Chapter XII: TECHNOLOGY.--The Tanning Materials of Europe.--The natural (2)

In France, we have a beautiful mountain dog--the dog of the Pyrenees--which is from 32 to 34 inches in height at the shoulders, and has a very thick white coat, spotted above with pale yellow or grayish fox color. It is very powerful, and is capable of successfully defending property or flocks against bears and wolves.

The Alpine dog is the type of the mountain dog. It is of the same size as the dog of the Pyrenees, and differs therefrom especially in its coloring. It is white beneath, with a wide patch of orange red covering the back and rump. The head and ears are of the same color, with the addition of black on the edges; but the muzzle is white, and a stripe of the same color advances upon the forehead nearly up to the nape of the neck. The neck also is entirely white. There are two varieties of the Alpine or St. Bernard dog, one having long hair and the other shorter and very thick hair. We give in Fig. 1 a portrait of Cano, a large St. Bernard belonging to Mr. Gaston Leonnard.

Although this breed originated at the celebrated convent of St. Bernard, it no longer exists there in a state of purity, and in order to find fine types of it we have to go to special breeders of Switzerland and England. The famous Plinnlimon, which was bought for $5,000 by an American two or three years ago, and about which there was much talk in the papers, even the political ones, was born and reared in England. It appears that it is necessary, too, to reduce the number of life-saving acts that it is said are daily performed by the St. Bernard dogs. This is no longer but a legend. There was, it is true, a St. Bernard named Barry, now exhibited in a stuffed state in the Berne Museum, which accomplished wonders in the way of saving life, but this was an exception, and the reputation of this animal has extended to all others of its kind. These latter are simply watch dogs kept by the monks for their own safety, and which do not go at all by themselves alone to search for travelers that have lost their way in the snow.

The Newfoundland dog, which differs from the preceding in its wholly black or black and white coat, was, it appears, also of mountain origin. According to certain authors, it is indigenous to Norway, and was carried to Newfoundland by the Norwegian explorers who discovered the island. Adapted to their new existence, they have become excellent water dogs, good swimmers, and better life savers by far than the majority of their congeners.

Is it from descending to the plain that the mountain dogs have lost their long hair and have become short haired dogs like the English dog or mastiff and the German or large Danish dogs? It is very probable. At all events, it is by this character of having short hair that mastiffs are distinguished from the mountain dogs. Again, the large breed of dogs are distinguished from each other by the following characters: The mastiff is not very high at the shoulders (30 inches), but he is very heavy and thick set, with powerful limbs, large head, short and wide muzzle and of a yellowish or café-au-lait color accompanying a black face; that is to say, the ears, the circumference of the eyes and the muzzle are of a very dark color. The German or large Danish dogs constitute but one breed, but of three varieties, according to the coat: (1) those whose coat is of a uniform color, say a slaty gray or isobelline of varying depth, without any white spots; (2) those having a fawn colored coat striped transversely with black like the zebra, but much less distinctly; (3) those having a spotted coat, that is to say, a coat with a white ground strewed with irregular black spots of varying size. These, like those of the first variety are generally small-eyed. Whatever be the variety to which they belong, the German or large Danish dogs are slimmer than, and not so heavy as, the mastiffs. Some, even, are so light that it might be supposed that they had some heavier blood in their veins. They have also a longer muzzle, although square, and are quicker in gait and motions.

The largest dogs are to be met with in this breed, and the beautiful Danish dog belonging to Prof. Charcot (Fig. 2) is certainly the largest dog in France and perhaps in Europe. It measures 36 inches at the shoulders and has an osseous and muscular development perfectly in keeping with its large stature, and at the same time has admirable proportions and lightness, and its motions are comparable to those of the finest horse.

Among the English dogs or mastiffs, we very frequently meet with individuals in which the upper incisors and canines are placed back of the corresponding ones in the lower jaw, this being due to a slight shortening of the bones of the upper jaw, not visible externally. This is the first degree of an artist of teratological development, which, since the middle ages, has become very marked in certain subjects, and has given rise to a variety in which this defect has become hereditary. Such is the origin of the breed of bulldogs. The latter were originally as large as the mastiffs. Carried to Spain under Philip II., they have there preserved their primitive characters, but the bulldogs remaining in England have continued to degenerate, so that now the largest are scarcely half the size of the Spanish bulldog, and the small ones attain hardly the size of the pug, although they preserve considerable width of chest and muscular strength.

POINTERS.

Man hunted for ages with dogs that he united in a pack; but these packs were of a very heterogeneous composition, since they included strong dogs, light dogs very swift of foot, shepherds' dogs, and others noted for acuteness of scent, and even mongrels due to a crossing with the wolf. It is from the promiscuousness of all these breeds that has arisen our ordinary modern dog.

The pointer is of relatively recent creation, and is due to the falconers. In our western countries, falconry dates from the fourth and fifth centuries, as is proved by the capitularies of Dagobert. This art, therefore, was not brought to us from the East by the crusaders in the twelfth and thirteenth centuries, as stated by Le Maout in his Natural History of Birds.

The falconer soon saw the necessity of having a dog of nice scent having for its role the finding or hunting up of game without pursuing it, in order to permit the falcons themselves to enter into the sport. This animal was called the bird dog, and was regarded as coming from various countries, especially from Spain, whence the name of spaniel that a breed of pointers has preserved. It is quite curious to find that for three or four centuries back there have been no spaniels in Spain. From Italy also and from southern climes comes what is called the _bracco_, whence doubtless is derived the French name _braque_ and English brach. Finally the _agasse_ of the Bretons was certainly also one of the progenitors of our present pointers. It was, says Oppian, a breed of small and very courageous dogs, with long hair, provided with strong claws and jaws, that followed hares on the sly under shelter of vine-stocks and reeds and sportively brought them back to their masters after they had captured them. We have certainly here the source of our barbets and griffons.

Finally the net hunters of the middle ages also contributed much to the creation of the pointer, for it is to them that we owe the setter. It is erroneously, in fact, that certain authors have attributed the creation of this dog to hunters with the arquebuse, since this weapon did not begin to be utilized in hunting until the sixteenth century. Gaston Phoebus, who died in 1391, shows, in his remarkable work, that the net hunters made use of Spanish setters and that it was they who created the true pointer--the animal that fascinates game by its gaze. By the same pull of their draw net they enveloped in its meshes both the setter and the prey that it held spellbound.

Upon the whole, we see that at the end of the middle ages there existed three types of pointers: spaniels, brachs and very hairy dogs, that Charles Estienne, in his Maison Rustique, of the sixteenth century, calls barbets. It is again with these three types that are connected all the present pointers, which we are going to pass rapidly in review.

_The Brach hounds_.--To-day we reserve the name of brachs for all pointers with short hair. The type of the old brach still exists in Italy, Spain, the south of France and in Germany. It is characterized by its large size, its robust form, its large head, its long, flat ears, its square muzzle separated from the forehead by a deep depression, its large nose, often double (that is to say, with nostrils separated by a deep vertical groove), its pendent lips, its thick neck, its long and strong paws provided with dew claws, both on the fore and the hind feet, and its short hair, which is usually white and marked with brown or orange-yellow spots. The old brach breed has been modified by the breeders of different countries, either by hygiene or by crossing with ordinary dogs, according to the manner of hunting, according to taste, and even according to fashion. Thus in England, where "time is money" reigns in every thing and where they like to hunt quickly and not leisurely, the brach has been rendered lighter and swifter of foot and has become the pointer. In France, while it has lost a little in size and weight, it has preserved its moderate gait and has continued to hunt near its master, "under the gun," as they say. The same is the case in Spain, Italy and Germany even. In France there are several varieties or sub-breeds of brach hounds. The old French brach, which is nothing more than the old type, preserved especially in the south, where it is called the Charles the Tenth brach, is about twenty-four inches in height, and has a white and a maroon coat, which is somewhat coarse. It often has a cleft nose and dew-claws on all the feet. The brach of the south scarcely differs from the preceding except in color. Its coat has a white ground covered with pale orange blotches and spots of the same color. The St. Germain brach is finer bred, and appears to be a pointer introduced into France in the time of Charles X. It has a very fine skin, very fine hair of a white and orange color. The Bourbon brach has the characters of the old French brach, with a white coat marked here and there with large brown blotches, and the white ground spotted with the same color; but what particularly characterizes this dog is that it is born with a stumpy tail, as if three-quarters of it had been chopped off. The Dupuy brach is slender and has a narrow muzzle, as if it had some harrier blood in its veins. It is white, with large dark maroon blotches. The Auvergne brach resembles the southern brach, but has a white and black coat spotted with black upon white. The pointer, or English brach (Fig. 3), descends from the old Spanish brach, but has been improved and rendered lighter and much swifter of foot by the introduction of the blood of the foxhound into its veins, according to the English cynegetic authors themselves. The old pointer was of a white and orange color, and was indistinguishable from our St. Germain. The pointer now fancied is white and maroon and has a stronger frame than the pointer of twenty years ago. The Italian brachs are heavy, with lighter varieties, usually white and orange color, more rarely _roan_, and provided with dew-claws, this being a sign of purity of breed according to Italian fanciers. The German brachs are of the type of the old brach, with a stiff white and maroon coat, the latter color being so extensively distributed in spots on the white as to make the coat very dark.

_Spaniels_.--The old type of spaniel has nearly disappeared, yet we still find a few families of it in France, especially in Picardy and perhaps in a few remote parts of Germany. The old spaniel was of the same build as the brach, and differed from it in that the head, while being short-haired, was provided with ears clothed with long, wavy hair. The same kind of hair also clothed the whole body up to the tail, where it constituted a beautiful tuft. The Picard spaniel is a little lighter than the old spaniel. It has large maroon blotches upon a white ground thickly spotted with maroon, with a touch of flame color on the cheeks, over the eyes, and on the legs. The Pont-Andemer spaniel is a Norman variety, with very curly hair, almost entirely maroon colored, the white parts thickly spotted with a little color as in the Picard variety, and a characteristic forelock on the top of the head.

In England, the spaniel has given rise to several varieties. In the first place there are several sub-breeds of setters, viz.: The English setter, still called laverack, which has large black or orange-colored blotches on the head, the rest of the body being entirely white, with numerous spots of the same color as the markings on the head (Fig. 4); the Irish setter, which is entirely of a bright yellowish mahogany color; and the Gordon setter, which is entirely black, with orange color on the cheeks, under the throat, within and at the extremity of the limbs (Fig. 5). Next come the field spaniels, a group of terrier spaniels, which includes the Clumber spaniel, which is white and orange color; the Sussex spaniel, which is white and maroon; the black spaniel, which is wholly black; and the cocker, which is the smallest of all, and is entirely black, and white and maroon, or white and orange-colored, or tricolored.

_Barbets and Griffons_.--To this latter category belong the dogs, _par excellence_, for hunting in swamps. The barbets are entirely covered with long curly hair, like the poodles, which are directly derived from them. They are white or gray, with large black or brown blotches. The griffons differ from the poodles in their coarse and stiff hair, which never curls. They have large brown blotches upon a white ground, which is much spotted or mixed, as in the color of the hair called roan. There is an excellent white and orange-colored variety. The griffons, neglected for a long time on account of the infatuation that was and is still had for English hunting dogs, are being received again with that favor which they have never ceased to be the object of in Germany and in Italy (where they bear the name of _spinone_). Breeders of merit, such as Mr. Korthals, in Germany, and Mr. E. Boulet, in France, are endeavoring to bring them into prominence (Fig. 6). Finally, we reckon also among hunting dogs some very happy crosses between the spaniels and the barbets, which in England are called retrievers or water spaniels.--_P. Megnin, in La Nature_.

* * * * *

RESTOCKING THE SEINE WITH FISH.

A few days ago, at Bougival, a short distance below the dam of the Marly machine, there were put into water 40,000 fry of California trout and salmon, designed to restock the Seine, which, in this region, has been depopulated by the explosions of dynamite which last winter effected the breaking up of the ice jam that formed at this place.

The operation, which is quite simple in itself, attracted a large number of inquisitive people by reason of the exceptional publicity given to the conflict provoked by a government engineer, who, under the pretext that he had not been consulted, made objections to the submersion of the little fish. As well known, the affair was terminated by a sharp reprimand from Mr. Yves Guyot, addressed to his overzealous subordinate.

It would have been a great pity, moreover, if this interesting experiment had not taken place, and had not come to corroborate the favorable results already obtained.

In three years the California salmon reaches a weight of eleven pounds, and, from this time, is capable of reproduction. Its flesh is delicious, and comparable to that of the trout, the development of which is less rapid, but just as sure.

The fry put into the water on Sunday were but two months old. The trout were, on an average, one and a half inches in length, and the salmon two and three-quarter inches. They were transported in three iron plate vessels, weighing altogether, inclusive of the water, 770 lb., and provided with air tubes through which, during the voyage, the employes, by means of pumps, assured the respiration of the little fish.

Our engraving represents the submersion at the moment at which the cylinders (of which the temperature has just been taken and compared with that of the Seine, in order to prevent too abrupt a transition for the fry) are being carefully let down into the river.--_L'Illustration_.

* * * * *

Figures show that the consumption of iron in general construction--other than railroads--in this country has grown from a little more than a million and a half of tons in 1879 to more than six million tons in 1889. Much of this increase has gone into iron buildings. By using huge iron frames and thin curtain walls for each story supported thereon, as is done in a building going up on lower Broadway, New York city, a good deal of space can be saved.

* * * * *

MODERN ARMOR.

By F.R. BRAINARD, U.S.N.

The building of a navy, which has been actively going on for the past few years, has drawn public attention to naval subjects, and recent important experiments with armor plates have attracted large attention, hence it may not be amiss to give a description of the manufacture and testing of armor. It would be interesting to wade through the history of armor, studying each little step in its development, but we shall simply take a hasty glance at the past, and then devote our attention to modern armor and its immediate future.

Modern armor has arrived at its present state of development through a long series of experiments. These experiments have been conducted with great care and skill, and have been varied from time to time as the improvements in the manufacture of materials have developed, and as the physical laws connected with the subject have been better understood. There has been very little war experience to draw from, and hence about all that is now known has been acquired in peaceful experiments.

The fundamental object to be obtained by the use of armor is to keep out the enemy's shot, and thus protect from destruction the vulnerable things that may be behind it. The first serious effort to do this dates with the introduction of iron armor. With this form of armor we have had a small amount of war experience. The combat of the Monitor and Merrimac, in Hampton Roads, in May, 1862, not only marked an epoch in the development of models of fighting ships, but also marked one in the use of armor. The Monitor's turret was composed of nine one-inch plates of wrought iron, bolted together. Plates built in this manner form what is known as laminated armor. (See Fig. 1.) The side armor of the hull was composed of four one-inch plates. The Merrimac's casemate was composed of four one-inch plates or two two-inch plates backed by oak. The later monitors had laminated armor composed of one-inch plates. The foregoing, with the Albemarle and Tennessee rams under the Confederate flag, are about the sum of our practical experience in the use of armor.

European nations took up the subject of armor and energetically conducted experiments which have cost large sums of money, but have given much valuable data. For a long time wrought iron was the only material used for armor, and the resisting power depending on the thickness; and the caliber and penetration of guns rapidly increasing, it was not long before a point was reached where the requisite thickness made the load of armor so great that it was impracticable for a ship to carry it. The question then arose as to what were the most important parts of a ship to protect. The attempted solutions of this question brought out various systems of distributions.

Armored ships were formerly of two classes; in one the guns were mounted in broadside, in the other in turrets. Every part of the ship was protected with iron to a greater or less thickness. In more modern ships the guns are mounted in an armored citadel, in armored barbettes or turrets, the engines, boilers and waterline being the only other parts protected. There may be said to be three systems of armor distribution. The belt system consists in protecting the whole waterline by an armored belt, the armor being thickest abreast of the engines and boilers. The guns are protected by breastworks, turrets or barbettes, the other parts of the ship being unprotected. The French use the belt system, and our own monitors may be classed under it. The central citadel system consists in armoring that part of the waterline which is abreast of the engines and boilers. Forward and aft the waterline is unprotected, but a protective deck extends from the citadel in each direction, preventing the projectiles from entering the compartments below. The hull is divided into numerous compartments by water-tight bulkheads, and, having a reserve of flotation, the stability of the ship is not lost, even though the parts above the protective deck, forward and aft, be destroyed or filled with water. The guns are protected by turrets or barbettes. The deflective system consists in inclining the armor, or in so placing it that it will be difficult or impossible to make a projectile strike normal to the face of the plate. A plate that is inclined to the path of a projectile will, of course, offer greater resistance to penetration than one which is perpendicular; hence, when there is no other condition to outweigh this one, the armor is placed in such a manner as to be at the smallest possible angle with the probable path of the projectile. This system is designed to cause the projectile to glance or deflect on impact. Deflective armor should be at such an angle that the projectiles fired at it cannot bite, and hence the angle will vary according to the projectile most likely to be used. In the usual form of deflective deck the armor is at such a small inclination with the horizon that it becomes very effective. Turret and barbette armor may be considered as deflective armor. The term inclined armor denotes deflective armor that is inclined to the vertical. The kinds of armor that are in use may be designated as rolled iron, chilled cast iron, compound, forged and tempered steel, and nickel steel. Iron armor consists of wrought iron plates, rolled or forged, and of cast iron or chilled cast iron, as in the Gruson armor. Compound armor consists of a forged combination of a steel plate and an iron plate. Steel armor consists of wrought steel plates. Nickel-steel armor consists of plates made from an alloy of nickel and steel.

I have spoken above of laminated armor. To secure the full benefit of this kind, the plates must be neatly fitted to each other; the surfaces must make close contact. This requires accurate machining, and hence is expensive. To overcome this point sandwiched armor was suggested. This consists in placing a layer of wood between the laminations, as shown in Fig. 2. It was found that laminated and sandwiched armor gave very much less resisting power than solid rolled plates of the same thickness. Wrought iron armor is made under the hammer or under the rolls, in the ordinary manner of making plates, and has been exhaustively studied and experimented with--more so than any other form of armor.

Chilled cast iron armor is manufactured by Gruson, in Germany, and is used in sea coast defense forts of Europe.

In 1867 several compound plates were made by Chas. Cammell & Co., of Sheffield, England, and were tested at Shoeburyness, in England, and at Tegel, in Russia. These plates were made by welding slabs of steel to iron; but the difficulties were so great that the idea was abandoned for the time.

Compound armor, as now manufactured, is of two types: Wilson's patent, a backing of rolled iron, faced with Bessemer steel; Ellis' patent, a backing of rolled iron, faced with a plate of hard rolled steel, cemented with a layer of Bessemer steel. Both these kinds are manufactured in England and France in sizes up to fifty tons weight. The Wilson process is used at the works of Messrs. Cammell & Co., of Sheffield, England, and the Ellis process at the Atlas Works of Sir John Brown & Co., of the same place. These are the two leading manufacturers of compound plate.

The method employed by Wilson in making compound plate is to first make a good wrought iron plate. To the surface of this and along each side of the length of the plate are fixed two small channel irons, as shown in Fig. 5. The plate is then raised to a welding heat in a gas furnace, and transferred to an iron flask or mould. Wedges are driven in between the back of the plate and the side of the mould, thus forcing the channel irons up snug against the opposite side of the mould. Moulding sand is then packed around the back and sides of the plate (see Fig. 6). The mould is lowered in a vertical position into a pit. Molten steel, manufactured by either the Siemens-Martin or Bessemer process, is then poured in through a trough that forms several streams, and forms the hard face of the plate. The molten steel as it runs down cleans the face of the wrought iron plate, scoring it in places, and, being of much higher temperature, the excessive heat carbonates the iron to a depth of one-eighth to three-sixteenths of an inch, forming a zone of mild steel between the hard steel and soft iron. The mould is placed in a vertical position to insure closeness of structure and the forcing of gases out of the steel. After solidifying, the whole plate is pressed, and passed through the rolls to obtain thorough welding. It is then bent, planed, fitted, tempered, and annealed to remove internal strains.

In 1887, Wilson took out a patent for improvements in his process of making compound plates. In this method of manufacture he takes a wrought iron, fibrous plate, fifteen inches thick, built up from a number of thin plates. While hot from the forging press, he places this plate in an iron mould (see Fig. 7) about 28 inches deep, and upon it runs "ingot iron" or very mild steel to a depth of thirteen inches. In this form of mould the plate rests on brickwork, and is held in place by two grooved side clamps or strips which are caused to grip the plate by means of screws which extend through the sides of the mould. After solidifying, the plate, which is twenty-eight inches thick, is reheated and rolled down to eighteen inches. This is the iron backing of the finished plate, and it is again put in the iron mould and heated, when a layer of hard steel is run on the exposed surface of the original wrought iron plate to a depth of eight inches. This makes a plate about twenty-eight inches thick. It is taken from the mould, reheated, rolled, hammered or pressed down to twenty inches. After cooling, it is bent, planed, and fitted as desired, then tempered and annealed to relieve internal strains.

The method employed by Ellis in making compound plates is to take two separate plates, one of good wrought iron and one of hard forged steel, placing the forged steel plate on the wrought iron plate, keeping them separate by a wedge frame or berm of steel around three sides, and placing small blocks of steel at various points near the middle of the plates (see Fig. 8). These blocks are called distance blocks. After covering all the exposed steel surfaces with ganister, the plates are put in a gas furnace and heated to a welding heat. They are then lowered into a vertical iron pit with the open side uppermost. The plates are held in position by hydraulic rams, which also prevent bulging. Molten steel of medium softness is then poured into the space between the plates, by means of a distributing trough having holes in the bottom, and after this has solidified, the whole plate is placed under the hydraulic press and reduced about twenty per cent. in thickness. The plate is then passed through the rolls, bent, planed, fitted, tempered, and annealed to reduce internal strains.

In heating the compound plates for rolling, the plate is placed in the furnace with the steel face down, so that the iron part gets well heated and the steel does not become too hot. Great care must be taken not to overheat the plate, and in working, many passes are given the plate with small closings of the rolls. The steel part of a compound plate is usually about one third of the full thickness of the plate.

Forged steel armor, tempered in oil, is fabricated at Le Creusot, France, by Schneider & Co., using open-hearth steel, and forging under the 100 ton hammer. The ingots are cast, with twenty-five per cent. sinking head and are cubical in form. The porter bar is attached to a lug on one side of the ingot. By means of a crane with a curved jib which gives springiness under the hammer, the ingot is thrust into the heating furnace. On arriving at a good forging heat it is swung around to the 100 ton hammer, under which it is worked down to the required shape. A seventy-five ton ingot requires about eight reheatings before being reduced to shape. Having been reduced to shape, the plate is carefully annealed, then raised to a high tempering heat, and the face tempered in oil. It is reannealed to take out the internal strains, care being taken not to reduce the face hardness more than necessary. The Schneider process of tempering is based upon the utilization of the absorption of heat caused by the fusing or melting of a solid substance, and of the fact that so long as a solid is melting or dissolving in a liquid substance, the liquid cannot get appreciably hotter, except locally around the heating surface. The body to be hardened is plunged at the requisite temperature into a bath containing the solid melting body, or is kept under pressure in the solid material of low melting point until the required extraction of heat has taken place, more solid material being added if necessary as that originally present melts and dissolves.

Nickel steel armor is made in a similar manner to the steel plates, the material used in casting the ingot being an alloy of nickel and steel containing between three and four per cent. of nickel.

The Harvey process of making armor consists in taking an all-steel plate and carbonizing the face. This carbonizing process is very similar to the cementation process of producing steel, and by it the face of the plate is made high in carbon and very hard.

The system invented by Sir Joseph Whitworth, of Manchester, England, consists in what might be called scale armor. A section of a sample of the armor represents four plates. The outer layer, one inch thick, is composed of steel of a tensile strength of 80 tons per square inch; the second layer, one inch thick, of steel whose tensile strength is 40 tons per square inch; the third and fourth layers, each one-half inch thickness, of mild steel. The outer layer is in small squares of about ten inches on a side, and is fastened to the second layer by bolts at the corners and one in the middle of each square. The surface is flush. (See Fig. 9.) The end sought by the above system is to break up the shot by the hard steel face and to restrict any starring or cracking of the metal to the limit of the squares or scales struck. The bolts are of high carbon and are extremely hard steel.

Armor plates must often be bent or curved to single or double curvature and sometimes to a warped surface to fit the form of the ship. There are several methods of bending plates. One method employs a cast iron slab of the required form, which is placed on the piston of a hydraulic press. The armor plate is placed face down on this slab, and on top of the plate are laid packing blocks of cast iron, of such sizes and shapes as to conform to the required curve. These blocks take against the upper table of the press, when the piston is forced up, and the hot plate is thus dished to the proper form.

In the French method of bending, an anvil or bed plate of the required curve is used, and the armor plate is forced to take the curve by being hammered all over its upper surface with a specially designed steam hammer.

The edges of the plate are trimmed by large, powerful slotting machines or circular saws; the latter, however, operate in exactly the same manner as a slotter, except that there is no return motion to the tool. Each tooth of the saw is but a slotting tool, and these teeth are, by screws, rendered capable of being nicely adjusted in the circumference of the saw.

The plates are fastened to the hulls and backing by heavy bolts, varying in size according to the weight of the individual plate. For the 6,000 ton armored ships, these bolts are from 2.75 to 3.1 inches in diameter and from 18.45 to 23 inches in length. They are tapped two or three inches into the armor and do not go through the plate. They pass through wrought iron tubes in the backing and set up with cups, washers and nuts against the inner skin of the ship.

At steel works where plates for our new navy are being manufactured, there are inspectors who look after the government's interests. Officers of the navy are detailed for this work, and their duty is to watch the manufacture of plates through each part of the process and to see that the conditions of the specifications and contract are complied with.

The inspection and testing of armor plates consists in examining them for pits, scales, laminations, forging cracks, etc., in determining the chemical analysis of specimens taken from different parts, in determining the physical qualities of specimens taken longitudinally and transversely, and the ballistic test. Specifications for these different tests are constantly undergoing change, and it would be impossible to state, with exactness, what the requirements are or will be in the near future. The ballistic test is the important one, and is made by taking one plate of a group and subjecting it to the fire of a suitable gun. The other tests are simply to insure, as far as practicable, that all the other plates of the group are similar to and are capable of standing as severe a ballistic test as the test plate.

The following will give an idea of the ballistic test as prescribed by the Bureau of Ordnance, Navy Department. The test plate, irrespective of its thickness, is to be backed by thirty-six inches of oak or other substantial wood. Near the middle region of the plate an equilateral triangle will be marked, each side of which will be three and one-half calibers long. The lower side of the triangle will be horizontal. Three shots will be fired, the points of impact being as near as possible the extremities of the triangle. The velocity of the shot will be such as to give the projectile sufficient energy to just pass through a wrought iron plate of equal thickness to the test plate, and through its wood backing. The velocity is calculated by the Gavre formula:

a
V² = --- { 3507 E² × 2265464 e^{1.4} }
w

[TEX: V^2 = \frac{a}{w} \{ 3507 \ E^2 \times 2265464 \ e^{1.4} \}]

V = the velocity of the projectile in feet per second.
a = the diameter of the projectile in inches.
w = the weight of the projectile in pounds.
E = the thickness of the backing in inches.
e = the thickness of the plate in inches.

Using the above formula we can make out a table as follows:

-------+-------+-------------+-------+-------+------+---------+ Plate. |Backi'g| Gun, service| w, | a, | V. | Energy, | Inches.|Inches.| shot. |Pounds.|Inches.| f. 8.| Impact. | | | | | | | f. tons.| -------+-------+-------------+-------+-------+------+---------+ 6 | 36 | 6" B.L.R. | 100 | 5.96 | 1389 | 1337 | 7 | 36 | 6" " | 100 | 5.96 | 1528 | 1619 | 8 | 36 | 8" " | 250 | 7.96 | 1213 | 2550 | 9 | 36 | 8" " | 250 | 7.96 | 1308 | 2966 | 10 | 36 | 8" " | 250 | 7.96 | 1399 | 3390 | 11 | 36 | 8" " | 250 | 7.96 | 1489 | 3839 | 12 | 36 | 10" " | 500 | 9.96 | 1247 | 5386 | 13 | 36 | 10" " | 500 | 9.96 | 1315 | 5987 | 14 | 36 | 10" " | 500 | 9.96 | 1381 | 6608 | 15 | 36 | 12" " | 850 | 11.96 | 1215 | 8699 | 16 | 36 | 12" " | 850 | 11.96 | 1269 | 9710 | 17 | 36 | 12" " | 850 | 11.96 | 1332 | 10454 | 18 | 36 | 12" " | 850 | 11.96 | 1374 | 11124 | 19 | 36 | 12" " | 850 | 11.96 | 1425 | 11965 | 20 | 36 | 12" " | 850 | 11.96 | 1476 | 12837 | -------+-------+-------------+-------+-------+------+---------+

No projectile or fragment of the plate or projectile must get wholly through the plate and backing. The plate must not break up or give such cracks as to expose the backing, previous to the third shot.

The penetration of projectiles of different forms into various styles of armor has been very thoroughly studied and many attempts have been made to bring the subject down to mathematical formulæ. These formulæ are based on several suppositions, and agree very closely with results obtained in actual experiments, but there are so many varying conditions that it is extremely doubtful if any formulæ will ever be written that will properly express the penetration.

Many different forms have been given to the heads of projectiles, as flat, ogival, hemispherical, conoidal, parabolic, blunt trifaced, etc.

The flat headed projectile has the shape of a right cylinder, and acts like a punch, driving the material of the armor plate in front of it. These projectiles are especially valuable when firing at oblique armor, for they will bite or cut into the armor when striking at an angle of thirty degrees.

The ogival head acts more as a wedge, pushing the metal aside, and generally will give more penetration in thick solid plates than the flat headed projectile. The ogival head is usually designed by using a radius of two calibers.

The hemispherical, conoidal, parabolic and blunt trifaced all give more or less of the wedging effect. The blunt trifaced has all the good qualities of the ogival of two calibers. It bites at a slightly less angle, and the three faces start cracks radiating from the point of impact.

Forged steel is the best material for armor-piercing projectiles, but many are made of chilled cast iron, on account of its great hardness and cheapness.

The best weight for a projectile is found by the formula

w = d³ (0.45 to 0.5)

w being the weight in pounds, d the diameter in inches and 0.45 to 0.5 having been determined by experiment.

With a light projectile we get a flat trajectory, and accuracy at short ranges is increased. With a heavy projectile the resistance of the air has less effect and the projectile is advantageously employed at long ranges.

In the following formulæ, used in calculating the penetration of projectiles in rolled iron armor,

g = the force of gravity.
w = the weight of projectile in pounds.
d = the diameter of projectile in inches.
v = the striking velocity in feet per second.
P = the penetration in inches.

Major Noble, R.A., gives

_________________
1.6 / w v²
P = /\ / ----------------
\/ [pi] g d 11334.4

[TEX: P = \sqrt[1.6]{\frac{w \ v^2}{\pi \ g \ d \ 11334.4}}]

U.S. Naval Ordnance Proving Ground uses

________________
2.035/ w v²
P = /\ / ---------------
\/ [pi] g d 3852.8

[TEX: P = \sqrt[2.035]{\frac{w \ v^2}{\pi \ g \ d \ 3852.8}}]

Col. Maitland gives

w v²
P = ------------
g d² 16654.4

[TEX: P = \frac{w \ v^2}{g \ d^2 \ 16654.4}]

Maitland's latest formula, now used in England, is

_
v /w
P = ----- \/ - - 0.14 d
608.3 d

[TEX: P = \frac{v}{608.3} \sqrt{\frac{w}{d}} - 0.14 \ d]

General Froloff, Russian army, gives

w v
P = ------
d² 576

[TEX: P = \frac{w \ v}{d^2 \ 576}]

for plates less than two and one-half inches thick, and

w v
P = ------ - 1.5
d² 400

[TEX: P = \frac{w \ v}{d^2 \ 400} - 1.5]

for plates more than two and one-half inches thick.

If [theta] be the angle between the path of the projectile and the face of the plate, then v in the above formulæ becomes v sin [theta].

When we come to back the plates, their power to resist penetration becomes greater, and our formula changes. The Gavre formula, given above, is used to determine the velocity necessary for a projectile to pass entirely through an iron plate and its wood backing.

Compound and steel armor are said to give about 29 per cent. more resisting power than wrought iron, but in one experiment at the proving ground, at Annapolis, a compound plate gave over 50 per cent. more resisting power than wrought iron.

The Italian government, after most expensive and elaborate comparative tests, has decided in favor of the Creusot or Schneider all-steel plates, and has established a plant for their manufacture at Terni, near Rome.

The French use both steel and compound plates; the Russians, compound; the Germans, compound; the Swedes and Danes use both. Spain has adopted and accepted the Creusot plate for its new formidable armored vessel, the Pelayo; and China too has recently become a purchaser of Creusot plates.

Certain general rules may be laid down for attacking armor. If the armor is iron, it is useless to attack with projectiles having less than 1,000 feet striking velocity for each caliber in thickness of plate. It is unadvisable to fire steel or chilled iron filled shells at thick armor, unless a normal hit can be made. When perforation is to be attempted, steel-forged armor-piercing shells, unfilled, should be used. They may be filled if the guns are of great power as compared to the armor. Steel and compound armor are not likely to be pierced by a single blow, but continued hammering may break up the plate, and that with comparatively low-powered guns.

Wrought iron must be perforated, and hard armor, compound or steel, must be broken up. Against wrought iron plates the projectile may be made of chilled cast iron, but hard armor exacts for its penetration or destruction the use of steel, forged and tempered. Against unarmored ships, and against unarmored portions of ironclads, the value of rapid-firing guns, especially those of large caliber, can hardly be overestimated.

The relative value of steel and compound armor is much debated, and at present the rivalry is great, but the weight of evidence and opinion seems to favor the all-steel plate. The hard face of a compound plate is supposed to break up the projectile, that is, make the projectile expend its energy on itself rather than upon the plate, and the backing of wrought iron is, by its greater ductility, to prevent the destruction of the plate. It seems probable that these two systems will approach each other as the development goes on. An alloy of nickel and steel is now attracting attention and bids fair to give very good results.

The problem to be solved, as far as naval armor is concerned, is to get the greatest amount of protection with the least possible weight and volume, and this reduction of weight and volume must be accomplished, in the main, by reducing the thickness of the plates by increasing the resisting power of the material. In the compound plate great surface hardness is readily and safely attained, but it has not yet been definitely determined what the proper proportionate thickness of iron and steel is.

A considerable thickness of steel is necessary to aid, by its stiffness, in preventing the very ductile iron from giving back to such an extent as to distort the steel face and thus tear or separate the parts of the plate. The ductile iron gives a very low resisting power, its duty being to hold the steel face up to its work. If now we substitute a soft steel plate in the place of the ductile iron, we will get greater resisting power, but our compound plate then becomes virtually an all-steel one, only differing in process of manufacture. The greatest faults of the compound plate are the imperfect welding of the parts and the lack of solidity of the iron. When fired at, the surface has a tendency to chip.

In the all-steel plate we have the greatest resisting power throughout, but there are manufacturing difficulties, and surface hardness equal to that of the compound plate has not been obtained. The manufacturing difficulties are being gradually overcome, and artillerists are in high hopes that the requisite surface hardness will soon be obtained.

The following may be stated as well proved:

1. That steel armor promises to replace both iron and compound.

2. That projectiles designed for the piercing of hard armor must be made of steel.

3. That the larger the plate, the better it is able to absorb the energy of impact without injury to itself.

4. That the backing must be as rigid as possible.

* * * * *

[FROM ENGINEERING.]

THE COMPRESSED AIR SYSTEM OF PARIS.

The demand for compressed air as a motive power is constantly increasing in Paris; the company, according to its official reports, is financially prosperous, and it seems difficult to understand how it should continue as an actively going concern, unless it at all events paid its way. The central station of St. Fargeau, originally started on modest lines, for maintaining a uniform time by pneumatic pressure throughout Paris, has grown rapidly to very large proportions, though it has never been able to supply the demand made on it for power; and at the present time a second and still larger station is being constructed in another part of Paris. We confess that we do not understand why such large sums of money should continue to be spent if the enterprise is not commercially a sound one, nor how men of such eminence in the scientific world as Professor Riedler should, without hesitation, risk their reputation on the correctness of the system, if it were the idle dream of an enthusiast, as many persons--chiefly those interested in electric transmission--have declared it to be.

In describing the developments that have taken place during the last two years, we shall confine ourselves entirely to the details of a report recently made on the subject by Professor Riedler. As soon as it became evident that a very largely increased installation was necessary, it was determined that the new central station should be as free as possible from the defects of the first one. These defects, which were the natural results of the somewhat hasty development of an experimental system, were of several kinds. In the first place, so large a growth had not been contemplated, and the extensions were made more or less piecemeal, instead of being on a regular plan; the location of the central station itself was very unfavorable, both as regards the facilities for obtaining coal and other supplies; the cost of water was excessive, and the amount available, inadequate.

This evil was partly remedied by elaborate arrangements for cooling the injection water so that it could be repeatedly used, a device costly and ineffective, and resulting in extravagant working, to say nothing of the high charges made by the Paris company for supplying water. To these drawbacks had to be added others of an even more serious character. The engines first laid down were not economical, and the compressors employed gave but a very inferior result; with each extension of the plant, the efficiency of both engines and compressors was increased, the most satisfactory, we believe, having been those supplied by the Societe Cockerill, and one of which was exhibited at the Paris exhibition in 1889. Still it was clearly recognized that much better results were possible, results which Professor Riedler claims have been attained and which will be embodied in the new installation now in progress.

This central station is located on the left bank of the Seine, close to the fortifications, opposite Vincennes and not far from the terminal stations of the Orleans and the Paris, Lyons, and Mediterranean Railways; the plan, Fig. 1, shows the position. The works are separated from the river by the quay, over which a bridge will be constructed for the transfer of coal from the landing stages belonging to the company, into the works; as will be readily seen from the plan, it would be quite easy to run junction lines to the two adjacent railways, but with all the advantages given by water carriage, it was considered unnecessary to incur the expense. The river also affords a constant and unlimited water supply, so that none of the difficulties existing at St. Fargeau Station in imperfect condensation and cooling will be met with.

The new installation, called the Central Station of the Quai de la Gare, is laid out on a very large scale, the total generating energy provided for being no less than 24,000 horse power; of this it is intended that 8,000 horse power will be in operation this year, and an extension of 10,000 horsepower in 1892; the power now in course of completion comprises four engines of 2,000 horse power each. Four batteries of boilers will provide steam for these engines. Figs. 2, 3, and 4 show the first section of the installation now in progress; the four groups of engines (three-cylinder condensing) are shown at 1, 2, 3, and 4; the four groups of boilers ranged behind them at F, F; the feed water heaters belonging to each group at V V.

The end of the building abuts against the Seine, and the position of the water conduits for inlet and discharge are indicated at C and A respectively. The installation, when completed, will include very extensive arrangements for transporting and storing coal, and the interior of the boiler houses will be furnished with an overhead system of rails and carriers for handling the coal automatically, as far as possible. All the principal mains and steam pipes are made in duplicate, not only for greater security, but in order that each set of engines and boilers may be connected interchangeably without delay. The Seine supplies an ample quantity of water, but not in a condition either for feeding the boilers, for condensation, or for the air compressors.