In attempting to develop a system of rural hygiene, by means of which the full value of the advantages of pure air and sunlight, of healthful exercise and sound sleep, may be realized, the first step should be a proper location of the house. For, while it is possible to have good health in houses not advantageously located, and while the influence of unsanitary surroundings is not as great as was formerly supposed, yet there can be no question but that some influences, whether they be great or small, must result directly from the situation of a dwelling. For example, it has been noticed that a house whose cellar was damp was an unhealthy house to live in, and early text-books on hygiene quote statistics at length to prove this fact.

The early theories connecting ill-health with conditions in and around the house have been handed down, and to-day some are accepted as true, although by the modern science of bacteriology most of the early notions have been upset. For example, it was considered dangerous to breathe night air in the vicinity of swamps, and in one of the Rollo Books, so much read by the children of the last generation, Uncle George requires Rollo, on a night journey through the Italian marshes, to stay inside the coach with the windows closed in order not to breathe the night air and so contract malarial fever. We know to-day that malarial fever comes only from mosquitoes, that night air has nothing to do with disease, and we hear the general advice of doctors that, except where it means the admission of mosquitoes, we should always sleep with our windows open in order to breathe as much night air as possible, because the night air is purer than any other air. These early traditions have not only concerned themselves with damp cellars and night air, but they have insisted that even the vicinity of a swamp or pond might lead to disease, and the State Department of Health of New York is in constant receipt of complaints because of alleged danger to health on account of some pond or swamp in the vicinity of houses.

Again, one tradition says that a house should not be located in the midst of a dense growth of trees, because the shade of the trees, however welcome in summer, will generate and maintain a condition of dampness in the house and, therefore, be injurious to the health of the inmates.

Another tradition is that a house ought not to be located in a valley, but that a hilltop, or at least a sidehill elevation, is preferable, the possible dampness of the valley being alleged again as the reason.

To-day, so far as is known, there is no direct evidence of dampness being primarily responsible for any disease, although, heretofore, such diseases as typhoid fever, yellow fever, bilious fever, malarial fever, cholera, and dysentery have all been attributed to miasms springing from damp soil. To-day we are assured by experts that none of these diseases are induced by dampness alone. One could spend his days immersed in water up to his chin and never contract any sickness of the types mentioned merely through that act. Later on, we shall show how the presence of swamps in the vicinity of a house is objectionable because of their providing breeding places for insects, but the dampness itself never has and never will cause disease. As a concrete example, it may be noted that the country of Holland, in large part lying below the level of the sea, with drainage canals and ditches everywhere in evidence, is, in spite of such manifest possibilities of dampness, one of the most healthy countries in the world, as already pointed out in Chapter I. This fact not only emphasizes the small effect of surface waters and damp soils in promoting disease, but also magnifies the value of cleanliness for which the Dutch people are so famous.

Damp soils.

Why is it, then, that damp soils and damp cellars are objected to? Chiefly, because of the inconvenience and discomfort they occasion. A damp cellar means conditions favorable to the development of mildew and rot; prevents vegetables from keeping a normal length of time; accounts for moldy, decaying odors throughout the house, and is generally disagreeable. One is tempted to say that such a condition is also unhealthy, and it is quite possible that a person living over a damp cellar which contains accumulations of decaying vegetables, and breathing air loaded with organic compounds, may gradually lose his normal vitality, and become thereby more readily susceptible to specific diseases, but the diseases themselves will not come from the dampness alone.

The discomfort and inconvenience, however, are quite sufficient reasons to make it eminently desirable to have the house and the cellar dry. With this in mind, the selection of the house site should be carefully made. Instinctively, and with reason, the immediate neighborhood of low, swampy, marshy ground, of stagnant ponds, or of sluggish streams should be avoided. It should not be necessary to warn prospective builders that low land, subject to inundation, even though this may happen only occasionally, is not a wise choice of a building site. Figure 2 shows an inundation in a small village of New York State in 1889. Floods are expected each spring and counted on as a part of the year's experience. The resulting exposure and the inevitable effluvia following the receding waters are both objectionable factors in hygienic living. Similarly, the vicinity of a stream carrying organic matter, such as sewage from a town above, should undoubtedly be avoided on account of possible odors in summer. Not long ago, the writer was told by the owner of a productive farm, situated below a small city in New York State, that in the summer time the windows of his house had all to be kept tightly shut at night, because of the effluvia from a stream a thousand feet distant, which carried the sewage from the city above.

Location of house.

A deep and narrow valley should be avoided, not so much because of the possible dampness in the valley, but because of the noticeably lessened amount of sunlight which such a location involves. For such a house, the morning sun comes up much later, and the afternoon sun disappears much earlier, and, since sunlight is the best foe to disease, the more sunlight enters a house, the healthier are those who live in it. On the other hand, the top of a hill exposes a house to strong and cold winds, not desirable on any account, and involving a large expense for heating in winter. Sloping ground, therefore, facing the south if possible, or better, some knoll which rises above the general surface of a southern slope, affords an ideal location. If the slope is toward the south, north winds are kept off, and every ray of the life-giving winter's sun is captured. If the house itself faces due south, the windows on the north have no sunlight. If, on the other hand, the house faces southeast or southwest, then all sides of the house will receive direct sunlight at some time of the day.

Objections to trees.

The vicinity of trees is not to be regarded as altogether evil, since they provide both shade in summer and a screen against winds in the winter. No disease comes from dampness because of their presence, and the worst thing which may be charged against a thick growth is that it keeps out the sun. Practically two points may, however, be urged against trees growing too close to a house. If near enough for leaves to drop on the roof, rain troughs and leaders become stopped up and cause trouble. A thick growth directly over a shingle roof allows organic matter to accumulate on the shingles, so that vegetation develops and the roof decays more rapidly than if exposed to sun and wind. Again, and it is no trivial matter, a house whose roof is easily accessible from trees is apt to become infested with squirrels, who get into the attic, run through the walls, and become a great nuisance. For these reasons, then, trees should be far enough away from the house to allow the sun to enter the windows freely and to keep away from the roof objectionable animals, large and small.

Space between houses.

It is a law or custom as ancient as the Romans that requires a proprietor to build his house so that the eaves should not overhang on the land of his neighbor. Our grandfathers, with the same idea, used to say that a man should be able to drive his team around his house on his own land. In our day it is highly desirable that a house should be built so as to leave as much land under control between the buildings and the lot line as possible. This, of course, does not apply to houses built on a farm of a hundred acres or more, but rather to the house in a small village where a few hundred people live closely together, under rural conditions. In such a village the water-supply usually comes from wells, and the wastes of the household are discharged into privies and cesspools. There is no law, unfortunately, which restricts the location of either of these two essential structures, and it is quite possible for a well, built within a few feet of a property line, to be ruined in quality by a cesspool, built later, on the other side of the line. It seems very unjust that, after the trouble and expense of building a well, a neighbor may render it worthless by the location of his cesspool, and yet, unless one can prove a direct underground connection between well and cesspool, no law is applicable to prevent the construction of the latter.

Besides such a menace to health, there are other objections to the immediate vicinity of neighbors which can be avoided by a judicious interposition of space. For example, the writer listened through a long evening, recently, to a hearing before a City Commissioner of Health, where one householder and a crowd of witnesses complained of the noise made by a kicking horse in an adjacent stable. The one witness who was not disturbed by the noise, and who lived in the vicinity, was unexpectedly found to be deaf.

It is wisdom also to have a reasonable space between a house and the highway, chiefly because the dust of the road is thereby kept from the house. There are people who find much enjoyment in watching passers-by on the road, and with them front windows would be as close to the road as possible, but it is wiser to have a front yard of at least fifty feet depth when possible.

Finally, the location on a sidehill, even when otherwise advantageous, is to be regarded with suspicion if the subsoil strata are horizontal and neighbors up the slope have cesspools in use. The writer knows of several cesspools, built in rock, which, so far as their owners were concerned, have worked successfully for many years, but the water leeching away through the rock was finally discovered to be the cause of continual dampness in neighboring cellars, on lower ground, to the manifest discomfort of those occupying the houses.

Composition of soils.

Having thus discussed the location of the house with reference to its surroundings, let us now more carefully examine the character of the soil or earth foundation on which the house shall be built. All soil is made up of varying proportions of mineral and vegetable matter in the interstices of which there are usually to be found more or less air, water, and watery vapor. The mineral substances of soil include almost all of the known minerals, although many of them are found in exceedingly small quantities. The most common and the most important mineral elements of the soil of New York State are carbon, silicon, aluminum, and calcium, which combine in various ways to make either sand, sandstone, clay, shale, limestone, or other rock. The particular form which these mineral elements assume is of interest in choosing a location for a house, for two reasons:—

In the first place, it has been asserted that the mineral constituents of a soil directly affect the health of persons living on that material. For instance, the earlier writers on hygiene gravely pointed out that very hard granite rocks, when weathered and disintegrated, became permeated by a fungus and caused malaria. We are, however, now so sure of the cause of malaria that we only laugh at a theory upheld by scientists of only twenty years ago.

Some constitutional diseases, including goiter and cancer, have been supposed to flourish in localities where an excess of calcium exists in the soil, and it is true that these diseases do have an unusual prevalence in certain limited districts; but no modern scientist ventures to say whether the boundaries of those districts are determined by the character of the soil constituents or by some other predisposing factor. The truth is that, in matters not absolutely determined by science, many theories usually have to be evolved and proved worthless before the real cause is found.

In the matter of appendicitis, for instance, it was formerly asserted that the seed of grapes was responsible for the local inflammation, and that one could never have appendicitis if such seeds were not swallowed. This theory is to-day almost forgotten, and one eminent surgeon has asserted that the prevalence of this disease in a district depends on the calcium in the soil, since it is to that mineral that hard water is due, although this has not been substantiated. No information is to-day available by which the fitness of a soil for securing sanitary conditions of building can be determined.

Cancer and soil conditions.

In the case of cancer, however, while no final conclusions can be drawn, there is some definite indication that the soil conditions have connection with the occurrence and continued appearance of cancer. It is known that this dread disease is abnormally prevalent in certain districts of the world where topography and climate are fairly alike. For example, the entire region between the Danube and the Alps from Vienna westward and between the Jura and Alps to Geneva furnishes the highest mortality from cancer in all Europe. The subsoil is clay with a thin covering of surface soil, the hillsides draining on to level valleys with meandering watercourses that frequently inundate and supersaturate the already moist soil.

This condition seems to prevail wherever cancer is abnormally prevalent. In England, in northwestern France, and in Spain the topography described in every case accompanies a high death-rate from cancer. It is of great interest to find that in New York State the two districts that are conspicuously affected by this disease have the same topography. The Unadilla Valley and some parts of the Allegheny Valley are noted for their cancer houses, and in both localities we find the same kinds of hillsides and water-soaked valleys as in Germany and France. It has also been noted that the older geological formations are free from the disease and that an occasional inundation does not seem to be a factor. Altogether there seems to be some ground for assuming a connection between cancer and soil conditions, at any rate until scientists have determined the real cause of the disease in those localities where it is now so markedly prevalent.

Topography.

The soil, however, with its mineral characteristics, does indirectly affect the health of the householder because different kinds of rock form themselves naturally into different surface formations, some healthy and some unhealthy. For example, localities where granite rock abounds and comes near the surface are usually healthy because the surface slope is great enough to carry off all drainage water rapidly. The air therefore is dry and not influenced by the immediate vicinity of swamps. The drinking water is soft, and malarial breeding places are usually absent.

Limestone rock, on the other hand, is commonly laid down in horizontal strata, and while a succession of strata may frequently give rapid slopes, marshes are very common, existing even on the tops of the hills. The drinking water is always to be suspected as to quality because, in the first place, it is hard from absorption of lime, and in the next place, cavities and seams in the rock allow polluting material to travel for long distances.

Sandstone, being porous, may be considered a healthy foundation, and sands and gravels of all sorts are usually free from marshy land.

Gravel has always been assumed to be the healthiest soil on which a house could be built, provided the ground water reaches its highest stage three or four feet below the cellar bottom.

Sand is equally desirable except in the cases where vegetable matter has been mixed with the sand, rendering decay imminent. Water drawn from such sands in the form of springs will contain large quantities of nitrates which may lead to excessive development of vegetable life and may have on the human system the same laxative effect as comes from drinking swamp water.

Clays and heavy alluvial soils are not usually considered desirable soils on which to build. Water does not run from such soils; they hold moisture, and hence are always damp, and marshes are very apt to exist in the vicinity.

Effects of cultivation.

It was formerly thought that extensive cultivation was objectionable from the standpoint of health, that manured fields in the vicinity of a house were undesirable, and that the turning up of a well-manured field with a plow in the spring was a very likely source of fever. It is a very common belief to-day that when water pipes are to be laid in city streets, thereby disturbing the soil and bringing fresh earth to the surface, typhoid or other fevers may be expected. There is, however, no ground for this belief, and the fact that laborers and their families live healthily in the midst of the thousands of acres of sewage-irrigated fields near Berlin, where the heavily manured fields are constantly being plowed, is a sure proof of this. The earlier text-books on hygiene all assert, however, the contrary; Parkes, for instance, says that irrigated lands, especially rice fields, which give a great surface for evaporation and also exhale organic matter into the air, are hurtful, and in northern Italy the rice grounds are required to be three quarters of a mile from the small towns to protect the village inhabitants against fevers. There is no ground, however, for such a requirement.

No evidence can be found that men who work in sewers and who breathe sewer air all the time are especially unhealthy. Statistics show that the laborers on the sewage fields of Paris and Berlin are actually healthier than the average person living within those cities.

No reason can be assigned, based on our present knowledge of bacteriology, why upturned earth or manured fields should be unhealthy except as the breeding of insects may be encouraged thereby. The two essentials, however, which should be considered are: first, the topography or the formation of the soil in order that the surface water may run off freely, and second, the character of the soil so that ground water may not remain too near the surface. Whether the soil is rock or gravel makes very little difference.

Made ground.

One kind of soil, however, is distinctly objectionable, although, fortunately, in the country such a soil is unusual: That is, a soil made up of refuse, whether it be garbage, street sweepings from a near-by city, or factory refuse.

The writer has in mind one enterprising landowner and farmer who offered a near-by city the free privilege of dumping the city garbage on his land. This was done for several years, and the low-lying districts of his farm were all filled to a more advantageous level. This garbage was then covered with about a foot of dirt and the land sold in building lots to enterprising laborers determined to own their own homes. According to the old theories of hygiene, the occupants of such houses should have died like rats, but no particular excess of sickness in the one hundred houses so located could be observed. One must, however, believe, as we shall see later, that the repeated breathing of air drawn from such polluted soil must be unhealthy, even though the mortality records fail to show it.

It is interesting in this connection to note that the organic matter in soil gradually disappears, just as a body buried in a grave will finally decompose. Experiments show that such organic matter as wheat straw or cloth in small pieces rots and decays in about three years. But this depends very largely on an excess of air. If the soil is open and the organic matter loose, oxidation takes place rapidly; but if a large pile of organic matter is buried in clay soil, it will take decades for it to disappear. The vegetable matter in soil is usually produced by the decay of plants which have either grown on the soil or have been washed down into its voids. A great deal was formerly written on the relation between this organic matter and the prevalence of malaria, and some earlier writers believed that the amount of malaria in a district was dependent upon the amount of vegetable dÉbris in the soil. Since we have learned that malaria is carried by mosquitoes, we are less interested in the amount of organic matter in the soil. Its mere presence is not likely to be injurious.

Water in the soil.

Only the hardest rocks are entirely solid, the others containing a certain percentage of voids or interstices. These voids are filled with air or water, as the case may be, and we may stop for a moment to inquire the effect of the presence of this air and water. In loose sands the amount of voids is 40 to 50 per cent of the total volume, in sandstone about 20 per cent, and in other rock reduced amounts. The volume of air, therefore, in the soil under a cellar to a depth of four or five feet, amounts to a good many cubic feet and would not be worth inquiring into except for the fact that it is continually in a state of motion. When the ground water, perhaps normally five feet below the cellar bottom, rises in the spring, this ground air is forced out, and in a cellar without a concrete foundation it rises into the cellar and penetrates into the house.

A house artificially warmed by stoves is continually discharging heated air from the tops of the rooms and colder air is being brought in from below to take its place. This air comes from the ground below, and in open soil may come from a great depth. A case has been noted where gas escaping from a main in a city street twenty feet from a cellar wall was, by the suction due to heat, drawn into the cellar and thence into the rooms of the house. It is possible that air from cesspools and broken drains in the vicinity of a house may, in this same way, contribute to the atmosphere breathed within the walls of the house. Gravelly and sandy soils, therefore, in order to maintain the superiority which they furnish for building construction, should not be polluted, since any pollution in the vicinity influences the quality of air which may get into the house. The method of preventing such ingress is plainly to water-proof the outside walls of the cellar and provide an air-tight floor over the cellar bottom. Methods of doing this will be discussed in the next chapter.

Moisture in soils.

The presence of water in the soil has usually been considered to be unhealthy because of the impression that it led to certain fevers. The writer has heard, for instance, of an attack of malaria being caused by a short visit to a damp vegetable cellar; and it is one of the triumphs of the century that the malarial parasite has been discovered, and the old theory of the dangers of moisture been done away with. A damp cellar has always been considered to be undesirable, but just why nobody knows. A damp cellar causes molds to form rapidly, thus destroying vegetables and other material which might naturally be stored there, but that the presence of moisture in a cellar in itself produces any organic emanation leading to disease is not true, although dampness is essential to the growth of certain organisms.

In the latter part of the nineteenth century, Dr. Bowditch, of Boston, showed that consumption developed most where the surrounding soil was moist, and generally it is the impression that dry air is the only proper air for a consumptive person to breathe. This theory, however, is being rapidly exploded, and patients now remain outdoors in any weather, and no kind of air is objected to by physicians, provided it is outdoor air. Some little time ago the writer was called by a Board of Health to investigate a certain swamp which had some odor, was considered a blot on the landscape in an unusually picturesque village, and was said to be responsible for a long list of contagious diseases. A house-to-house inquiry in the vicinity showed that among some dozen families, only one illness in the last few years could be remembered, and that was an old lady who had been on the verge of the grave for forty years.

It is curious to note the many examples which are cited by the earlier sanitarians to prove the dangerous effect of damp soil. For example, Pettenkofer, a very prominent German hygienist, says that in two royal stables near Munich, with the same arrangements as to stalls, feed, and attendance, and the same class of horses, fever affected the horses very unequally. In one stable, fever was continually prevalent; in the other, no fever was found. Horses sent from the unhealthful to the healthful stables did not communicate the disease. The difference between the two places, says Pettenkofer, was that in the healthful stables the ground water was five to six feet below the surface, while in the unhealthful ones it was only two and a half feet from the surface. A system of drainage by which the ground water was brought to the same level under both stables made them equally healthful. The writer cannot help but feel that some other factor was involved, and while he has no doubt that excessive dampness in stables or cellars is undesirable, he does not believe that such dampness can be directly the cause of fevers of any sort.

It is not desirable, however, to live over a wet cellar nor to maintain a house in a constant condition of dampness, partly on account of its bad effect on the house and partly because such dampness may, by reducing the vitality of the household, become a predisposing factor in disease.

Drainage.

From whatever source dampness may come, it can be guarded against by giving to the surface of the ground in the vicinity of the house, on all sides, sufficient slope away from the walls so that there will be no tendency for water to accumulate against the cellar walls. On the top of a hill this is very easy to do, and the natural surface grade takes care of the surface water without difficulty. On a sidehill or in a valley artificial grading has to be resorted to, except on one side.

Too much emphasis cannot be laid on the necessity for grading the ground surface away from the house. In some cases it may be sufficient to dig a broad shallow trench protected from wash by sods (Fig. 3). In other cases it may be desirable to pave the ditch with cobble stones or to build a cement gutter. In constructing such a surface drain, proper allowance must be made for the accumulation of snow and the resulting amount of water in the spring, so that the distance in which the ground slopes away from the house ought to be, if possible, at least ten feet, so that there can be no standing water to penetrate the house walls. The slope necessary to carry surface water away need not be great. A fall of one foot in one hundred will be ample, even on grassy areas, and if the surface is that of a macadam road or the gutters of a drive, this grade may be cut in two. A slope of more than one foot in one hundred is permissible up to a maximum of seven or eight feet per hundred, more than this being Æsthetically objectionable and tending to make the house appear too high. Whenever gutters are built in driveways or ditches to intercept water coming down the slopes, a suitable outlet must be provided to carry the water thus collected either into underground pipes, by which the water is led to some stream or gulley, or directly into some well-marked surface depression.

Ground water.

The soil always contains water at a greater or less depth, and the elevation of this "ground water," as it is called, varies throughout the year partly with the rainfall and partly with the elevation of the water level in the near-by streams.

It is not at all unusual for this ground water to rise and fall six feet or more within the year, high levels coming usually in the spring and fall, and low levels in the late summer and winter. It is easily possible, then, that a house cellar may seem dry at the time of construction in summer and may develop water to a foot or more in depth after occupancy. The presence of such an amount of water in a cellar, whether injurious to health or not, is objectionable, and a subsoil trench should be provided in order to limit the height to which ground water may rise.

If a system of drainpipes is led around a house extending outward to include the surrounding yard, then the ground water will always be maintained at the level of those pipes, provided the system has a free outlet. Indeed, the question of an outlet for a drainage system is a most important factor, and no system of underdrains can be effective unless a stream or gulley or depression of some kind is available into which the drains may discharge. It is for this reason, quite as much as for any other, that the location of a house on a perfectly level bottom land is objectionable, since the ground there may be normally full of water with no existing depression into which it may be drained.

In the next chapter the proper method of laying drains close to the cellar wall, for the purpose of taking away the dampness from those walls, is described, but another system of drains is desirable, covering more area and more thoroughly drying the ground, provided the ground water needs attention at all. These drains should be laid like all agricultural drainage; and while substitution of broken stone, bundles of twigs, wooden boxes, or flat stone may be made, the only proper material to be used is burnt clay in the form of tile. These tiles are made in a variety of patterns, but the most common in use to-day is one which is octagonal outside and circular inside. They are about one foot in length and may be had from two to six inches inside diameter. The ordinary size for laterals is four-inch diameter, while the mains into which these laterals discharge are generally of six-inch diameter. These tiles are laid in trenches about fifteen feet apart, although in porous soil, such as coarse sand or gravel, this distance may be increased to twenty feet. If the tiles are laid more than four feet below the surface, this distance may be increased, and if the tiles are five feet deep, the distance apart of the several lines may be fifty feet.

The grade of the line must be carefully taken care of, and while it is possible to lay a line of tile with a carpenter's level and a sixteen-foot straightedge, it is much safer to have an engineer's or architect's level and set grade stakes, as in regular sewer work. A fall of one fourth of an inch to the foot is a proper grade, although a greater slope is not objectionable. It is sometimes desirable in soft ground to lay down a board six inches wide in the bottom of a trench on which to rest the tile, but, unless the ground is very soft, this is not necessary. Care must be taken, however, if the board is not used, to have the bottom of the trench very carefully smoothed so that a perfectly even grade in the tile is maintained. There are three ways of laying out a line of trench as shown in the following sketches (Fig. 4). It is usually sufficient to run parallel lines of tile from fifteen to fifty feet apart over the area which it is desired to drain, and let the ends of these lines enter a cross line which shall carry off the water led into it. This cross line should be six inches in diameter as a general rule, unless there is more than a mile of small drains, in which case the size of the cross pipe ought to be increased to eight inches. This cross line then becomes the main outlet, and great care must be taken to see that it has a perfectly free delivery at all times of the year. In cities and sometimes in small villages it is possible to discharge this outlet pipe into a regular public sewer, provided the sewer is deep enough, and provided the municipal ordinances allow such a connection. Otherwise, the outfall must be carried to a natural depression.

In level ground, the problem of finding a suitable outlet is a serious one, and in many cases impossible of solution, so that the householder, being unable to find an outlet, must put up with the ground water and be as patient as possible during its prevalence. It does not do to trust one's eye to find a practicable outlet, since even a trained eye is easily deceived. An engineer with a level can tell in a few moments where a proper point of discharge may be found, and it is absurd to begrudge the small amount which it will cost, in view of the large expense involved in digging a long trench to no purpose.

Some years ago the writer was able to note the conditions in a house where the cellar excavation went three feet into limestone rock. The strata were perfectly level and the cellar floor of natural rock was apparently all that could be desired, smooth and flat, without involving any expense for concrete. One wall came where a vertical seam in the rock existed, and since this natural rock face was smooth and vertical and just where the cellar wall should go, it seemed unnecessary to dig it out and lay up masonry in its place. So it was left and the house built. When the spring rains came, however, the cellar was turned into a pond, water dripping everywhere from the vertical rock face, and coming up through the cellar bottom like springs. It cost a great deal more then to make the changes and improvements necessary in order to secure a dry cellar than it would have done at the outset. This serves as an illustration of the need of taking every precaution at the beginning to insure a dry and well-drained soil around and below the cellar walls.