The evaporation of the juices of trees and other plants is doubtless their most important thermoscopic function, and as recent observations lead to the conclusion that the quantity of moisture exhaled by vegetables has been hitherto underrated, we must ascribe to this element a higher value than has been usually assigned to it as a meteorological influence.

The exhalation and evaporation of the juices of trees, by whatever process effected, take up atmospheric heat and produce a proportional refrigeration. This effect is not less real, though to common observation less sensible, in the forest than in meadow or pasture land, and it cannot be doubted that the local temperature is considerably affected by it. But the evaporation that cools the air diffuses through it, at the same time, a medium which powerfully resists the escape of heat from the earth by radiation. Visible vapors, fogs and clouds, it is well known, prevent frosts by obstructing radiation, or rather by reflecting back again the heat radiated by the earth, just as any mechanical screen would do. On the other hand, fogs and clouds intercept the rays of the sun also, and hinder its heat from reaching the earth. The invisible vapors given out by leaves impede the passage of heat reflected and radiated by the earth and by all terrestrial objects, bat oppose much less resistance to the transmission of direct solar heat, and indeed the beams of the sun seem more scorching when received through clear air charged with uncondensed moisture than after passing through a dry atmosphere. Hence the reduction of temperature by the evaporation of moisture from vegetation, though sensible, is less than it would be if water in the gaseous state were as impervious to heat given out by the sun as to that emitted by terrestrial objects.

Total Influence of the Forest on Temperature.

It has not yet been found practicable to measure, sum up, and equate the total influence of the forest, its processes and its products, dead and living, upon temperature, and investigators differ much in their conclusions on this subject. It seems probable that in every particular case the result is, if not determined, at least so much modified by local conditions which are infinitely varied, that no general formula is applicable to the question. In the report to which I referred on page 163, Gay-Lussac says; "In my opinion we have not yet any positive proof that the forest has, in itself, any real influence on the climate of a great country, or of a particular locality. By closely examining the effects of clearing off the woods, we should perhaps find that, far from being an evil, it is an advantage; but these questions are so complicated when they are examined in a climatological point of view, that the solution of them is very difficult, not to say impossible." Becquerel, on the other hand, considers it certain that in tropical climates the destruction of the forests is accompanied with an elevation of the mean temperature, and he thinks it highly probable that it has the same effect in the temperate zones. The following is the substance of his remarks on this subject: "Forests act as frigorific causes in three ways:

"1. They shelter the ground against solar irradiation and maintain a greater humidity.

"2. They produce a cutaneous transpiration by the leaves.

"3. They multiply, by the expansion of their branches, the surfaces which are cooled by radiation.

"These three causes acting with greater or less force, we must, in the study of the climatology of a country, take into account the proportion between the area of the forests and the surface which is bared of trees and covered with herbs and grasses.

"We should be inclined to believe, a priori, according to the foregoing considerations, that the clearing of the woods, by raising the temperature and increasing the dryness of the air, ought to react on climate. There is no doubt that, if the vast desert of the Sahara were to become wooded in the course of ages, the sands would cease to be heated as much as at the present epoch, when the mean temperature is twenty-nine degrees [Centigrade, = 85 degrees Fahr.]. In that case, the ascending currents of warm air would cease, or be less warm, and would not contribute, by descending in our latitudes, to soften the climate of Western Europe. Thus the clearing of a great country may react on the climates of regions more or less remote from it.

"The observations by Boussingault leave no doubt on this point. This writer determined the mean temperature of wooded and of cleared points, under the same latitude, and at the same elevation above the sea, in localities comprised between the eleventh degree of north and the fifth degree of south latitude, that is to say, in the portion of the tropics nearest to the equator, and where radiation tends powerfully during the night to lower the temperature under a sky without clouds." [Footnote: Becquerel, Des Climats, etc., pp. 139-141.]

The result of these observations, which has been pretty generally adopted by physicists, is that the mean temperature of cleared land in the tropics appears to be about one degree Centigrade, or a little less than two degrees of Fahrenheit, above that of the forest. On page 147 of the volume just cited, Becquerel argues that, inasmuch as the same and sometimes a greater difference is found in favor of the open ground, at points within the tropics so elevated as to have a temperate or even a polar climate, we must conclude that theforests in Northern America exert a refrigerating influence equally powerful. But the conditions of the soil are so different in the two regions compared, that I think we cannot, with entire confidence, reason from the one to the other, and it is much to be desired that observations be made on the summer and winter temperature of both the air and the ground in the depths of the North American forests, before it is too late.

Recent inquiries have introduced a new element into the problem of the influence of the forest on temperature, or rather into the question of the thermometrical effects of its destruction. I refer to the composition of the soil in respect to its hygroscopicity or aptitude to absorb humidity, whether in a liquid or a gaseous form, and to the conducting power of the particles of which it is composed. [Footnote: Composition, texture, and color of soil are important elements to be considered in estimating the effects of the removal of the forest upon its thermoscopic action. "Experience has proved," says Becquerel, "that when the soil is bared, it becomes more or less heated [by the rays of the sun] according to the nature and the color of the particles which compose it, and according to its humidity, and that, in the refrigeration resulting from radiation, we must take into the account the conducting power of those particles also. Other things being equal, siliceous and calcareous sands, compared in equal volumes with different argillaceous earths, with calcareous powder or dust, with humus, with arable and with garden earth, are the soils which least conduct heat. It is for this reason that sandy ground, in summer, maintains a high temperature even during the night. We may hence conclude that when a sandy soil is stripped of wood, the local temperature will be raised. After the sands follow successively argillaceous, arable, and garden ground, then humus, which occupies the lowest rank.

"The retentive power of humus is but half as great as that of calcareous sand. We will add that the power or retaining heat is proportional to the density. It has also a relation to the magnitude of the particles. It is for this reason that ground covered with siliceous pebbles cools more slowly than siliceous sand, and that pebbly soils are best suited to the cultivation of the vine, because they advance the ripening of the grape more rapidly than chalky and clayey earths, which cool quickly. Hence we see that in examining the calorific effects of clearing forests, it is important to take into account the properties of the soil laid bare."—Becquerel, Des Climats et des Sols boises, p. 137.]

The hygroscopicity of humus or vegetable earth is much greater than that of any mineral soil, and consequently forest ground, where humus abounds, absorbs the moisture of the atmosphere more rapidly and in larger proportion than common earth. The condensation of vapor by absorption develops heat, and consequently elevates the temperature of the soil which absorbs it, together with that of air in contact with the surface. Von Babo found the temperature of sandy ground thus raised from 68 degrees to 80 degrees F., that of soil rich in humus from 68 degrees to 88 degrees. The question of the influence of the woods on temperature does not, in the present state of our knowledge, admit of precise solution, and, unhappily, the primitive forests are disappearing so rapidly before the axe of the woodman, that we shall never be able to estimate with accuracy the climatological action of the natural wood, though all the physical functions of artificial plantations will, doubtless, one day be approximately known.

But the value of trees as a mechanical screen to the soil they cover, and often to ground far to the leeward of them, is most abundantly established, and this agency alone is important enough to justify extensive plantation in all countries which do not enjoy this indispensable protection.



Influence of Forests as Inorganic on the Humidity of the Air and the Earth.

The most important hygroscopic as well as thermoscopic influence of the forest is, no doubt, that which it exercises on the humidity of the air and the earth, and this climatic action it exerts partly as dead, partly as living matter. By its interposition as a curtain between the sky and the ground it both checks evaporation from the earth, and mechanically intercepts a certain proportion of the dew and the lighter showers, which would otherwise moisten the surface of the soil, and restores it to the atmosphere by exhalation; [Footnote: Mangotti had observed and described, in his usual picturesque way, the retention of rain-water by the foliage and bark of trees, but I do not know that any attempts were made to measure the quantity thus intercepted before the experiments of Becquerel, communicated to the Academy of Sciences in 1866. These experiments embraced three series of observations continued respectively for periods of a year, a month, and two days. According to Becquerel's measurements, the quantity falling on bare and on wooded soil respectively was as 1 to 0.07; 1 to 0.5; and 1 to 0.6, or, in other words, he found that only from five-tenths to sixty-seven hundredths of the precipitation reached the ground.—Comptes Rendus de l'Academie des Sciences, 1866. It seemed, indeed, improbable that in rain-storms which last not hours but whole days in succession, so large a proportion of the downfall should continue to be intercepted by forest vegetation after the leaves, the bark, and the whole framework of the trees were thoroughly wet, but the conclusions of this eminent physicist appear to have been generally accepted until the very careful experiments of Mathieu at the Forest-School of Nancy were made known. The observations of Mathieu were made in a plantation of deciduous trees forty-two years old, and were continued through the entire years 1866, 1867, and 1868. The result was that the precipitation in the wood was to that in an open glade of several acres near the forest station as 043 to 1,000, and the proportion in each of the three years was nearly identical. According to Mathieu, then, only 57 thousandths or 5.7 per cent of the precipitation is intercepted by trees.—Surrell, Etude sur les Torrents, 2d ed., ii., p. 98.

By order of the Direction of the Forests of the Canton of Berne, a series of experiments on this subject was commenced at the beginning of the year 1869. During the first seven months of the year (the reports for which alone I have seen), including, of course, the season when the foliage is most abundant, as well as that when it is thinnest, the pluviometers in the woods received only fifteen per cent less than those in the open grounds in the vicinity.—Risler, in Revue des Eaux et Forets, of 10th January, 1870.] while in heavier rains, the large drops which fall upon the leaves and branches are broken into smaller ones, and consequently strike the ground with less mechanical force, or are perhaps even dispersed into vapor without reaching it. [Footnote: We are not, indeed, to suppose that the condensation of vapor and the evaporation of water are going on in the same stratum of air at the same time, or, in other words, that vapor is condensed into rain-drops, and rain-drops evaporated, under the same conditions; but rain formed in one stratum may fall through another, where vapor would not be condensed. Two saturated strata of different temperatures may be brought into contact in the higher regions, and discharge large rain-drops, which, it not divided by some obstruction, will reach the ground, though passing through strata which would vaporize them if they were in a state of more minute division.]

The vegetable mould, resulting from the decomposition of leaves and of wood, serves as a perpetual mulch to forest-soil by carpeting the ground with a spongy covering which obstructs the evaporation from the mineral earth below, [Footnote: The only direct experiments known to me on the evaporation from the SURFACE of the forest are those of Mathieu.—Surrell, Etude sur les Torrents, 2d ed., ii, p. 99.

These experiments were continued from March to December, inclusive, of the year 1868. It was found that during those months the evaporation from a recipient placed on the ground in a plantation of deciduous trees sixty-two years old, was less than one-fifth of that from a recipient of like form and dimensions placed in the open country.] drinks up the rains and melting snows that would otherwise flow rapidly over the surface and perhaps be conveyed to the distant sea, and then slowly gives out, by evaporation, infiltration, and percolation, the moisture thus imbibed. The roots, too, penetrate far below the superficial soil, conduct water along their surface to the lower depths to which they reach, and thus by partially draining the superior strata, remove a certain quantity of moisture out of the reach of evaporation. The Forest as Organic.

These are the principal modes in which the humidity of the atmosphere is affected by the forest regarded as lifeless matter. Let us inquire how its organic processes act upon this meteorological element. The commonest observation shows that the wood and bark of living trees are always more or less pervaded with watery and other fluids, one of which, the sap, is very abundant in trees of deciduous foliage when the buds begin to swell and the leaves to develop themselves in the spring. This fluid is drawn principally, if not entirely, from the ground by the absorbent action of the roots, for though Schacht and some other eminent botanical physiologists have maintained that water is absorbed by the leaves and bark of trees, yet most experiments lead to the contrary result, and it is now generally held that no water is taken in by the pores of vegetables. Late observations by Cailletet, in France, however, tend to the establishment of a new doctrine on this subject which solves many difficulties and will probably be accepted by botanists as definitive. Cailletet finds that under normal conditions, that is, when the soil is humid enough to supply sufficient moisture through the roots, no water is absorbed by the leaves, buds, or bark of plants, but when the roots are unable to draw from the earth the requisite quantity of this fluid, the vegetable pores in contact with the atmosphere absorb it from that source.

Popular opinion, indeed, supposes that all the vegetable fluids, during the entire period of growth, are drawn from the bosom of the earth, and that the wood and other products of the tree are wholly formed from matter held in solution in the water abstracted by the roots from the ground. This is an error, for the solid matter of the tree, in a certain proportion not important to our present inquiry, is received from the atmosphere in a gaseous form, through the pores of the leaves and of the young shoots, and, as we have just seen, moisture is sometimes supplied to trees by the atmosphere. The amount of water taken up by the roots, however, is vastly greater than that imbibed through the leaves and bark, especially at the season when the sap is most abundant, and when the leaves are yet in embryo. The quantity of water thus received from the air and the earth, in a single year, even by a wood of only a hundred acres, is very great, though experiments are wanting to furnish the data for even an approximate estimate of its measure; for only the vaguest conclusions can be drawn from the observations which have been made on the imbibition and exhalation of water by trees and other plants reared in artificial conditions diverse from those of the natural forest. [Footnote: The experiments of Hales and others on the absorption and exhalation of vegetables are of high physiological interest; but observations on sunflowers, cabbages, hops, and single branches of isolated trees, growing in artificially prepared soils and under artificial conditions, furnish no trustworthy data for computing the quantity of water received and given off by the natural wood.]

Flow of Sap.

The amount of sap which can be withdrawn from living trees furnishes, not indeed a measure of the quantity of water sucked up by their roots from the ground—for we cannot extract from a tree its whole moisture—but numerical data which may aid the imagination to form a general notion of the powerful action of the forest as an absorbent of humidity from the earth.

The only forest-tree known to Europe and North America, the sap of which is largely enough applied to economical uses to have made the amount of its flow a matter of practical importance and popular observation, is the sugar maple, Acer saccharinum, of the Anglo-American Provinces and States. In the course of a single "sugar season," which lasts ordinarily from twenty-five to thirty days, a sugar maple two feet in diameter will yield not less than twenty gallons of sap, and sometimes much more. [Footnote: Emerson (Trees of Massachusetts. p. 403) mentions a maple six feet in diameter, as having yielded a barrel, or thirty-one and a half gallons, of sap in twenty-four hours, and another, the dimensions of which are not stated, as having yielded one hundred and seventy-five gallons in the course of the season.

The Cultivator, an American agricultural journal, for June, 1842, states that twenty gallons of sap were drawn in eighteen hours from a single maple, two and a half feet in diameter, in the town of Warner, New Hampshire, and the truth of this account has been verified by personal inquiry made in my behalf. This tree was of the original forest growth, and had been left standing when the ground around it was cleared. It was tapped only every other year, and then with six or eight incisions. Dr. Williams (History of Vermont, i., p. 01) says: "A man much employed in milking maple sugar, found that, for twenty-one days together, a maple-tree discharged seven and a half gallons per day."

An intelligent correspondent, of much experience in the manufacture of maple sugar, writes me that a second-growth maple, of about two feet in diameter, standing in open ground, tapped with four incisions, has, for several seasons, generally run eight gallons per day in fair weather. He speaks of a very large tree, from which sixty gallons were drawn in the course of a season, and of another, something more than three feet through, which made forty-two pounds of wet sugar, and must have yielded not less than one hundred and fifty gallons.] This, however, is but a trifling proportion of the water abstracted from the earth by the roots during this season; for all this fluid runs from two or three incisions or auger-holes, so narrow as to intercept the current of comparatively few sap vessels, and besides, experience shows that large as is the quantity withdrawn from the circulation, it is relatively too small to affect very sensibly the growth of the tree. [Footnote: Tapping does not check the growth, but does injure the quality of the wood of maples. The wood of trees often tapped is lighter and less dense than that of trees which have not been tapped, and gives less heat in burning. No difference has been observed in the bursting of the buds of tapped and untapped trees.] The number of large maple-trees on an acre is frequently not less than fifty, [Footnote: Dr. Rush, in a letter to Jefferson, states the number of maples fit for tapping on an acre at from thirty to fifty. "This," observes my correspondent, "is correct with regard to the original growth, which is always more or less intermixed with other trees; but in second growth, composed of maples alone, the number greatly exceeds this. I have had the maples on a quarter of an acre, which I thought about an average of second-growth 'maple orchards,' counted. The number was found to be fifty-two, of which thirty-two were ten inches or more in diameter, and, of course, large enough to tap. This gives two hundred and eight trees to the acre, one hundred and twenty-eight of which were of proper size for tapping."] and of course the quantity of moisture abstracted from the soil by this tree alone is measured by thousands of gallons to the acre. The sugar orchards, as they are called, contain also many young maples too small for tapping, and numerous other trees—two of which, at least, the black birch, Betula lenta, and yellow birch, Betula excelsa, both very common in the same climate, are far more abundant in sap than the maple [Footnote: The correspondent already referred to informs me that a black birch, tapped about noon with two incisions, was found the next morning to have yielded sixteen gallons. Dr. Williams (History of Vermont, i., p. 91) says: "A large birch, tapped in the spring, ran at the rate of five gallons an hour when first tapped. Eight or nine days after, it was found to run at the rate of about two and a half gallons an hour, and at the end of fifteen days the discharge continued in nearly the same quantity. The sap continued to flow for four or five weeks, and it was the opinion of the observers that it must have yielded as much as sixty barrels [l,800 gallons]."]—are scattered among the sugar-trees; for the North American native forests are remarkable for the mixture of their crops. The sap of the maple, and of other trees with deciduous leaves which grow in the same climate, flows most freely in the early spring, and especially in clear weather, when the nights are frosty and the days warm; for it is then that the melting snows supply the earth with moisture in the justest proportion, and that the absorbent power of the roots is stimulated to its highest activity.

When the buds are ready to burst, and the green leaves begin to show themselves beneath their scaly covering, the ground has become drier, the absorption by the roots is diminished, and the sap, being immediately employed in the formation of the foliage, can be extracted from the stem in only small quantities.

Absorption and Exhalation by Foliage.

The leaves now commence the process of absorption, and imbibe both uncombined gases and an unascertained but probably inconsiderable quantity of aqueous vapor from the humid atmosphere of spring which bathes them.

The organic action of the tree, as thus far described, tends to the desiccation of air and earth; but when we consider what volumes of water are daily absorbed by a large tree, and how small a proportion of the weight of this fluid consists of matter which, at the period when the flow of sap is freest, enters into new combinations, and becomes a part of the solid framework of the vegetable, or a component of its deciduous products, it becomes evident that the superfluous moisture must somehow be carried back again almost as rapidly as it flows into the tree. At the very commencement of vegetation in spring, some of this fluid certainly escapes through the buds, the nascent foliage, and the pores of the bark, and vegetable physiology tells us that there is a current of sap towards the roots as well as from them. [Footnote: "The elaborated sap, passing out of the leaves, is received into the inner bark, . . . and a part of what descends finds its way even to the ends of the roots, and is all along diffused laterally into the stem, where it meets and mingles with the ascending crude sap or raw material. So there is no separate circulation of the two kinds of sap; and no crude sap exists separately in any part of the plant. Even in the root, where it enters, this mingles at once with some elaborated sap already there."—Gray, How Plants Grow, Section 273.]

I do not know that the exudation of water into the earth, through the bark or at the extremities of these latter organs, has been proved, but the other known modes of carrying off the surplus do not seem adequate to dispose of it at the almost leafless period when it is most abundantly received, and it is possible that the roots may, to some extent, drain as well as flood the water-courses of their stem. Later in the season the roots absorb less, and the now developed leaves exhale an increased quantity of moisture into the air. In any event, all the water derived by the growing tree from the atmosphere and the ground is parted with by transpiration or exudation, after having surrendered to the plant the small proportion of matter required for vegetable growth which it held in solution or suspension. [Footnote: Ward's tight glazed cases for raising and especially for transporting plants, go far to prove that water only circulates through vegetables, and is again and again absorbed and transpired by organs appropriated to these functions.

Seeds, growing grasses, shrubs, or trees planted in proper earth, moderately watered and covered with a glass bell or close frame of glass, live for months, and even years, with only the original store of air and water. In one of Ward's early experiments, a spire of grass and a fern, which sprang up in a corked bottle containing a little moist earth introduced as a bed for a snail, lived and flourished for eighteen years without a new supply of either fluid. In these boxes the plants grow till the enclosed air is exhausted of the gaseous constituents of vegetation, and till the water has yielded up the assimilable matter it held in solution, and dissolved and supplied to the roots the nutriment contained in the earth in which they are planted. After this, they continue for a long time in a state of vegetable sleep, but if fresh air and water be introduced into the cases, or the plants be transplanted into open ground, they rouse themselves to renewed life, and grow vigorously, without appearing to have suffered from their long imprisonment. The water transpired by the leaves is partly absorbed by the earth directly from the air, partly condensed on the glass, along which it trickles down to the earth, enters the roots again, and thus continually repeats the circuit. See Aus der Natur, 21, B. S. 537.] The hygrometrical equilibrium is then restored, so far as this: the tree yields up again the moisture it had drawn from the earth and the air, though it does not return it each to each; for the vapor carried off by transpiration greatly exceeds the quantity of water absorbed by the foliage from the atmosphere, and the amount, if any, carried back to the ground by the roots.

The present estimates of some eminent vegetable physiologists in regard to the quantity of aqueous vapor exhaled by trees and taken up by the atmosphere are much greater than those of former inquirers. Direct and satisfactory experiments on this point are wanting, and it is not easy to imagine how they could be made on a sufficiently extensive and comprehensive scale. Our conclusions must therefore be drawn from observations on small plants, or separate branches of trees, and of course are subject to much uncertainty. Nevertheless, Schleiden, arguing from such analogies, comes to the surprising result, that a wood evaporates ten times as much water as it receives from atmospheric precipitation. [Footnote: Fur Baum und Wald, pp. 46, 47, notes. Pfaff, too, experimenting on branches of a living oak, weighed immediately after being cut from the tree, and again after an exposure to the air for three minutes, and computing the superficial measure of all the leaves of the tree, concludes that an oak-tree evaporates, during the season of growth, eight and a half times the mean amount of rain-fall on an area equal to that shaded by the tree.] In the Northern and Eastern States of the Union, the mean precipitation during the period of forest growth, that is from the swelling of the buds in the spring to the ripening of the fruit, the hardening of the young shoots, and the full perfection of the other annual products of the tree, exceeds on the average twenty-four inches. Taking this estimate, the evaporation from the forest would be equal to a precipitation of two hundred and forty inches, or very nearly one hundred and fifty standard gallons to the square foot of surface.

The first questions which suggest themselves upon this statement are: what becomes of this immense quantity of water and from what source does the tree derive it We are told in reply that it is absorbed from the air by the humus and mineral soil of the wood, and supplied again to the tree through its roots, by a circulation analogous to that observed in Ward's air-tight cases. When we recall the effect produced on the soil even of a thick wood by a rain-fall of one inch, we find it hard to believe that two hundred and forty times that quantity, received by the ground between early spring and autumn, would not keep it in a state of perpetual saturation, and speedily convert the forest into a bog.

No such power of absorption of moisture by the earth from the atmosphere, or anything approaching it, has ever been shown by experiment, and all scientific observation contradicts the supposition. Schubler found that in seventy-two hours thoroughly dried humus, which is capable of taking up twice its own weight of water in the liquid state, absorbed from the atmosphere only twelve per cent. of its weight of humidity; garden-earth five and one-fifth per cent. and ordinary cultivated soil two and one-third per cent. After seventy-two hours, and, in most of his experiments with thirteen different earths, after forty-eight hours, no further absorption took place. Wilhelm, experimenting with air-dried field-earth, exposed to air in contact with water and protected by a bell-glass, found that the absorption amounted in seventy-two hours to two per cent. and a very small fraction, nearly the whole of which was taken up in the first forty-eight hours. In other experiments with carefully heat-dried field-soil, the absorption was five per cent. in eighty-four hours, and when the water was first warmed to secure the complete saturation of the air, air-dried garden-earth absorbed five and one-tenth per cent. in seventy-two hours.

In nature, the conditions are never so favorable to the absorption of vapor as in those experiments. The ground is more compact and of course offers less surface to the air, and, especially in the wood, it is already in a state approaching saturation. Hence, both these physicists conclude that the quantity of aqueous vapor absorbed by the earth from the air is so inconsiderable "that we can ascribe to it no important influence on vegetation." [Footnote: Wilhelm, Der Boden und das Wasser, pp. 14,20.] Besides this, trees often grow luxuriantly on narrow ridges, on steep declivities, on partially decayed stumps many feet above the ground, on walls of high buildings, and on rocks, in situations where the earth within reach of their roots could not possibly contain the tenth part of the water which, according to Schleiden and Pfaff, they evaporate in a day. There are, too, forests of great extent on high bluffs and well-drained table-lands, where there can exist, neither in the subsoil nor in infiltration from neighboring regions, an adequate source of supply for such consumption. It must be remembered, also, that in the wood the leaves of the trees shade each other, and only the highest stratum of foliage receives the full influence of heat and light; and besides, the air in the forest is almost stagnant, while in the experiments of Unger, Marshal, Vaillant, Pfaff and others, the branches were freely exposed to light, sun, and atmospheric currents. Such observations can authorize no conclusions respecting the quantitative action of leaves of forest trees in normal conditions.

Further, allowing two hundred days for the period of forest vital action, the wood must, according to Schleiden's position, exhale a quantity of moisture equal to an inch and one-fifth of precipitation per day, and it is hardly conceivable that so large a volume of aqueous vapor, in addition to the supply from other sources, could be diffused through the ambient atmosphere without manifesting its presence by ordinary hygrometrical tests much more energetically than it has been proved to do, and in fact, the observations recorded by Ebermayer show that though the RELATIVE humidity of the atmosphere is considerably greater in the cooler temperature of the wood, its ABSOLUTE humidity does not sensibly differ from that of the air in open ground. [Footnote: Ebermeyer, Die Physikalischen, Einwirkungen des Waldes, i., pp. 150 et seqq. It may be well here to guard my readers against the common error which supposes that a humid condition of the AIR is necessarily indicated by the presence of fog or visible vapor. The air is rendered humid by containing INVISIBLE vapor, and it becomes drier by the condensation of such vapor into fog, composed of solid globules or of hollow vesicles of water—for it is a disputed point whether the particles of fog are solid or vesicular. Hence, though the ambient atmosphere may hold in suspension, in the form of fog, water enough to obscure its transparency, and to produce the sensation of moisture on the skin, the air, in which the finely divided water floats, may be charged with even less than an average proportion of humidity.]

The daily discharge of a quantity of aqueous vapor corresponding to a rain-fall of one inch and a fifth into the cool air of the forest would produce a perpetual shower, or at least drizzle, unless, indeed, we suppose a rapidity of absorption and condensation by the ground, and of transmission through the soil to the roots and through them and the vessels of the tree to the leaves, much greater than has been shown by direct observation. Notwithstanding the high authority of Schleiden, therefore, it seems impossible to reconcile his estimates with facts commonly observed and well established by competent investigators. Hence the important question of the supply, demand, and expenditure of water by forest vegetation must remain undecided, until it can be determined by something approaching to satisfactory direct experiment. [Footnote: According to Cezanne, Surrell, Etude sur les Torrents, 2e edition, ii., p. 100, experiments reported in the Revue des Eaux et Forets for August, 1868, showed the evaporation from a living tree to be "almost insignificant." Details are not given.]

Balance of Conflicting Influences of Forest on Atmospheric Heat and Humidity.

We have shown that the forest, considered as dead matter, tends to diminish the moisture of the air, by preventing the sun's rays from reaching the ground and evaporating the water that falls upon the surface, and also by spreading over the earth a spongy mantle which sucks up and retains the humidity it receives from the atmosphere, while, at the same time, this covering acts in the contrary direction by accumulating, in a reservoir not wholly inaccessible to vaporizing influences, the water of precipitation which might otherwise suddenly sink deep into the bowels of the earth, or flow by superficial channels to other climatic regions. We now see that, as a living organism, it tends, on the one hand, to diminish the humidity of the air by sometimes absorbing moisture from it, and, on the other, to increase that humidity by pouring out into the atmosphere, in a vaporous form, the water it draws up through its roots. This last operation, at the same time, lowers the temperature of the air in contact with or proximity to the wood, by the same law as in other cases of the conversion of water into vapor.

As I have repeatedly said, we cannot measure the value of any one of those elements of climatic disturbance, raising or lowering of temperature, increase or diminution of humidity, nor can we say that in any one season, any one year, or any one fixed cycle, however long or short, they balance and compensate each other. They are sometimes, but certainly not always, contemporaneous in their action, whether their tendency is in the same or in opposite directions, and, therefore, their influence is sometimes cumulative, sometimes conflicting; but, upon the whole, their general effect is to mitigate extremes of atmospheric heat and cold, moisture and drought. They serve as equalizers of temperature and humidity, and it is highly probable that, in analogy with most other works and workings of nature, they, at certain or uncertain periods, restore the equilibrium which, whether as lifeless masses or as living organisms, they may have temporarily disturbed. [Footnote: There is one fact which I have nowhere seen noticed, but which seems to me to have an important bearing on the question whether forests tend to maintain an equilibrium between the various causes of hygroscopic action, and consequently to keep the air within their precincts in an approximately constant condition, so far as this meteorological element is concerned. I refer to the absence of fog or visible vapor in thick woods in full leaf, even when the air of the neighboring open grounds is so heavily charged with condensed vapor as completely to obscure the sun. The temperature of the atmosphere in the forest is not subject to so sudden and extreme variations as that of cleared ground, but at the same time it is far from constant, and so large a supply of vapor as is poured out by the foliage of the trees could not fail to be sometimes condensed into fog by the same causes as in the case of the adjacent meadows, which are often covered with a dense mist while the forest-air remains clear, were there not some potent counteracting influence always in action. This influence, I believe, is to be found partly in the equalization of the temperature of the forest, and partly in the balance between the humidity exhaled by the trees and that absorbed and condensed invisibly by the earth.] When, therefore, man destroys these natural harmonizors of climatic discords, he sacrifices an important conservative power, though it is far from certain that he has thereby affected the mean, however much he may have exaggerated the extremes of atmospheric temperature and humidity, or, in other words, may have increased the range and lengthened the scale of thermometric and hygrometric variation.

Special Influence of Woods on Precipitation.

With the question of the action of forests upon temperature and upon atmospheric humidity is intimately connected that of their influence upon precipitation, which they may affect by increasing or diminishing the warmth of the air and by absorbing or exhaling uncombincd gas and aqueous vapor. The forest being a natural arrangement, the presumption is that it exercises a conservative action, or at least a compensating one, and consequently that its destruction must tend to produce pluviometrical disturbances as well as thermometrical variations. And this is the opinion of perhaps the greatest number of observers. Indeed, it is almost impossible to suppose that, under certain conditions of time and place, the quantity and the periods of rain should not depend, more or less, upon the presence or absence of forests; and without insisting that the removal of the forest has diminished the sum-total of snow and rain, we may well admit that it has lessened the quantity which annually falls within particular limits. Various theoretical considerations make this probable, the most obvious argument, perhaps, being that drawn from the generally admitted fact, that the summer and even the mean temperature of the forest is below that of the open country in the same latitude. If the air in a wood is cooler than that around it, it must reduce the temperature of the atmospheric stratum immediately above it, and, of course, whenever a saturated current sweeps over it, it must produce precipitation which would fall upon it, or at a greater or less distance from it.

We must here take into the account a very important consideration. It is not universally or even generally true, that the atmosphere returns its condensed humidity to the local source from which it receives it. The air is constantly in motion,

—howling tempests scour amain From sea to land, from land to sea;

[Footnote: Und Sturme brausen um die Wette Vom Meer aufs Land, vom Land aufs Meer. Goethe, Faust, Song of the Archangels.]

and, therefore, it is always probable that the evaporation drawn up by the atmosphere from a given river, or sea, or forest, or meadow, will be discharged by precipitation, not at or near the point where it rose, but at a distance of miles, leagues, or even degrees. The currents of the upper air are invisible, and they leave behind them no landmark to record their track. We know not whence they come, or whither they go. We have a certain rapidly increasing acquaintance with the laws of general atmospheric motion, but of the origin and limits, the beginning and end of that motion, as it manifests itself at any particular time and place, we know nothing. We cannot say where or when the vapor, exhaled to-day from the lake on which we float, will be condensed and fall; whether it will waste itself on a barren desert, refresh upland pastures, descend in snow on Alpine heights, or contribute to swell a distant torrent which shall lay waste square miles of fertile corn-land; nor do we know whether the rain which feeds our brooklets is due to the transpiration from a neighboring forest, or to the evaporation from a far-off sea. If, therefore, it were proved that the annual quantity of rain and dew is now as great on the plains of Castile, for example, as it was when they were covered with the native forest, it would by no means follow that those woods did not augment the amount of precipitation elsewhere. The whole problem of the pluviometrical influence of the forest, general or local, is so exceedingly complex and difficult that it cannot, with our present means of knowledge, be decided upon a priori grounds. It must now be regarded as a question of fact which would probably admit of scientific explanation if it were once established what the actual fact is.

Unfortunately, the evidence is conflicting in tendency, and sometimes equivocal in interpretation, but I believe that a majority of the foresters and physicists who have studied the question are of opinion that in many, if not in all cases, the destruction of the woods has been followed by a diminution in the annual quantity of rain and dew. Indeed, it has long been a popularly settled belief that vegetation and the condensation and fall of atmospheric moisture are reciprocally necessary to each other, and even the poets sing of

Afric's barren sand, Where nought can grow, because it raineth not, And where no rain can fall to bless the land, Because nought grows there.

[Footnote: Det golde Strog i Afrika, Der Intet voxe kan, da ei det regner, Og, omvendt, ingen Regn kan falde, da Der Intet voxer. Paudan-Muller, Adam Hamo, ii., 408.]

Before going further with the discussion, however, it is well to remark that the comparative rarity or frequency of inundations in earlier or later centuries is not necessarily, in most cases not probably, entitled to any weight whatever, as a proof that more or less rain fell formerly than now; because the accumulation of water in the channel of a river depends far less upon the quantity of precipitation in its valley, than upon the rapidity with which it is conducted, on or under the surface of the ground, to the central artery that drains the basin. But this point will be more fully discussed in a subsequent chapter.

In writers on the subject we are discussing, we find many positive assertions about the diminution of rain in countries which have been stripped of wood within the historic period, but these assertions very rarely rest upon any other proof than the doubtful recollection of unscientific observers, and I am unable to refer to a single instance where the records of the rain-gauge, for a considerable period before and after the felling or planting of extensive woods, can be appealed to in support of either side of the question. The scientific reputation of many writers who have maintained that precipitation has been diminished in particular localities by the destruction of forests, or augmented by planting them, has led the public to suppose that their assertions rested on sufficient proof. We cannot affirm that in none of these cases did such proof exist, but I am not aware that it has ever been produced. [Footnote: Among recent writers, Clave, Schacht, Sir John F. W. Herschel, Hohenstein, Barth, Asbjornsen, Boussingault, and others, maintain that forests tend to produce rain and clearings to diminish it, and they refer to numerous facts of observation in support of this doctrine; but in none of these does it appear that these observations are supported by actual pluviometrical measure. So far as I know, the earliest expression of the opinion that forests promote precipitation is that attributed to Christopher Columbus, in the Historie del S. D. Fernando Colombo, Venetia, 157l, cap. lviii., where it is said that the Admiral ascribed the daily showers which fell in the West Indies about vespers to "the great forests and trees of those countries," and remarked that the same effect was formerly produced by the same cause in the Canary and Madeira Islands and in the Azores, but that "now that the many woods and trees that covered them have been felled, there are not produced so many clouds and rains as before."

Mr. H. Harrisse, in his very learned and able critical essay, Fernand Colomb, sa Vie et ses Oeuvres, Paris, 1872, has made it at least extremely probable that the Historie is a spurious work. The compiler may have found this observation in some of the writings of Columbus now lost, but however that may be, the fact, which Humboldt mentions in Cosmos with much interest, still remains, that the doctrine in question was held, if not by the great discoverer himself, at least by one of his pretended biographers, as early as the year 1571.]

The effect of the forest on precipitation, then, is by no means free from doubt, and we cannot positively affirm that the total annual quantity of rain is even locally diminished or increased by the destruction of the woods, though both theoretical considerations and the balance of testimony strongly favor the opinion that more rain falls in wooded than in open countries. One important conclusion, at least, upon the meteorological influence of forests is certain and undisputed: the proposition, namely, that, within their own limits, and near their own borders, they maintain a more uniform degree of humidity in the atmosphere than is observed in cleared grounds. Scarcely less can it be questioned that they tend to promote the frequency of showers, and, if they do not augment the amount of precipitation, they probably equalize its distribution through the different seasons. [Footnote: The strongest direct evidence which I am able to refer to in support of the proposition that the woods produce even a local augmentation of precipitation is furnished by the observations of Mathieu, sub-director of the Forest-School at Nancy. His pluviometrical measurements, continued for three years, 1866-1868, show that during that period the annual mean of rain-fall in the centre of the wooded district of Cinq-Tranchees, at Belle Fontaine on the borders of the forest, and at Amance, in an open cultivated territory in the same vicinity, was respectively as the numbers 1,000, 957, and 853.

The alleged augmentation of rain-fall in Lower Egypt, in consequence of large plantations by Mehemet Ali, is very frequently appealed to as a proof of this influence of the forest, and this case has become a regular common-place in all discussions of the question. It is, however, open to the same objection as the alleged instances of the diminution of precipitation in consequence of the felling of the forest.

This supposed increase in the frequency and quantity of rain in Lower Egypt is, I think, an error, or at least not an established fact. I have heard it disputed on the spot by intelligent Franks, whose residence in that country began before the plantations of Mehemet Ali and Ibrahim Pacha, and I have been assured by them that meterological observations, made at Alexandria about the begiuning of this century, show an annual fall of rain as great as is usual at this day. The mere fact that it did not rain during the French occupation is not conclusive. Having experienced a gentle shower of nearly twenty-four hours' duration in Upper Egypt, I inquired of the local governor in relation to the frequency of this phenomenon, and was told by him that not of drop of rain had fallen at that point for more than two years previous.

The belief in the increase of rain in Egypt rests almost entirely on the observations of Marshal Marmont, and the evidence collected by him in 1836. His conclusions have been disputed, if not confuted, by Joinard and others, and are probably erroneous. See Foissac, Meteorologie, German translation, pp. 634-639.

It certainly sometimes rains briskly at Cairo, but evaporation is exceedingly rapid in Egypt—as any one who ever saw a Fellah woman wash a napkin in the Nile, and dry it by shaking it a few moments in the air, can testify; and a heap of grain, wet a few inches below the surface, would probably dry again without injury. At any rate, the Egyptian Government often has vast quantities of wheat stored at Boulak in uncovered yards through the winter, though it must be admitted that the slovenliness and want of foresight in Oriental life, public and private, are such that we cannot infer the safety of any practice followed in the East merely from its long continuance.

Grain, however, may be long kept in the open air in climates much less dry than that of Egypt, without injury, except to the superficial layers; for moisture does not penetrate to a great depth in a heap of grain once well dried and kept well aired. When Louis IX. was making his preparations for his campaign in the East, he had large quantities of wine and grain purchased in the Island of Cyprus, and stored up for two years to await his arrival. "When we were come to Cyprus," says Joinville, Histoire de Saint Louis, Section 72, 73, "we found there greate foison of the Kynge's purveyance. . . The wheate and the barley they had piled up in greate heapes in the feeldes, and to looke vpon, they were like vnto mountaynes; for the raine, the whyche hadde beaten vpon the wheate now a longe whyle, had made it to sproute on the toppe, so that it seemed as greene grasse. And whanne they were mynded to carrie it to Egypte, they brake that sod of greene herbe, and dyd finde under the same the wheate and the barley, as freshe as yf menne hadde but nowe thrashed it."]

Total Climatic Influence of the Forest.

Aside from the question of local disturbances and their compensations, it does not seem probable that the forests sensibly affect the general mean of atmospheric temperature of the globe, or the total quantity of precipitation, or even that they had this influence when their extent was vastly greater than at present. The waters cover about three-fourths of the face of the earth, and if we deduct the frozen zones, the peaks and crests of lofty mountains and their craggy slopes, the Sahara and other great African and Asiatic deserts, and all such other portions of the solid surface as are permanently unfit for the growth of wood, we shall find that probably not one-tenth of the total superficies of our planet was ever, at any one time in the present geological period, covered with forests. Besides this, the distribution of forest land, of desert, and of water, is such as to reduce the possible influence of the woods to a low expression; for the forests are, in large proportion, situated in cold or temperate climates, where the action of the sun is comparatively feeble both in elevating temperature and in promoting evaporation; while, in the torrid zone, the desert and the sea—the latter of which always presents an evaporable surface—enormously preponderate. It is, upon the whole, not probable that so small an extent of forest, so situated, could produce a sensible influence on the general climate of the globe, though it might appreciably affect the local action of all climatic elements. The total annual amount of solar heat absorbed and radiated by the earth, and the sum of terrestrial evaporation and atmospheric precipitation, must be supposed constant; but the distribution of heat and of humidity is exposed to disturbance in both time and place by a multitude of local causes, among which the presence or absence of the forest is doubtless one.

So far as we are able to sum up the results, it would appear that, in countries in the temperate zone still chiefly covered with wood, the summers would be cooler, moister, shorter, the winters milder, drier, longer, than in the same regions after the removal of the forest, and that the condensation and precipitation of atmospheric moisture would be, if not greater in total quantity, more frequent and less violent in discharge. The slender historical evidence we possess seems to point to the same conclusion, though there is some conflict of testimony and of opinion on this point.

Among the many causes which, as we have seen, tend to influence the general result, the mechanical action of the forest, if not more important, is certainly more obvious and direct than the immediate effects of its organic processes. The felling of the woods involves the sacrifice of a valuable protection against the violence of chilling winds and the loss of the shelter afforded to the ground by the thick coating of leaves which the forest sheds upon it and by the snow which the woods prevent from blowing away, or from melting in the brief thaws of winter. I have already remarked that bare ground freezes much deeper than that which is covered by beds of leaves, and when the earth is thickly coated with snow, the strata frozen before it fell begin to thaw. It is not uncommon to find the ground in the woods, where the snow lies two or three feet deep, entirely free from frost, when the atmospheric temperature has been for several weeks below the freezing-point, and for some days even below the zero of Fahrenheit. When the ground is cleared and brought under cultivation, the leaves are ploughed into the soil and decomposed, and the snow, especially upon knolls and eminences, is blown off, or perhaps half thawed, several times during the winter. The water from the melting snow runs into the depressions, and when, after a day or two of warm sunshine or tepid rain, the cold returns, it is consolidated to ice, and the bared ridges and swells of earth are deeply frozen. [Footnote: I have seen, in Northern New England, the surface of the open ground frozen to the depth of twenty-two inches, in the month of November, when in the forest-earth no frost was discoverable; and later in the winter, I have known an exposed sand-knoll to remain frozen six feet deep, after the ground in the woods was completely thawed.] It requires many days of mild weather to raise the temperature of soil in this condition, and of the air in contact with it, to that of the earth in the forests of the same climatic region. Flora is already plaiting her sylvan wreath before the corn-flowers which are to deck the garland of Ceres have waked from their winter's sleep; and it is probably not a popular error to believe that, where man has substituted his artificial crops for the spontaneous harvest of nature, spring delays her coming. [Footnote: The conclusion arrived at by Noah Webster, in his very learned and able paper on the supposed change in the temperature of winter, read before the Connecticut Academy of Arts and Sciences in 1799, was as follows: "From a careful comparison of these facts, it appears that the weather, in modern winters, in the United States, is more inconstant than when the earth was covered with woods, at the first settlement of Europeans in the country; that the warm weather of autumn extends further into the winter months, and the cold weather of winter and spring encroaches upon the summer; that, the wind being more variable, snow is less permanent, and perhaps the same remark may be applicable to the ice of the rivers. These effects seem to result necessarily from the greater quantity of heat accumulated in the earth in summer since the ground has been cleared of wood and exposed to the rays of the sun, and to the greater depth of frost in the earth in winter by the exposure of its uncovered surface to the cold atmosphere."—Collection of Papers by Noah Webster, p. 162.]

There are, in the constitution and action of the forest, many forces, organic and inorganic, which unquestionably tend powerfully to produce meteorological effects, and it may, therefore, be assumed as certain that they must and do produce such effects, UNLESS they compensate and balance each other, and herein lies the difficulty of solving the question. To some of these elements late observations give a new importance. For example, the exhalation of aqueous vapor by plants is now believed to be much greater, and the absorption of aqueous vapor by them much less, than was formerly supposed, and Tyndall's views on the relations of vapor to atmospheric heat give immense value to this factor in the problem. In like manner the low temperature of the surface of snow and the comparatively high temperature of its lower strata, and its consequent action on the soil beneath, and the great condensation of moisture by snow, are facts which seem to show that the forest, by protecting great surfaces of snow from melting, must inevitably exercise a great climatic influence. If to these influences we add the mechanical action of the woods in obstructing currents of wind, and diminishing the evaporation and refrigeration which such currents produce, we have an accumulation of forces which MUST manifest great climatic effects, unless—which is not proved and cannot be presumed—they neutralize each other. These are points hitherto little considered in the discussion, and it seems difficult to deny that as a question of ARGUMENT, the probabilities are strongly in favor of the meteorological influence of the woods. The EVIDENCE, indeed, is not satisfactory, or, to speak more accurately, it is non-existent, for there really is next to no trustworthy proof on the subject, but it appears to me a case where the burden of proof must be taken by those who maintain that, as a meteorological agent, the forest is inert.

The question of a change in the climate of the Northern American States is examined in the able Meteorological Report of Mr. Draper, Director of the New York Central Park Observatory, for 1871. The result arrived at by Mr. Draper is, that there is no satisfactory evidence of a diminution in the rainfall, or of any other climatic change in the winter season, in consequence of clearing of the forests or other human action. The proof from meteorological registers is certainly insufficient to establish the fact of a change of climate, but, on the other hand, it is equally insufficient to establish the contrary. Meteorological stations are too few, their observations, in many cases, extend over a very short period, and, for reasons I have already given, the great majority of their records are entitled to little or no confidence. [Footnote: Since these pages were written, the subject of forest meteorology has received the most important contribution ever made to it, in several series of observations at numerous stations in Bavaria, from the year 1866 to 1871, published by Ebermayer, at Aschaffenburg, in 1873, under the title: Die Physikalischen Einwirkungen des Waldes auf Luft und Boden, und seine Klimatologische und Hygienische Bedeutung. I. Band. So far as observations of only five years' duration can prove anything, the following propositions, not to speak of many collateral and subsidiary conclusions, seem to be established, at least for the localities where the observations were made:

1. The yearly mean temperature of wooded soils, at all depths, is lower than that of open grounds, p. 85.

This conclusion, it may be remarked, is of doubtful applicability in regions of excessive climate like the Northern United States and Canada, where the snow keeps the temperature of the soil in the forest above the freezing-point, for a large part and sometimes the whole of the winter, while in unwooded ground the earth remains deeply frozen.

2. The yearly mean atmospheric temperature, other things being equal, is lower in the forest than in cleared grounds, p. 84.

3. Climates become excessive in consequence of extensive clearings, p. 117.

4. The ABSOLUTE humidity of the air in the forest is about the same as in open ground, while the RELATIVE humidity is greater in the former than in the latter case, on account of the lower temperature of the atmosphere in the wood, p. 150.

5. The evaporation from an exposed surface of water in the forest is sixty-four per cent. less than in unwooded grounds, pp. 159,161.

6. About twenty-six per cent. of the precipitation is interrupted and prevented from reaching the ground by the foliage and branches of forest trees, p. 194.

7. In the interior of thick woods, the evaporation from water and from earth is much less than the precipitation, p. 210.

8. The loss of the water of precipitation intercepted by the trees in the forest is compensated by the smaller evaporation from the ground, p. 219.

9. In elevated regions and during the summer half of the year, woods tend to increase the precipitation, p. 202.]

Influence of the Forest on the Humidity of the Soil.

I have hitherto confined myself to the influence of the forest on meteorological conditions, a subject, as has been seen, full of difficulty and uncertainty. Its comparative effects on the temperature, the humidity, the texture and consistence, the configuration and distribution of the mould or arable soil, and, very often, of the mineral strata below, and on the permanence and regularity of springs and greater superficial water-courses, are much less disputable as well as more easily estimated and more important, than its possible value as a cause of strictly climatic equilibrium or disturbance.

The action of the forest on the earth is chiefly mechanical, but the organic process of absorption of moisture by its roots affects the quantity of water contained in the vegetable mould and in the mineral strata near the surface, and, consequently, the consistency of the soil. In treating of the effects of trees on the moisture of the atmosphere, I have said that the forest, by interposing a canopy between the sky and the ground, and by covering the surface with a thick mantle of fallen leaves, at once obstructed insulation and prevented the radiation of heat from the earth. These influences go far to balance each other; but familiar observation shows that, in summer, the forest-soil is not raised to so high a temperature as open grounds exposed to irradiation. For this reason, and in consequence of the mechanical resistance opposed by the bed of dead leaves to the escape of moisture, we should expect that, except after recent rains, the superficial strata of woodland-soil would be more humid than that of cleared land. This agrees with experience. The soil of the natural forest is always moist, except in the extremest droughts, and it is exceedingly rare that a primitive wood suffers from want of humidity. How far this accumulation of water affects the condition of neighboring grounds by lateral infiltration, we do not know, but we shall see, in a subsequent chapter, that water is conveyed to great distances by this process, and we may hence infer that the influence in question is an important one.

It is undoubtedly true that loose soils, stripped of vegetation and broken up by the plough or other processes of cultivation, may, until again carpeted by grasses or other plants, absorb more rain and snow-water than when they were covered by a natural growth; but it is also true that the evaporation from such soils is augmented in a still greater proportion. Rain scarcely penetrates beneath the sod of grass-ground, but runs off over the surface; and after the heaviest showers a ploughed field will often be dried by evaporation before the water can be carried off by infiltration, while the soil of a neighboring grove will remain half saturated for weeks together. Sandy soils frequently rest on a tenacious subsoil, at a moderate depth, as is usually seen in the pine plains of the United States, where pools of rain-water collect in slight depressions on the surface of earth the upper stratum of which is as porous as a sponge. In the open grounds such pools are very soon dried up by the sun and wind; in the woods they remain unevaporated long enough for the water to diffuse itself laterally until it finds, in the subsoil, crevices through which it may escape, or slopes which it may follow to their outcrop or descend along them to lower strata.

Drainage by Roots of Trees.

Becquerel notices a special function of the forest to which I have already alluded, but to which sufficient importance has not, until very recently, been generally ascribed. I refer to the mechanical action of the roots as conductors of the superfluous humidity of the superficial earth to lower strata. The roots of trees often penetrate through subsoil almost impervious to water, and in such cases the moisture, which would otherwise remain above the subsoil and convert the surface-earth into a bog, follows the roots downwards and escapes into more porous strata or is received by subterranean canals or reservoirs. [Footnote: "The roots of vegetables," says d'Hericourt, "perform the office of draining in a manner analogous to that artificially practised in parts of Holland and the British islands. This method consists in driving deeply down into the soil several hundred stakes to the acre; the water filters down along the stakes, and in some cases as favorable results have been obtained by this means as by horizontal drains."-Annales Forestieres, 1837, p. 312.] When the forest is felled, the roots perish and decay, the orifices opened by them are soon obstructed, and the water, after having saturated the vegetable earth, stagnates on the surface and transforms it into ponds and morasses. Thus in La Brenne, a tract of 200,000 acres resting on an impermeable subsoil of argillaceous earth, which ten centuries ago was covered with forests interspersed with fertile and salubrious meadows and pastures, has been converted, by the destruction of the woods, into a vast expanse of pestilential pools and marshes. In Sologne the same cause has withdrawn from cultivation and human inhabitation not less than 1,100,000 acres of ground once well wooded, well drained, and productive.

It is an important observation that the desiccating action of trees, by way of drainage or external conduction by the roots, is greater in the artificial than in the natural wood, and hence that the surface of the ground in the former is not characterized by that approach to a state of saturation which it so generally manifests in the latter. In the spontaneous wood, the leaves, fruits, bark, branches, and dead trunks, by their decayed material and by the conversion of rock into loose earth through the solvent power of the gases they develop in decomposition, cover the ground with an easily penetrable stratum of mixed vegetable and mineral matter extremely favorable to the growth of trees, and at the same time too retentive of moisture to part with it readily to the capillary attraction of the roots.

The trees, finding abundant nutriment near the surface, and so sheltered against the action of the wind by each other as not to need the support of deep and firmly fixed stays, send their roots but a moderate distance downwards, and indeed often spread them out like a horizontal network almost on the surface of the ground. In the artificial wood, on the contrary, the spaces between the trees are greater; they are obliged to send their roots deeper both for mechanical support and in search of nutriment, and they consequently serve much more effectually as conduits for perpendicular drainage.

It is only under special circumstances, however, that this function of the forest is so essential a conservative agent as in the two cases just cited. In a champaign region insufficiently provided with natural channels for the discharge of the waters, and with a subsoil which, though penetrable by the roots of trees, is otherwise impervious to water, it is of cardinal importance; but though trees everywhere tend to carry off the moisture of the superficial strata by this mode of conduction, yet the precise condition of soil which I have described is not of sufficiently frequent occurrence to have drawn much attention to this office of the wood. In fact, in most soils, there are counteracting influences which neutralize, more or less effectually, the desiccative action of roots, and in general it is as true as it was in Seneca's time, that "the shadiest grounds are the moistest." [Footnote: Seneca, Questiones Naturales, iii. 11, 2.]

It is always observed in the American States, that clearing the ground not only causes running springs to disappear, but dries up the stagnant pools and the spongy soils of the low grounds. The first roads in those States ran along the ridges, when practicable, because there only was the earth dry enough to allow of their construction, and, for the same reason, the cabins of the first settlers were perched upon the hills. As the forests have been from time to time removed, and the face of the earth laid open to the air and sun, the moisture has been evaporated, and the removal of the highways and of human habitations from the bleak hills to the sheltered valleys, is one of the most agreeable among the many improvements which later generations have witnessed in the interior of the Northern States. [Footnote: The Tuscan poet Ginati, who hod certainly had little opportunity of observing primitive conditions of nature and of man, was aware that such must have been the course of things in new countries. "You know," says he in a letter to a friend, "that the hills were first occupied by man, because stagnant waters, and afterwards continual wars, excluded men from the plains. But when tranquillity was established and means provided for the discharge of the waters, the low grounds were soon covered with human habitations."— Letters, Firenze, 1864, p. 98.]

Recent observers in France affirm that evergreen trees exercise a special desiccating action on the soil, and cases are cited where large tracts of land lately planted with pines have been almost completely drained of moisture by some unknown action of the trees. It is argued that the alleged drainage is not due to the conducting power of the roots, inasmuch as the roots of the pine do not descend lower than those of the oak and other deciduous trees which produce no such effect, and it is suggested that the foliage of the pine continues to exhale through the winter a sufficient quantity of moisture to account for the drying up of the soil. This explanation is improbable, and I know nothing in American experience of the forest which accords with the alleged facts. It is true that the pines, the firs, the hemlock, and all the spike-leaved evergreens prefer a dry soil, but it has not been observed that such soils become less dry after the felling of their trees. The cedars and other trees of allied families grow naturally in moist ground, and the white cedar of the Northern States, Thuya occidentalis, is chiefly found in swamps. The roots of this tree do not penetrate deeply into the earth, but are spread out near the surface, and of course do not carry off the waters of the swamp by perpendicular conduction. On the contrary, by their shade, the trees prevent the evaporation of the superficial water; but when the cedars are felled, the swamp—which sometimes rather resembles a pool filled with aquatic trees than a grove upon solid ground—often dries up so completely as to be fit for cultivation without any other artificial drainage than, in the ordinary course of cultivation, is given to other new soils. [Footnote: A special dessicative influence has long been ascribed to the maritime pine, which has been extensively planted on the dunes and sand-plains of western France, and it is well established that, under certain conditions, all trees, whether evergreen or deciduous, exercise this function, but there is no convincing proof that in the cases now referred to there is any difference in the mode of action of the two classes of trees. An article by D'Arbois de Jubainville in the Revue des Eaux et Forets for April, 1869, ascribing the same action to the Pinus sylvestris, has excited much attention in Europe, and the facts stated by this writer constitute the strongest evidence known to me in support of the alleged influence of evergreen trees, as distinguished from the draining by downward conduction, which is a function exercised by all trees, under ordinary circumstances, in proportion to their penetration of a bibulous subsoil by tap or other descending roots. The question has been ably discussed by Beraud in the Revue des Deux Mondes for April, 1870, the result being that the drying of the soil by pines is due simply to conduction by the roots, whatever may be the foliage of the tree. See post: Influence of the Forest on Flow of Springs. It is however certain, I believe, that evergreens exhale more moisture in winter than leafless deciduous trees, and consequently some weight is to be ascribed to this element.]

The Forest in Winter.

The influence of the woods on the flow of springs, and consequently on the supply for the larger water-courses, naturally connects itself with the general question of the action of the forest on the humidity of the ground. But the special condition of the woodlands, as affected by snow and frost in the winter of excessive climates, like that of the United States, has not been so much studied as it deserves; and as it has a most important bearing on the superficial hydrology of the earth, I shall make some observations upon it before I proceed to the direct discussion of the influence of the forest on the flow of springs.

To estimate rightly the importance of the forest in our climate as a natural apparatus for accumulating the water that falls upon the surface and transmitting it to the subjacent strata, we must compare the condition and properties of its soil with those of cleared and cultivated earth, and examine the consequently different action of these soils at different seasons of the year. The disparity between them is greatest in climates where, as in the Northern American States and in the extreme North of Europe, the open ground freezes and remains impervious to water during a considerable part of the winter; though, even in climates where the earth does not freeze at all, the woods have still an important influence of the same character. The difference is yet greater in countries which have regular wet and dry seasons, rain being very frequent in the former period, while, in the latter, it scarcely occurs at all. These countries lie chiefly in or near the tropics, but they are not wanting in higher latitudes; for a large part of Asiatic and even of European Turkey is almost wholly deprived of summer rains. In the principal regions occupied by European cultivation, and where alone the questions discussed in this volume are recognized as having, at present, any practical importance, more or less rain falls at all seasons, and it is to these regions that, on this point as well as others, I chiefly confine my attention.

Importance of Snow.

Recent observations in Switzerland give a new importance to the hygrometrical functions of snow, and of course to the forest as its accumulator and protector. I refer to statements of the condensation of atmospheric vapor by the snows and glaciers of the Rhone basin, where it is estimated to be nearly equal to the entire precipitation of the valley. Whenever the humidity of the atmosphere in contact with snow is above the point of saturation at the temperature to which the air is cooled by such contact, the superfluous moisture is absorbed by the snow or condensed and frozen upon its surface, and of course adds so much to the winter supply of water received from the snow by the ground. This quantity, in all probability, much exceeds the loss by evaporation, for during the period when the ground is covered with snow, the proportion of clear dry weather favorable to evaporation is less than that of humid days with an atmosphere in a condition to yield up its moisture to any bibulous substance cold enough to condense it. [Footnote: The hard snow-crust, which in the early spring is a source of such keen enjoyment to the children and youth of the North—and to many older persons in whom the love of nature has kept awake a relish for the simple pleasures of rural life—is doubtless due to the congelation of the vapor condensed by the snow rather than to the thawing and freezing of the superficial stratum; for when the surface is melted by the sun, the water is taken up by the absorbent mass beneath before the temperature falls low enough to freeze it.]

In our Northern States, irregular as is the climate, the first autumnal snows pretty constantly fall before the ground is frozen at all, or when the frost extends at most to the depth of only a few inches. [Footnote: The hard autumnal frosts are usually preceded by heavy rains which thoroughly moisten the soil, and it is a common saying in the North that "the ground will not freeze till the swamps are full."] In the woods, especially those situated upon the elevated ridges which supply the natural irrigation of the soil and feed the perennial fountains and streams, the ground remains covered with snow during the winter; for the trees protect the snow from blowing from the general surface into the depressions, and new accessions are received before the covering deposited by the first fall is melted. Snow is of a color unfavorable for radiation, but, even when it is of considerable thickness, it is not wholly impervious to the rays of the sun, and for this reason, as well as from the warmth of lower strata, the frozen crust of the soil, if one has been formed, is soon thawed, and does not again fall below the freezing-point during the winter. [Footnote: Dr. Williams, of Vermont, made some observations on the comparative temperature of the soil in open and in wooded ground In the years 1789 and 1791, but they generally belonged to the warmer months, and I do not know that any extensive series of comparisons between the temperature of the ground in the woods and in the fields has been attempted in America. Dr. Williams's thermometer was sunk to the depth of ten inches, and gave the following results:

Temperature Temperature Time. of ground in of ground in Difference. pasture. woods. May 23...................... 52 46 6 " 28...................... 57 48 9

June 15...................... 64 51 13 " 27...................... 62 51 11 July 16...................... 62 51 11 " 30...................... 65 1/2 55 1/2 10 Aug. 15...................... 68 58 10 " 31...................... 59 1/2 55 4 1/2 Sept.15...................... 59 1/2 55 4 1/2 Oct. 1...................... 59 1/2 55 4 1/2 " 15...................... 49 49 0 Nov. 1...................... 43 43 0 " 16...................... 43 1/2 43 1/2 0

On the 14th of January, 1791, in a winter remarkable for its extreme severity, he found the ground, on a plain open field where the snow had been blown away, frozen to the depth of three feet and five inches; in the woods where the snow was three feet deep, and where the soil had frozen to the depth of six inches before the snow fell, the thermometer, at six inches below the surface of the ground, stood at 39 degrees. In consequence of the covering of the snow, therefore, the previously frozen ground had been thawed and raised to seven degrees above the freezing-point.—William's Vermont, i., p. 74.

Boussingault's observations are important. Employing three thermometers, one with the bulb an inch below the surface of powdery snow; one on the surface of the ground beneath the snow, then four inches deep; and one in the open air, forty feet above the ground, on the north side of a building, he found, at 5 P.M., the FIRST thermometer at -1.5 degrees Centigrade, the second at 0 degrees, and the THIRD at + 2.5 degrees; at 7 A.M. the next morning, the first stood at -12 degrees, the second at -3.5 degrees and the third at -3 degrees; at 5.30 the same evening No. 1 stood at -1.4 degrees, No. 2 at 0 degrees, and No. 3 at + 3 degrees. Other experiments were tried, and though the temperature was affected by the radiation, which varied with the hour of the day and the state of the sky, the upper surface of the snow was uniformly colder than the lower, or than the open air.

According to the Report of the Department of Agriculture for May and June, 1872, Mr. C. G. Prindle, of Vermont, in the preceding winter, found, for four successive days, the temperature immediately above the snow at 13 degrees below zero; beneath the snow, which was but four inches deep, at 19 degrees above zero; and under a drift two feet deep, at 27 degrees above.

On the borders and in the glades of the American forest, violets and other small plants begin to vegetate as soon as the snow has thawed the soil around their roots, and they are not unfrequently found in full flower under two or three feet of snow.—American Naturalist, May, 1869, pp. 155, 156.

In very cold weather, when the ground is covered with light snow, flocks of the grouse of the Eastern States often plunge into the snow about sunset, and pass the night in this warm shelter. If the weather moderates before morning, a frozen crust is sometimes formed on the surface too strong to be broken by the birds, which consequently perish.] The snow in contact with the earth now begins to melt, with greater or less rapidity, according to the relative temperature of the earth and the air, while the water resulting from its dissolution is imbibed by the vegetable mould, and carried off by infiltration so fast that both the snow and the layers of leaves in contact with it often seem comparatively dry, when, in fact, the under-surface of the former is in a state of perpetual thaw. No doubt a certain proportion of the snow is given off to the atmosphere by direct evaporation, but in the woods, the protection against the sun by even leafless trees prevents much loss in this way, and besides, the snow receives much moisture from the air by absorption and condensation. Very little water runs off in the winter by superficial water-courses, except in rare cases of sudden thaw, and there can be no question that much the greater part of the snow deposited in the forest is slowly melted and absorbed by the earth.