tion stops the diverging torpedo; a few lines stretched between booms would change its direction and preserve the vessel from all danger. This reasoning, supported by the experiences of the Turco-Russian war, leads us to completely reject the diverging torpedo for boat use.

The actions which we have cited in the beginning of this essay give us three complete successes in ten attacks made with spar torpedoes. But, as we have seen, the torpedo cutters used by the Americans were very defective, and the actions at Roustchouk and Nicopoli were carried on under very unfavorable conditions for the assailants. The destruction of the Housatonic and of the Seïfi shows that the spar torpedo is, at night, a terribly efficient arm against vessels at anchor unprotected by an exterior belt of obstructions; that of the Albemarle, and even the action of Sulina prove that sometimes even that defense is insufficient. We know, in fact, that on the night of June 10, 1877, the Russian launches, commanded by Lieutenants Rojdestvensky and Poustchine, pushed over the obstructions which surrounded the Idjalieh. If their torpedoes did not destroy that vessel, it was because they were struck by shot, or because they struck some obstacle placed by the Turks inside of the obstructions. These obstructions were, probably, the primary cause of the Russian failure, delaying the boats for some minutes under the fire of the Idjalieh and producing amongst them a momentary demoralization which prevented their handling their torpedoes with the necessary coolness. Again, no matter how quickly the spar is worked, it will always be most difficult to submerge it in the short space of time occupied by the passage from the obstruction to the vessel. In the Nicopoli affair also, the nets and booms presented an unsurpassable obstruction to Lieutenant Niloff's boat. Relative weakness against vessels defended by exterior obstructions is the principal objection that can be made against the spar torpedo.

The Russians attribute the premature explosion of the two torpedoes which they used against the Idjalieh to the shock giyen by a projectile; while official reports from Roustchouk and Nicopoli state that the torpedoes of the Choutka and Mina could not be ignited, because their conductors had been shot away. An explanation for these four failures, may probably be found in the faulty construction of the torpedoes themselves, in the hurry of the moment, or in some mismanagement committed under the enemy's fire. In fact experiments prove that the shock of bullets or canister balls can only ignite a torpedo when the cap is directly struck, furthermore two wire conductors are a very small target for small-arms; nothing prevents even these from

being protected in the spar of the torpedo. If we place entire confidence in the Russian reports, a truly extraordinary chance caused, on four different occasions, accidents which the fallibility of man, even of the ablest and bravest, explains much more readily. We cannot therefore consider the accidents to which the enemy's fire exposes spar torpedoes as a serious objection to their use.

Unfortunately we have only been able to give two actions in which the assailants used the Whitehead torpedo; one of these was a failure, the other a complete success for the Russians.

What were the real facts on the occasion of the second action off Batoum? Commander Chardonneau thinks that the Whitehead of the Tchesme struck a rock; but, since that torpedo was fired at a distance of only sixty metres from the Turkish ironclad, it would be necessary to suppose that that vessel must have been very close herself to the danger; the least change in her position would have put her on it. It is not probable that her captain would have chosen, in the vast harbor of Batoum, such a perilous anchorage. Shall we suppose that the Russian torpedoes met some sub-marine obstructions placed by the enemy? The Turks might very possibly have attempted some such system of defense, but no document has ever been adduced to support this theory.

It therefore becomes necessary to look elsewhere for an explanation of the Russian want of success. Now, when they made their attack, the night was very dark; therefore they might have made considerable error in regard to the distance which separated them from the enemy. All seamen know that on a dark night a large vessel often appears immeasurably larger in every way than she really is. The Russian officers state that they were able to see, on the water, the track of the torpedo from the Tchesme to off the beam of the Turkish vessel where it exploded, but the arrangement of the mechanism of the Whitehead is such a delicate operation, that, the trajectory of the torpedo does not always follow a straight course. It is, therefore, very possible that the Russians fired their torpedoes at a greater range than sixty metres from the Turkish ironclad and that the latter describing a curve, passed astern of the vessel striking the rocks, off the neighboring shore. This supposition seems still more admissible from the fact that the Russian officers state in their reports that the Sinope's Whitehead took the direction of the stern of the Turk. The arrangements fitted for firing the torpedoes from the Tchesme and Sinope were, moreover, probably defective; a movable tube under the keel of the cutter, or the expedient of an auxiliary raft, does not seem a very convenient method of

pointing the Whitehead. Whether the Russians modified these arrangements before the third affair at Batoum, is not known, but the rapidity with which the Turkish vessel sank leads us to believe that the Whiteheads, fired on this occasion from the Tchesme and Sinope, both reached the object against which they were directed.

On the Danube, the Russians did not attempt to use the Whitehead, as any body of water with a strong current or tide is very unfavorable to the action of this torpedo, both boats and torpedoes being subjected to many exterior influences which go to destroy all accuracy of fire, while even a rough sea makes the firing of a Whitehead a delicate and sometimes impossible operation in an open roadstead; the differ ence in density between salt and fresh water renders new adjustments necessary in ascending rivers. Therefore, deviations arising from an incorrect appreciation of distance, a faulty regulation of the rudder, a false trajectory-caused by exterior influences acting on the direction, or immersion of the torpedo, are grave objections inherent to the Whitehead.

The following deductions concerning the employment of spar and Whitehead torpedoes are based upon the experiences from the use of these weapons during the Turco-Russian war.

When a vessel is anchored in a strong current, it is very difficult to establish a belt of floating obstructions about her, or even a defensive system with swinging booms and nets attached to her hull. We have just observed that similar circumstances, especially on a dark night with a rough sea, also render the firing of the Whitehead very uncertain. Spar torpedoes should then be used in preference; the darkness. will moreover allow the launches to avoid, with greater ease, the watchfulness of the guard boats and the sentries of the enemy's vessel. When, on the contrary, the night is not too dark; when the sea is calm and the feebleness of the current has allowed the attacked vessel to surround herself with exterior defenses, the use of the Whitehead presents splendid chances of success which it would be impossible to ask of spartorpedoes. The opportunity of attacking a vessel underway, with torpedo-boats, may occur at night as well as during the day. No action pleads for nor against the Whitehead under these circumstances. But the affair of Nicopoli shows that a vessel, having poor speed, can protect herself efficiently, against the explosion of torpedoes, by means of a system of swinging booms and nets. The use of the Whitehead in this case is clearly indicated. We must remember, however, that the speed of the enemy's vessel adds considerably to the exterior causes

which render the use of the Whitehead from boats very uncertain. In conclusion, until further investigation, it would seem wise to place in the flotillas intended for coast defense some launches armed with spar torpedoes, and others with Whiteheads. Finally, the following simile seems appropriate: the spar torpedo is a dagger which a determined man, at the peril of his own life, strikes into the heart of his enemy which even a solid coat of mail will not protect; the Whitehead torpedo is the perfected projectile, which, being easily fired from a distance, kills the enemy which it strikes, but very often misses its mark and is uselessly lost in the distance.

Up to the present we have only enumerated torpedo-boat combats from the side of the attack; let us see if they offer any useful points in regard to the defence. If the Minnesota, the Wabash, and the Memphis were neither sunk nor seriously injured by the "davids" of the Confederates it was owing to the extreme vigilance of their crews; but the Housatonic, which was also on her guard, as she slipped her chain at the instant of the explosion, was nevertheless destroyed by Lieutenant Dixon's torpedo boat. The Turkish vessels anchored near Matchin and off Soukoum-Kale had guard boats; notwithstanding, the Seïfi was sunk, and the Assar-i-Chefket only escaped because the torpedo did not act properly. It is therefore evident, that, the greatest vigilance is not always a sufficient guarantee against the attack of torpedoboats. All the actions which we have reviewed show the decided insufficiency in such cases of a defense depending entirely on great-gun and small arm fire. On the other hand, if the Albemarle was sunk in spite of her surrounding obstructions, we have the simple rope protection of the Idjalieh frustrating entirely the Russian attack, and also the fact, that the monitor attacked off Nicopoli owed her salvation to the swinging booms and nets attached to her hull. To the use of such systems of defense the objection is raised of the necessity of a vessel threatened by torpedo boats always being ready to get underway; the example of the Housatonic proves however that the last mentioned method is not sufficient to withdraw a vessel from the torpedo of a fast cutter. Would the getting underway of a vessel be impeded by a system of swinging booms and nets placed around the hull? No: because the Turkish monitor, attacked off Nicopoli, used this means of defence, even when she was underway. We cannot lay too much stress on such an example, which demonstrated practically the possibility of defending exteriorly the hull of a vessel under steam, and consequently the necessity of giving serious attention to this sub

ject. We know that at sea, even at slow speed, the use of nets and swinging booms would be impossible; but it will generally occur that a vessel will have to defend herself in this way in rivers, passes and harbors, where the water is generally smooth.

When not provided with swinging booms and nets fastened to the hull, a vessel at anchor can, under certain circumstances, protect herself efficiently by means of obstructions made fast to boats or to buoys anchored around her. It was thus that the Turkish obstructions at Sulina foiled the attempts of the Russian torpedo-boats. The Ottoman ships collected at Batoum were in an excellent position to have followed this method. Their bows being secured by anchors to seaward, and their sterns being moored to the shore, they had every facility for establishing a good line of obstructions some metres from their hulls. It is evident that a vessel at anchor in variable currents and winds, where she would swing about her anchor, could not use this system on account of the great space required to be kept clear.

For these reasons the events of the Turco-Russian war seem to indicate the following methods as most advantageous in protecting ships. against the attack of torpedo-boats, combining them according to circumstances: one watch sleeping at quarters during the night; mitrailleurs and light guns, in as great numbers as possible, loaded and aimed beforehand; the engine with steam up and ready to move, and the cables ready to slip; a system of swinging booms and nets secured to the hull; obstructions similar to those used by the Turks at Sulina; the guard boats at these obstructions to be armed with mitrailleurs or guns; finally, at a distance of at least one hundred metres from each beam, steam launches under way. It would be the duty of these launches to give the alarm, and to rush at full speed against the enemy before he could use the Whiteheads. A vessel protected in this way would not be invincible, but she would have, we think, a great many chances on her side.

Finally, an examination of the actions cited above lead us to the following conclusions:

First, the attacks made by torpedo-boats against ships during the American War of Secession, and the late struggle between Turkey and Russia, took place under such circumstances that their results should be taken into serious consideration; these results appear to establish the following facts. Second, the danger to a torpedo-boat attacking a vessel has been greatly over-estimated. Third, such an attack, even when the assailant makes use of comparatively insignificant means,