience of the characteristics of the fogs of the downs assures me of the part they take in assisting in the replenish- ment of dew-ponds. Much fog is deposited 199 DEW-PONDS mechanically in a hollow as it passes over, and there is every possibility also that the silent discharge of downland electricity from the rounded surfaces of the downs may result in the precipitation of fog. The term " mist-pond " has been found in Surrey, Kent, and Wiltshire by Messrs. W. Johnson and W. Wright, as recorded by them.* Mr. T. W. Shore found these ponds known by the name of " cloud-ponds " in Hampshire, and the word " fog-pond " has been known at Hampstead for at least half a century. So far as popular tradition goes, any one of these names is as deserving of credit as the word " dew-pond." We have, as I now hold, in the word " mist-pond " the clue to the perennial nature of the ponds. As a matter of fact, I have spoken with farmers with whom the word " dew-pond " is in familiar use, who agree that when they speak of dew they include in that word condensation from mist and clouds also. * "Neolithic Man in North-East Surrey," p. 47. 200 SUMMARY CONCLUSIONS Chemical analyses of a number of upland pond-waters have shown that they contain a large proportion of sodium chloride and other salts. These must have been brought in from the sea, and the particles must be present in comparatively large quantities in downland air. These are seized upon as nuclei of condensation when the night-mists form on the downs, and as the mists blow up in the early morning from the sea they pass across the pond-depressions and are deposited in quantities there. 201 APPENDIX SOME details of the various ponds which I have had under observation are appended. No. i. One Tree pond, overlooking Rag Bottom, 750 feet O.D. : Frequently dries up ; very shallow trough ; fed by runnel from track during heavy rains. Circumference of water-edge on June 28, 1908, 75 feet ; on December 29, 1909, 104 feet, after heavy rains. The rain of the last three months of 1909 increased the areas of the various ponds enormously. This will be seen in the cases of other ponds as shown below. No. 3. Potamogeton pond, top of Burnt House Bostel, 730 feet O.D. : Analysis made ; much weed and grass and ranunculus when first seen ; cleaned out in September 1909. Circumference of water, 142 J feet on August 23, 1908 ; 6 feet to grass bank ; drainage from runnel only in heavy rains. Circumference on December 29, 1909, 165 feet, all the long grass in the pond being covered. No. 4. Grass pond, south-east from No. 3, overlooking Home Bottom, 716 feet O.D. : Analysis made ; much grass in pond and reeds round edge when first seen ; 202 APPENDIX grass almost to water's edge ; cleaned out September 1909. Shingly soil ; no drainage from road except in heavy rains. A breeding-place for may-flies. The cleaning of this and No. 3 pond evidently stopped sus- pected leakage, and on December 29, 1909, the circum- ference had increased to 138 feet. No. 5. Ditchling Beacon, top of Ditchling Bostel, and overlooking Big Bottom, 750 feet O.D. : Seaford Cliff seen to the south-east ; some tufts of reeds growing in pond, which would be good for dew ; many newts and limncea and planorbis snails ; a chalk-puddled pond. This pond is at the head of a long winding valley, and has not been known to run dry. At times of heavy rains there is probably much drainage from the chalky road. The water is very clear. Many dragon-flies ; some sludge in pond. On December 29, 1900, all the reeds were covered by the water, and the surface had encroached on the grass bank. The circumference was 245 feet. No. 6. High Park Corner, near, 620 feet O.D. : A large pond of very clear water ; chalk-puddled ; water- edge marked by a ridge ; no weed projects above surface ; on summit of down. Circumference of water, 227^ feet on August 12, 1908 ; 20 feet to grass bank. No. 7. Red-worm pond : Dale Hill : A very foul, clay-puddled pond ; mud permeated with errant red- worms ; much surface drainage from soil ; on side of hill ; many rushes in pond ; puddled with red loamy clay. No. 8. Bushes pond : Dale Hill : Clay-puddled ; 203 APPENDIX water not clear ; fed by drainage from field, and dew from overhanging trees and bushes ; near No. 7. No. 9. Waterhall Valley, over, 400 feet O.D. : A beautiful pond, comparable with No. 6 ; water very clear ; little, if any, drainage from down ; weed beneath surface. Circumference of water, 260 feet ; 15 feet to edge of grass ridge on September 13, 1908. No. 10. Oblong pond, over Plumpton : 720 feet O.D. : Puddled with white loamy chalk ; margin covered with angular flints ; the shape is unusual, the dimensions being 50 feet long, with an average breadth of 25 feet, on September 6, 1908 ; 147^ feet in circumference. A good deal of weed beneath the surface. No. ii. Circular pond, near to No. 10, 720 feet O.D. : Very shallow ; much reduced in drought ; comparable with No. i. Circumference, 140 feet on September 6, 1908. Pond-weed under surface. No. 12. Cemented pond, on the Buhner Down, 390 feet O.D. : Analysis made. No. 13. Chalky pond, near North Bottom, on slope of hill, 570 feet O.D. : 117^ feet in circumference ; water very clear ; margin of smoothed white chalk. Not on summit of down ; analysis made. No 14. Rottingdean, near Old Mill : Dry many years ; covered with grass ; on highest ground of neighbourhood. An old stone-breaker stated that he had never known this pond to have contained water. He had served in the Baltic Fleet in Crimean days, and had settled in the neighbourhood soon afterwards. 204 APPENDIX No. 15. Rottingdean, north-west of : Dry, and over- grown with grass ; a clay-puddled pond. No. 16. Rottingdean, north-east of: A cemented pond ; circumference (July 8, 1909), 130 feet ; 13 feet across margin ; water foul, receiving drainage from farmyard. No. 17. Upper Standean, 560 feet O.D. : Water clean ; chalk-puddled ; analysis made. Circumference of water August 14, 1909, 160 feet ; chalky margin, 16 feet across. On December 29, 1909, the circumference had increased to 206 feet, and the margin had decreased to I2j feet across. No. 18. Ewebottom Hill pond, 450 feet O.D. : Cir- cumference of water on July 14, 1909, 165 feet ; 20 feet across chalky bank ; hard puddle. Two well-defined ridges round pond at 5 feet and 9 feet from water's edge, showing limits of former extensions of the pond. A dry runnel on north-east side, but grass-grown ; quantity of pond-weed under surface, with eggs of fresh-water snails. On December 29, 1909, the circumference had increased to 201 J feet. Analysis made. No. 19. South Hill, Pangdean, 550 feet O.D. : Dry ; concreted, but concrete broken up. Concrete made of fine gravel and occasional shell-fragments, about f inch thick. No. 20. Over Donkey Bottom, Standean, 400 feet O.D. : Dry ; a well-made pond, on side of hill. No. 21. Piddingworth, 580 feet O.D. : Analysis made ; water green with algae ; water not clean, very green ; much used by cattle. Circumference of water, 170 feet 205 APPENDIX on August 8, 1909 ; 15 feet across white-puddle margin. Analysis made. No. 22. Greystones pond, Standean ; 620 feet O.D. : Analysis made ; circumference of water on August I, 1909, 95 feet ; weeds almost to water's edge ; large crater-like trough ; surrounded by large masses of Tertiary sandstones (sarsens) ; much pond-weed and grass in water. Circumference on December 29, 1909, I22j feet ; water's edge to rise of bank, 22j feet. No. 23. Plumpton plain, 680 feet : Dry ; overgrown with grass. Nos. 24-26. On Bulmer Huff, 500 to 530 feet O.D. : All dry and overgrown. No. 29. Shambledean Bottom, 350 feet O.D. : Analysis made ; circumference, 77^ feet on August 22, 1909 ; 15 feet across margin ; much drainage. No. 30. Stanmer Down, 430 feet O.D. : No runnels ; analysis taken ; 112^ feet in circumference on September 4, 1909 ; 20 feet across margin. This is a carefully concreted pond, apparently kept watertight by constant repairs ; on the summit of the down. No. 31. High Park Wood : Pond dry September 7, 1909 ; 500 feet O.D. ; 63 feet bank to bank. In centre of pond is a perforated pipe projecting about I foot, from which the water collected is made to flow to a farm lower down. No. 32. Pyecombe Golf Links : Pond dry September 8, 1909. 206 INDEX AITKEN, Dr. J., 44, 47, 69 Analyses of pond-waters, 168, 172 Ancient encampments' water- supply, 12 " Artificial rain-ponds," 27 BANKS of ponds, 187 Blythe, E., 43 Brabourne, 27 Burgess, W. T., analyses by, 173 CHALK districts, ponds in, 13 Chalk-puddling, 14, 95 Chalk streams, 1 5 Chanctonbury, 60, 83 Cholsey pond, 30 Cissbury, 25 Clay-puddle, 109 Clayton downs, 105, 112 " Cloud-ponds," 200 Clutterbuck, 27, 84 Concreted ponds, 1 52 Construction of ponds, 83 Convection currents, 74 Cornish, 26, 77 Cost of a pond, 89, 141 DEW, 39 estimate of , 54, 126 how formed, 41, 191 Dew -pond sections, 106 Dines, G., 54 Drought of 1911, 174 Dust-particles in air, 71 ELECTRICITY in rain, 79 Essex ponds, 93 Evaporation, 157 Experimental pond, 139, 198 Experiments in ponds, 112 et seq. FALMER downs, 27 Flemings, 32 Fogs on Sussex Downs, 79 Fog-pond, 200 Frost on palings, 48 GRACE, C. S., analyses by, 171 HIGH Park corner, 114 Hubbard, 26, 88, 100 INDIA, ice in, 52 JACK and Jill, 36 Johnson, W., 27 KEYMER Slipe, 131 LATHAM, Baldwin, 158 Lewes ponds, 91 Ligurian Alps, 67 2O7 INDEX Lowcock, P. R., on evaporation, 162 Low-level ponds, 9 MAIDEN Castle, 92 Marlborough ponds, 97 Margins of ponds, 137 Mist, 38 Mist-ponds, 58, 200 NEOLITHIC man, 17 ONE Tree pond, 119, 133 PACKE, C., 27 Palaeolithic man, 16 Pans, 17 in South Africa, 179 Puddling by cattle, 104 RAIN, 38, 135, 199 Red-worms, 109 Reid, Clement, 35, 59 SALISBURY Plain ponds, 177 Salt-dust in atmosphere, 165 as carrier of vapour, 183 Scott, R. H., 77 Slade, 26, 85 Sheep on downs, 31 drinking from ponds, 37 Snow, use of, 141 Straw, 52, 56, 109, 183 Sussex downs, 78 TABLE Mountain, 63 Thermometric results, 193 UPPER Standean pond, 1 50 VEGETATION around ponds, 61, 184 WATERHALL pond, 1 1 3 Water-level in chalk, 20 Wells, Dr. W. C., 40, 47 White, Gilbert, 28, 37, 43, 50 Wiltshire ponds, 94 Winklebury Camp, 19 Wood as a foundation, 127 Worms, 33 YORKSHIRE Wolds, 98 Printed by BALLANTYNE, HANSON & Co. LTD. At the Ballantyne Press LONDON AND EDINBURGH UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. J)ut/w; H^*&A C*i , ^1119560 MAY 8 '* LOAA; r \j TN NTER-LIBRARY LOAN MAY! 3 1970 rr "> iL tt ir af It- ^ ^ LD 21-95t-ll,'50(2877sl6)476 10907 UNIVERSITY OF CALIFORNIA LIBRARY