singular idea of a M. Flachon de la Jomarière, who proposes to pour upon the besiegers, when they are about to crown the covered-way, an enormous quantity of water from powerful engines, which, he says, will make the soil so liquid that it cannot be worked.

His New System of Improving Fortifications is, of course, grounded on these principles. "The spirit,' says A. Carnot, of the new system of fortification, consists in procuring, by the particular combination of the parts which compose it, numerous debouches on all the avenues of the place, so that the besiegers may not be able to establish themselves near it without being exposed to be suddenly attacked, at all times, by all the garrison. From this the enemy will not be able to present himself any where, without keeping troops constantly drawn out, ready to repulse any sortie the besieged may unexpectedly make, and which they may renew whenever they please. The besiegers will therefore be obliged to accumulate troops on all parts of the immense circumference which they must occupy, to embrace the defences of the place; and as in the défense raprochée all this development of force is within the influence of vertical fire, showers of projectiles will carry off some men every moment, and at length entirely crush the besiegers.'

This torrent of vertical fire' is thrown from casemated mortar-batteries, the positions of which are determined from an acknowledged defect in Vauban's systems, viz. the deficiency of fire on the prolongations of the capitals of the bastions, but which fault M. Cormontaingne has remedied by constructing redoubts in the reentering places of arms.

M. Carnot's ideas of the irresistible effect, and exclusive advantage, of this profusion of vertical fire in defence, are such, he asserts, p. 445, that it will change entirely the character of the operations of a siege. According to the existing practice,' he says, 'the besiegers are covered, and the besieged exposed. In the new system, on the contrary, the besieged are covered, but the besiegers exposed to a profusion of feux verticaux, which will reach them behind their parapets and lodgments, enabling the besieged to defend their out-works, without occupying them, merely by pouring upon them torrents of vertical fire when the assailants move forward to the attack.'

M. Carnot then arranges his new system; the casemated mortar-batteries are placed in interior enclosures in the gorges of the bastions, so as to fire in the direction of their capitals. There are nine casemates in each battery: of these, seven contain mortars or pierriers, two in each; the other two (the extreme casemates) are each armed with three guns, for the defence of the ditch of the retranchement général. The escarpe of the retranchement général is a detached wall placed in front of the rampart, leaving a chemin des rondes eighteen feet wide. The exterior slopes of the ramparts are all forty-five degrees. The bastions are also covered by a detached wall erected near the base of the exterior slope of the rampart, leaving a chemin des rondes six feet wide.

The ditch of the bastion is thirty-six feet wide. Counterguards are placed before the bastions. The demi-lunes are works of the same profile as the counterguards. Sometimes M. Carnot calls his counterguards and demi-lunes glacis coupés, and under this name recommends them for improving the defences of existing places. The cavaliers are placed in front of the tenailles, and communicate with them by caponnières.

The counterguards and demi-lunes have ditches thirty-six feet wide at bottom, the counter slopes forming a reverse glacis of forty-three yards which M. Carnot calls glacis en contrepente.

In old fortresses M. Carnot proposes to convert a portion of each bastion into a counterguard, by making a ditch, about thirty-six feet wide across the bastion, from the middle of each flank, in the directions of lines of defence; the two branches of the ditch meeting, consequently, on the capital. The part thus enclosed is formed into a bastion, by making parapets upon the interior lines of the ditch, which thus become the faces of a bastion so small that its flanks are but sixty feet long-sufficient only to receive three guns. The new ditch is consequently very little defended by flank-fire; but this, consistently with the principles already noticed, M. Carnot has here also sacrificed to the superiority of vertical fire.

For the purpose, chiefly, of being able to make sorties with facility, M. Carnot proposes to convert the glacis into a glacis en contrepente, and, with the earth furnished by the excavation, to form the upper part of the old glacis into a counterguard or glacis coupé, raised nearly as high as the body of the place. The interior slope of the new work occupies the greater part of the old covered-way. The traverses are removed; and, instead of palisades, a brick wall furnished with loop-holes is constructed very near the counterscarp. The exterior slope of the glacis coupé is so abrupt that no part of it can be seen from the body of the place; and the greater part of the advanced ditch formed by this alteration cannot be seen at all.

Colonel Sir Howard Douglas has published some spirited and scientific Observations on the Motives, Errors, and Tendency of M. Carnot's Principles. We have already quoted this writer, and it is but fair to add, that he seems to make a formidable attack upon the principles and constructions of that able engineer. We abstract a sufficient portion of his remarks and experiments to place the whole subject fairly before the reader.

'It is quite clear,' observes this writer, 'that M. Carnot has formed his theory upon the parabolic hypothesis, which, I must inform such readers as are not acquainted with these matters, is the theory of a projectile's flight in a non-resisting medium. This theory, considerably erroneous in all cases, is particularly and greatly so with small projectiles; and its deductions, as applied to the velocity of descent of small balls used in very elevated short ranges, are quite fallacious. The velocity of the ball in a horizontal direction (which by this theory would be constant, and to the projectile velocity as radius to the cosine

of the angle of elevation) being inconsiderable, it is evident that the effect of vertical fire must depend upon the velocity of descent in the direction of the curve. Estimating this according to the parabolic theory (as the secant of the angle of elevation), the motion would be slowest at the vertex of the curve, and the velocities of the projectile be equal at equal distances from that point. According to this supposition we should assign to the descent of small balls, discharged at an elevation of seventy-five degrees, or eighty degrees, such accelerated velocities, as would, if true, be quite sufficient to do good service in the way M. Carnot suggests; but the fact is, that there can be no acceleration beyond a limit which, with small balls, is very much less than is generally imagined. From the vertex of the curve, where all the vertical motion is lost, the ball begins to descend by an urging force which is nearly constant, viz. its own weight. This force would produce equal increments of velocity, in equal times in vacuo, but in air, the descent of the ball being resisted more and more as the velocity accelerates, the urging force will, at a certain velocity, be opposed by an equal resistance of air, after which there can be no further acceleration of motion, and the ball will continue to descend with a velocity nearly terminal.

"When I began to consider this interesting problem, as applied to vertical fire, I was soon satisfied that M. Carnot had entirely overlooked terminal velocity; and I shall show, from his own words, that this is the case. It is not ne cessary to exhibit here the investigations by which I have established the impotency of M. Carnot's vertical fire; I shall only state the results, not to embarrass the conclusions with abstruse matter. The solutions are computed from the theorems given in Dr. Hutton's tracts, and, although the results may differ a little from the truth, yet it is quite clear, that in the descent of the balls there can be no acceleration of motion beyond a certain limit; that with small balls this velocity is very much less than persons who have not investigated this curious problem would imagine; and that M. Carnot has evidently overlooked this circumstance.

The velocity which a musket ball has acquired when the resistance becomes equal to the weight, or urging force of descent, is only about 180 feet in a second. The potential altitude, or the height from which the ball must descend in vacuo, to acquire a velocity equal nearly to the terminal velocity, is 523 feet. Hence, in the first place, it would be a waste of means to use the full charge; for a musket ball fired upwards, with the ordinary quantity of powder, would be projected to a greater height than 523 feet; and it is evident that all above this is unnecessary. The indentation which a musket ball, moving with a velocity of 180 feet per second, makes on a piece of elm timber, is about of an inch this might, perhaps, be sufficient to knock a man down, if by great chance it were to fall upon his head; but in no other case would it put him 'hors de combat.'

'Now, as to the four-ounce balls. The diameter of a French four-ounce ball is one

'M. Carnot recommends that the balls should be made of hammered iron; but adds, that, as the charge of powder for a mortar is sinall, balls of cast-iron may resist the explosion without breaking, and will answer as well. Now this observation shows that the author had not considered the effect of the air's resistance, nor doubted a sufficiency of force in his vertical fire for the weight of a ball of hammered iron is greater than that of a ball of cast-iron of equal diameter, and the superior weight or urging force of the former would generate greater terminal velocity than a lighter ball of the same size could acquire; the momenta of the two balls in question would be as nineteen to eighteen.

'Four-ounce balls, discharged at elevations even considerably above forty-five degrees, to the distance of 120 yards, would not inflict a mortal wound, excepting upon an uncovered head. They would not have force sufficient to break any principal bone; there would be no penetration, but merely a contusion. This certainly would not oblige the besiegers to cover themselves with blindages, as M. Carnot imagines; for a strong cap or hat, and a cover of thick leather for the back and shoulders, would be sufficient protection from the effects of his vertical fire with small balls. As the quantity of balls required to feed mortars discharging 600 balls at a time would be very considerable, M. Carnot observes that cubes of iron of eight or ten lines side, cut from square bars of this dimension, may be substituted. These, he says, may be fired from mortars, howitzers, or stonemortars, and will produce the same effect as balls (page 491, Carnot).

'Let us consider this: Ten lines French are equal to 89523 in. English. The content of the cube is Its weight is

'Now take a ball of the Its diameter is

Its terminal velocity is
Its potential altitude is

⚫71746 3.0822 ounces. same weight:

1.111 inches.

. 185 feet per sec. 534 feet

"We have no experiments from which we can ascertain the terminal velocity of square shot; but, from comparative experiments with round and flat surfaces, we know that the resistance of the air to the flat end of a cylinder is more than double the resistance to a ball of the same diameter. Thus, although the urging force of a ball and cube of the same weight be the same, yet the surfaces upon which the resistance acts (and very irregularly in regard to the cube) are very different:

The surface of the ball is the cube is.

3.87045 4.80862

'From this, together with what has been said respecting the descent of balls, we know, and that is enough for our present purpose, that the terminal velocity of the cube must be much less than 185 feet per second; and consequently its effect or momentum inferior to that of a 3-08 ounce ball. The motion of a cubical shot will, besides, be quite irregular, descending sometimes with an angle, then a face, then an edge foremost, tumbling over and over in oblique, irregular directions, without any certainty, excepting that the velocity and effect will be much less than those of a round shot of equal weight.'

Our author smiles at the preference of the French writer for cross bows and ancient weapons of attack and defence; and compares the far shorter time in which Calais, Tournai, Thouars, Naples, &c., have fallen before fire arms. lle contends that sieges became uniformly shorter as gunnery was improved. Upon the main topic of this writer, he adds, "I give the results of some very careful experiments, made purposely to ascertain the precise effects of those natures of vertical fire, which M. Carnot proposes to adopt as the principal means of defence.'

Experiments with different charges of stones from

a ten-inch iron mortar. Elevation 45°.

1. Charge 10 oz. of powder, and fifty flint stones, each about 14 oz. The average range was 107 yards; but most of the stones were blown to pieces.

2. Charge 12 oz. of powder, and forty stones of hard granite of about 1 lb. each. The nearest stones fell at forty, and the furthest at 120 yards; the transverse spread was thirty yards.

3. Charge 16 oz. of powder, and forty-six stones, as before. The nearest stones fell at fifty, and the furthest at 130 yards from the mortar: the spread was forty-five yards. One stone went off to the right in an angle of about Forty-five degrees, and fell at the distance of 100 yards in that direction, very near a spectator placed, as he thought, in perfect safety. Experiments with a brass pierrier. sixteen inches. Elevation 45°.

Diameter

1. Charge 24 lbs. of powder (which filled the chamber), and 100 granite stones of 11b. each, piled up to the mouth of the pierrier in a basket with a bottom of wood. The nearest effect was twenty-eight yards; the furthest 300: the spread was seventy yards. Many of the stones broke.

2. Charge 1 lb. of powder, and seventy-five granite stones of 1 lb. each. The nearest effect was twelve yards, and the furthest 180. The spread was fifty yards.

In both cases it could not well be ascertained where the greatest effect was, on account of the great dispersion of the stones, many of which broke even with the reduced charge.

Applying these experiments to the new defences of M. Carnot, Sir Howard found that the nearest effect would take place in the gorge of the bastion; and that the furthest effect, P, plate I., would not reach the crest of the glacis, even with the full charge of powder. It appears, therefore, says he, that neither the third parallel, VOL. IX

nor the couronnement of the glacis, are within the reach of stones forced to the utmost, from pierriers in the casemated battery; and the horizontal area of all those parts of the attack which come within its influence is so small, compared with the vast magnitude of the oval surface upon which the stones fall, that, it may be relied upon, not one stone in 1000 would take effect upon the besiegers.

A substitution of large balls and grenades, adds this writer, fired from mortars, would be less uncertain and more formidable; but even with these the dispersion is very great. 100 iron balls, of one pound each, were discharged from a ten-inch iron mortar, at forty-five degrees elevation, with a charge of 11b. 4 oz. of powder. The spread was fifty yards; the nearest effect 150, and the furthest 210 yards: the longitudinal dispersion was therefore sixty yards, and consequently the area of the surface affected by the descent of the balls, supposing it to be an ellipse whose axes are sixty and fifty yards, was 20,476 square feet. The chances of hitting would therefore be very remote, whilst the expenditure of iron would be immense. At the rate of discharge which M. Carnot mentions, page 231, it would require a provision of nearly 1,500,000 lbs. of iron for the

seven casemates of one batterie de gorge.

'In regard to the display of vigor and resolution in personal conflict, which M. Carnot seems to think comparatively deficient in modern defences, it is clear,' says the above writer,' that the invention of gunpowder has narrowed the opportunities of displaying those qualities in the operations of a siege, properly conducted, more than in any other military enterprise. There is no opportunity for personal conflict, excepting in sorties, which, we have already shown, prove too frequently but a waste of life and valor, and in the defence of breaches, where also there is that to encounter which the ancients were not exposed to.

M. Carnot's object in quoting so many sieges, is, to show that the defence of places by 'armes blanches' has constantly been more brilliant, more efficacious, of longer duration than by armes à feu. What, he says (p. 239), has the invention of powder, or the new process of attack, to do with the vigor and resolution that were used by the ancients? These, he observes, may alter the means but not the principles of resistance.

'Now here we differ from M. Carnot; and to close properly with this assertion, we have, rather fully, compared the ancient with the modern means of attack, for the purpose of showing that the general principles as well as means of defence are altered, and that both are inferior to those of attack when directed by scientific intelligence, and furnished with sufficient means. What can personal vigor and resolution do against the establishment of the ricochet batteries, and all the process of attack, until it come near enough to be checked by sorties? The defence by armes blanches' can only be applied to the defence of a breach; but a breach may always be made, whatever be the vigor, resolution, or strength of the garrison. The only means to oppose and retard the opening of a breach are by a powerful fire of artillery in the 2 II

Long-established practice; and that the defence of the breaches at Badajoz, which has thrown some popular lustre on M. Carnot's work, could not have succeeded against an attack conducted, throughout, according to regular process. M. Carnot may perhaps dispute illustrations from British talent and experience, or we should have presented him with other references to facts contained in colonel Jones's excellent work, in support of other parts of our reasoning.'

PART II.

OF FIELD FORTIFICATION.

Field, or temporary fortification, having the same general objects as more permanent works, only differs from them in the means that may be accessible for attacking or defending them.

Field works are thrown up, merely for a short time: often in haste, without either choice or preparation of the materials employed; with very few means at hand, and sometimes in presence, as it were, of the enemy; besides, there are many cases in which they are not intended to resist an attack supported by cannon, and, when they are, the nature of the guns which will probably be brought against them may be different, according to the importance of the works. Lastly, field-works are usually attacked by troops formed into columns; which, advancing rapidly in the direction of their capitals, threaten many points at once; therefore, the dispositions for their defence, ought to be different from those of permanent works, &c. The maxims or general rules that are to be observed in them are;

1. In general, a saliant angle should not be less than sixty degrees, especially when it is undefended by any flank fire.

2. The saliants being the most exposed points, particularly when they are not flanked, their defence ought to be carefully attended to; when the ground, and intended object of the work you construct, will allow you to direct the saliants towards some natural obstacles which prevent the enemy approaching them on the prolongation of the capitals, you ought to avail yourself of that advantage; but, if you cannot direct the saliants thus, they must be protected, if possible, by some artificial obstacles.

3. In tracing field-works, let there be as many flank defences as possible.

4. When one part of a fortification is to flank another, it must be so disposed as to make with it an angle not less than ninety degrees, and ex ceeding as little as possible ninety degrees; in order that the ditch and counterscarp of the part flanked, may be defended by a direct fire from that which flanks it.

5. The length of the lines of defence ought not to exceed eighty toises at most.

6. Avoid the second flank defence, unless you are obliged to have recourse to it.

7. Be careful not to suffer any cover in the vicinity of a work, under which the assailants may approach unperceived.

8. Dead angles are to be avoided as much as possible.

9. A fortification must always be proportioned to the number of men who are to defend it;

and, the length of the parapet remaining the same, you ought to enclose within it the greatest possible surface.

10. Before you begin a work, you ought to ascertain whether you have sufficient means for completing it in time.

We can only find room for a sketch of the principal or out-line of field-works.

Of reduns or flèches. As redans or flêches, plate VI. fig. 3, can be quickly and easily constructed, they are frequently used in the field, where few means are at hand; besides, in many circumstances, the intended object of a work does not require that it should be able to afford an obstinate defence. Weak indeed is that which a redan can make, particularly when isolated; for then, independent of being easily carried in front, owing to the undefended sector fag, its gorge b c is also greatly exposed, and you ought not to rely on the defence of a redan, unless it is supported in its rear; such as, for instance, redans thrown up in front of an army you intend to intrench, and on the banks of a river to cover a bridge, or defend a ford.

Sometimes redans are placed in front of a main work, either to cover its communications with the country, or to defend some parts of the ground which cannot be seen from it, and would be of advantage to the enemy in directing their attacks: or in short to procure a cross fire on the capitals of the main work, and keep the enemy at a distance from it. Redans so disposed are called lunettes.

No fixed rules can be given with regard to the length and direction of the faces of a redan, since both vary according to the ground, the intended object of the work, and the strength of the detachment that it is to cover, &c.

Of redoubts.-Redoubts, as well as redans, are frequently used in the field; where, as isolated works, they are employed, when the post or detachment to be intrenched being abandoned to its own strength, and without any protection in its rear that may prevent its being turned, it becomes necessary to enclose it entirely, so as to secure it from the attacks which the enemy may make upon it on all sides. Redoubts are extremely proper for covering an advanced post, a grand guard, or a communication; for defending a defile, a height; for protecting a retreat,

the

passage of a river, ford, or bridge; for supporting the wings of an army, a line of frontiers, &c.; independent of being easily constructed, they have also the advantage of affording a very good defence when supported from without, and even of being sometimes effectually used instead of fortins or field-forts, which in general require more time and materials for their construction, and a more numerous garrison for their defence.

The requisite length of the sides of a redoubt depends, not only on the extent which the parapet must have, in order that the garrison may man it properly, but on the necessary interior space for containing the men. It should also be considered, whether the troops are to reside in the work, or to remain there for a short time; as it happens, for instance, when a work is sufficiently near a main body of troops to communicate easily with them, and receive reinforce