Chapter II., but probably the worst earth in which a canal can be made

is peat-bog land; the method of procedure is then different to that required in making a railway or a road, and in such soil the construction of a canal should be avoided, as it necessitates most experienced and skilful treatment and extensive drainage, which causes subsidence; difficult maintenance to preserve the channel and retain the water of navigation, control the drainage waters, and keep unimpaired the towing-path, which must be firmly covered throughout. Fine sand is also an unfavourable soil as it so readily becomes a quicksand, but almost all other earths usually met with can be so protected that failure should be an improbable contingency, except when the cuttings are of extraordinary depth, as on the Panama Canal; the chief danger being when there is considerable diversity, irregularity, fissuring, looseness, and upheaval of the soil, for varieties of almost every surface earth may be encountered, each with its own characteristics and behaviour when dry or water-charged, and requiring great attention, especially at the joints, and not a few separate treatment, not only of the soil but of the same earth in a cutting and in an embankment.

It sometimes happens that there is not enough clayey earth to form a canal bank, but if the whole of the excavation, except the mere surface earth, is used there is sufficient, and that it must be employed for reasons of economical construction. In such a case the clayey earth should be deposited upon the water side, and the gravelly or sandy soil on the land side and as a towing-path covering, great care being exercised that no stratification of the earth takes place, or it may separate, and water seams be created. The bottom and slopes of a canal embankment must be covered with an impervious layer, puddle or concrete being most frequently used for the bottom, and puddle or a coating of well rammed vegetable soil sown closely with grass seeds for the slopes, or stone pitching in a wide canal liable to wash: also in loose or doubtful soil. In canal cuttings when the water is drawn off the counter-thrust against the slopes is removed, and unless this is maintained the earth, if in a delicate or loosened condition, may slip, as a flow may be caused of previously dammed up waters. Prior to any works being commenced upon a canal embankment it is advisable to strengthen it for a distance of not less than about 60 feet in length on each side of a proposed railway bridge or other structure, particularly when it will be subject to vibration.

Water will soon find a weak place in any earthwork and, certainly, when an engineer can maintain heavy earthworks in treacherous soils in canal construction, where part of one slope is submerged in an embankment and the other dry, and in a cutting part wet and part dry, he should be able to do so in any analogous situation, making due provision for vibration, the chief disturbing agency canals are not subject to, but which is so potent in its effect upon railway earthwork.

Canal and reservoir embankments are so similar as regards the stability of earthwork that they are here considered under one head. The object of a canal, reservoir, or river embankment is in a desired position to hold water without leakage, subsidence, or deterioration of the earth, or other works connected with the general construction. In dock, canal, or any earthworks made for the purpose of containing or expelling water, it is obvious the position is entirely dissimilar to that of railway works: for whereas a slip or a subsidence in an embankment upon railway work may be of slight moment and give comparatively little trouble or anxiety, any movement in earthwork constructed for the purposes named is of grave importance, as it may mean the destruction of the work, loss of life and property, not only upon the site but also in the surrounding district. No leakage or marked weeping of such an embankment should be disregarded, as unless arrested it will gradually deteriorate the earth until the equilibrium becomes so affected that the embankment fails. As a rule, upon railway works, serious slips, except in very unstable soil, such as drift earth, or those quickly decomposed by atmospheric influences, seldom occur, _vide_ Chapter II., until a year or two has elapsed, the process of disintegration from water being necessarily slower in its action, the material having an opportunity to return to its normal condition, and the surface being more equally exposed. In earthwork for hydraulic purposes, although not subject to frequent and sudden vibration, the slopes being unequally exposed and part generally either permanently or temporarily submerged or alternately wet and dry, and the top and land side unsubmerged and fully open to the action of the weather, percolation cannot be equal and regular upon its surfaces, and slips and subsidences therefore usually occur within a short time, varying from a few weeks to a few months; and after about a year or a cycle of seasons they are not frequent, unless the inherent design and construction of an embankment is faulty, permitting leakage through the submerged slope or bottom, pools of water to collect upon the top, and the bank to become fissured and unequally wet or dry. Especial care should be exercised to ensure a complete and impermeable connection with the earth upon which it is placed.

In embankments to contain or expel water time should always be allowed for an embankment to consolidate before the admission of the water, but in the case of a cutting the presence of the navigation water as soon after construction as practicable may be an advantage, depending upon the nature of the earth in each case; as it may protect the submerged portions of the slopes from the deteriorating effects of being in a constantly changing degree of dryness and wetness, and shield them from the sun and drying winds. Except under peculiar circumstances the unstable places will be known in a few months, and provided a canal is properly constructed, it will not cause much trouble after six months or a year from earthwork movement if the embankments and cuttings receive ordinary attention, as the navigation water will quickly indicate the unconsolidated places. In clay, clay marls, clay loams, and where loamy soil is intermixed with permeable and water-bearing strata, movement of the earth in canals will speedily occur unless proper precautions have been taken in the construction, as they are seldom homogeneous.

Earthworks to contain or expel water should be made proof against even improbable deterioration and accident, and in proportion as the soil is less solid and firm it should be consolidated by ramming or other means, or be covered and protected to prevent it cracking, and only firm and binding material should be used in canal embankments or those holding or expelling water in order to prevent slips and subsidences. Chapter II. treats of some conditions under which slips and subsidences may be expected. No precaution should be omitted that will render the work solid and uniform, and nothing should be left to chance. The location is of great importance, for the earth may vary in stability within a very short distance, not only in character but as regards the superimposition of the strata. Upheaved and distorted beds should be avoided, and all loose and fissured earths, especially rock or other soils having more or less vertical seams: for all embankments constructed upon fissured soils are sure to cause anxiety and trouble, and will always be liable to become in a dangerous state from an excessive rainfall. Impermeable horizontally deposited earth of considerable thickness is that to be desired, or firm ground that cannot slide upon another stratum or be affected by any artificially brought addition to the percolating waters consequent upon the construction of waterworks. Having carefully selected a site, which for waterworks purposes must almost invariably be upon high ground and may have to be upon the top cap soil of a hill and be peculiarly exposed, the next step to guard against a slip or a subsidence is to prepare the foundations so that no leakage or trickling of water can undermine or gradually deteriorate the seat of the embankment; and, therefore, a thorough connection between the deposited embankment and the ground must be established and the whole be prevented from movement.

The selection of the best available earth is one requiring careful consideration, and it may be necessary to experiment to test the capability of the soil to be made watertight by compression or other comparatively inexpensive means. In some cases none may be obtainable; if so, the only course to pursue may be to consolidate the earth as much as possible, and protect it with an impermeable homogeneous and durable covering, and one that experience has shown can be trusted to equally resist percolation. The whole practice of stable earth dam construction is comprised in the employment of homogeneous, fine, and tenacious earth uniformly deposited in thin layers, gently watered sufficiently to aid ramming and consolidation, rolled, pressed, or trodden down by the passage of carts, men, or animals, and in its due surface protection. Heavy rolling is to be preferred, for it is more effective than ramming, as may be judged by the greater compression; the thickness of the layers and the weight being so regulated that it compresses the earth to its state of maximum solidity, and does not pulverize it, as the compression is more uniform, and irregularly compact masses are not created, which not only destroy uniformity of condition of the mass, but cause seams and veins and destroy homogeneity. After rolling, the compressed layer should be gently watered, as the weight of the roller will have made the top crust drier than the lower portion, unless it has been made into too moist a state. The thickness of the layers should not exceed, when the soil is to be heavy steam-rolled, about 6 inches in earth, and 4 inches in clay soils; they may be compressed to about four-fifths to two-thirds of their normal thickness, the degree of compression to ensure maximum solidity indicating the openness of the original earth. In sand, the layers when simply wetted and rammed with an ordinary rammer, should be about 4 to 5 inches in thickness, and ordinary earth about 2 inches. The volume of the sand will be reduced from 10 to 15 per cent. Simple ramming of sand will only reduce it about 6 to about 9 per cent., and water about 4 to 6 per cent., making the total compression as before stated, the quantity of water used being some 20 per cent. of the volume of the sand. Embankments of moderate height have been made of sand when no other earth was available, except clay loamy soil in comparatively small quantities, the surface of the embankment being worked by thin bars, and the loam being incorporated with the sand and made to fill its interstices, surface protection being thus afforded, and all the usual means to promote consolidation being adopted. Wet earth well mixed with a grout of quicklime has also been used to make a firm embankment of little height. Clay is not a good material to use unless it is incorporated with a considerable percentage of sand to prevent fissures and to lessen expansion. Many prefer the sand to be in the proportion of about two-thirds of the mass, only one-third being clay; the soil being then a loam, and cannot be classed as a clay.

There is much diversity of opinion as to the relative value of a central puddle wall or protection of the slope and toe. Such a wall is placed in the centre, not only to give support to an earth embankment and prevent any through leakage, but also to keep the puddle in a uniformly moist condition, and therefore to prevent it fissuring and cracking from exposure to the sun or air. In an embankment erected upon an earth foundation it may be advantageous and warrant the expense, but when the embankment of earth is placed upon rock it should not be erected, for it is impossible to make a watertight joint between rock and clay puddle, and should it be placed in a rock trench leakage will not be prevented by it, and water will accumulate when it does not at once pass through the seat of an embankment until the surface of its base becomes softened, and the whole mass almost worthless either as a means of support or of prevention of through percolation; and, moreover, there is the danger that the clay will draw away from the sides of the rock. What is wanted is, as it were, to insert as far as the top of an embankment an artificial rock having a thorough connection with the foundation, consequently a Portland cement concrete wall should be adopted in such a case and not one consisting of clay puddle. Two great advantages of a Portland cement concrete central support over that of a clay puddle wall are that it does not settle and the weight is evenly distributed, whereas a puddle wall in subsiding may draw away from the earth of the embankment and cause cavities and cracks and consequent leakage. The chief consideration is to produce an embankment having impermeability, solidity, and homogeneity, and a thorough connection at the foundations; and as the central puddle wall system rarely does so, and usually only under circumstances which would probably hardly warrant the expense of its adoption, it is in a state of obsolescence; and should central wall support be necessary, which may not infrequently be the case, Portland cement concrete is being adopted in its place, a solid and durable material being used instead of one varying in shape and size, according to its state of dryness or wetness, liable to fissure, and incapable of being permanently and firmly joined to any material. This system may be described as one of arresting the percolation of water in the interior of an embankment; but the great aim is to prevent any percolation into the mass, as when once a passage is made the water will escape along the line of least resistance, and a fissure become a cavity, a cavity a breach.

The protection of the slope and toe has for its object the prevention of any percolation into the mass, and it may effect all that is requisite, but its efficiency and completeness much depends upon the preparation and deposition of the layers during the construction of an embankment, and also the time that can be allowed for consolidation, so as to prevent any cracking or fissuring of the face covering from settlement of the embankment. It is obvious in the case of a reservoir embankment or a slope alternately submerged and unsubmerged, that an exposed clay puddle covering cannot be used, as heat or the sun’s rays will cause it to crack, although if constantly wet it would succeed. A puddled clay covering with a broken stone bed placed upon it to receive dry or mortar-set stone pitching has been frequently adopted; a simple cement concrete facing about 6 to 12 inches in thickness, depending upon the depth of the water and the nature of the soil, or in conjunction with an asphalt coating of ordinary thickness. In the two latter cases care must be taken that they do not separate from the embankment either from the force of pent-up water, frost, or shrinking of the earth embankment.

As any perforation of the surface must be prevented, in certain districts the covering should be capable of resisting the attacks of rodents and crustacea, and, therefore, a stone pitched or concrete covered slope is necessary, or the puddle towards the surface must be well incorporated with small stones or ashes or other tough material. In European countries the effects of the burrowing of rats may be insignificant, but in warmer climates, as, for instance, on the coast of Coromandel and in some parts of Bengal, where rats may measure as much as 2 feet in length, their attacks are not to be disregarded with impunity.

The causes of failure of water-containing embankments afford an indication of the direction in which especial care should be exercised in their construction. Assuming a reservoir embankment to be of the necessary form and bulk, and to be properly constructed, the principal causes of failure are as follows:—

1. Leakage along the line of a culvert or pipe passing through the lower portion of an embankment.

2. Leakage under the seat of an embankment.

3. Water overflowing the top and eroding the land slope and so destroying the equilibrium.

4. Bursting of springs over the site. _Vide_ Chapter XII. on “boils” in loose soils.

With regard to the first, the most frequent cause of the failure of a water enclosure embankment; the culvert or outlet passages have been rendered unnecessary by conducting the waters in a tunnel passage under the seat of the embankment and without interfering with it; but this method is expensive. One of the causes of failure is the weight of the embankment owing to unequal settlement producing a breach in a culvert, the probable result of insecure foundations or want of a firm concrete base to the culvert to evenly distribute the weight; or it being placed upon clay puddle, which should never be done; or the embankment being constructed without due care. All culverts should be sufficiently large for a man to easily pass through and should be equally watertight within and without, for a leakage from the culvert to the bank is equally dangerous, and means should be provided so that it can be closed in short lengths. Reports on the temporary failures of reservoir embankments almost invariably state, failure occurred from water penetrating between the puddle and the culvert, from water percolating between the rock foundation and the central puddle wall, or the embankment gave way as water issued through interstices in it caused by settlement of the masonry outlet passage.

Careless construction has often been shown by inspection to be the reason of so many embankments yielding along the line of the culvert; but when the latter is properly designed and built it seldom causes a temporary failure of a well-made reservoir embankment. Respecting the second cause of failure, it generally proceeds from want of care in thoroughly binding an embankment to the solid ground and protecting the toe; by “boils” in the foundations; or by a porous earth seam, such as fine sand, existing under the stratum upon which the embankment is erected becoming a quicksand on flowing water reaching it.

In connection with the third, it may be said that it is not a frequent cause of failure, as provision is almost always made to prevent it, but it may occur in a reservoir embankment from extraordinary circumstances, or in a much less degree from sufficiently high waves being generated upon a lake or impounding reservoir that the top of the bank may be loosened, and the water dash over it and erode the land slope, and convert it into a kind of tail-race; tarpaulins have been temporarily laid upon the surface on an emergency to protect a soft place, also sand bags and planking to prevent an overflow. In an embankment erected for such a purpose the top width is more severely strained, consequent upon the greater exposure, than in a reservoir embankment, and the height must be sufficient to prevent water passing over it.

With reference to the fourth cause of failure, “boils” in loose soil and the bursting of springs are referred to in Chapter XII.

When a reservoir is emptied the weight of the water on its bottom is removed, but the load from the embankment is the same, and should the ground be soft the embankment may subside towards the reservoir and the bed be uplifted; hence it may be advisable not to draw off the water unless the bottom is weighted or movement prevented.

Due regard of the different causes of failure herein enumerated and care in construction, will reduce to a minimum the probability of a slip or a subsidence in an earth embankment erected to contain or expel water.

The temporary or permanent diversion of rivers or streams being so often necessary in public works, a few paragraphs are here devoted to it so far as regards earthslips and subsidences. If it be possible when the soil is very porous and incapable of retaining water, it is advisable not to divert a river.

Some of the most vulnerable places in a newly-formed river-bank are:

The ends that join it to the old bank or to the land, to which it should be thoroughly connected.

The toe of the slope and seat, which should be tied into the bed of the river or be well protected by making the slope flat towards the base.

Any abrupt bends or angles should always be avoided as they increase erosive action.

The wind and water line which requires especial protection.

Provided these points are remembered and the usual precautions taken in forming a river-bank to make it thoroughly sound and homogeneous, a slip or subsidence of serious moment is improbable.

In order to protect the sandy bed of a river and to prevent the banks slipping and subsiding, it may be necessary to guard against scour of the bed and consequently of the toe of the banks. Stone thrown in will settle and compress the bed by weighting and consolidation. By periodical depositions the sand becomes more protected and the quantity of stone required is reduced, but especial care should be taken to preserve the normal bed, to offer no obstruction, and not to cause whirlpools or to interfere with the current except to direct and train it, or the erosive action so created will cause movement. Stones simply cast in and allowed to sink and find a permanent bed until the regular surface of the bottom of a river is so reached, have been proved in many instances to be a sure protection in sandy soils provided eddies do not exist. The preservation of the slopes is particularly referred to in