sea and in the broad light of the sun. To gain the weather-gauge, to double on the van division and throw the rear or lee division out of action, were favorite manoeuvres in the days of the old sailing vessels, and the brilliant success that has followed such strategy in the past still has its effect upon the tactics of to-day; but when we consider that hereafter fleets will go into action with a speed of at least ten knots, and that the right vessel of the largest fleet will be within easy range of the guns of the vessel on the left, it seems idle to talk of feinting on one wing and attacking the other, or to expect more than momentary advantage from concentration, at the expense, perhaps, of confusion and disorder in your own fleet. And here I will stop a moment to consider the great stress that is laid upon concentration; this seems to me a fallacy that we are most likely to fall into by following too closely the military idea. In the English Naval Prize Essay of 1880 (a very valuable paper), concentration is one of the main points insisted upon. In one instance given, of an imaginary naval battle of the future, B's fleet advances in column of vessels to attack A's fleet in line abreast (the favored formation). As soon as A discovers that it is B's intention to pierce his line in his present formation, he makes signal to his fleet, in accordance with a previous arrangement, and at the proper distance, determined by sextant, and after due consideration of the table of helm angles, he hauls down the signal, and the right and left wings simultaneously put their helms respectively to starboard and port. The result theoretically is the ramming of B's vessels, which, with a fatuity to be encountered only on paper, continue on their way; but, as the essayist in another place states that the difference between ramming and being rammed is only a ship's length, it seems more than probable that such a plan would bring the two wings of A's fleet together in dangerous, if not disastrous, confusion, and there would be as much likelihood of their ramming each other as of their ramming the enemy. A careful reading of a sextant, or an accurate application of the table of helm angles, at such a moment is hardly to be counted upon. I do not mean to discredit the usefulness of such tables, nor do I think there is any danger that we can know too much about such matters, or have too many scientific or mechanical aids to judgment. But it seems to me that the true time for the application of all these aids is during the evolutionary drills, all of which are simply to educate the eye and train the judgment of the captain for the time when he must throw aside such aids and act instantaneously, and without hesitation, upon the decision of the moment.

It is not, however, surprising that there should be a wide difference of opinion on the question of naval tactics; the whole subject is purely tentative. There has been only one combat between fleets since the introduction of steam-the battle of Lissa. In that action the Austrians in line abreast, attacked the Italians in line ahead; the latter appeared to have no especial plan of battle, while that of the Austrians seemed, as has been said by an English writer on naval tactics, very similar to those of Donnybrook Fair, "whenever you see a head, hit it." As far as tactical lessons go, there is little to be learned from Lissa, except perhaps the fact that to ram successfully, vessels must have speed and use it. The Austrians pierced the Italian line, but, strange to say, effected no damage in doing so. After this manœuvre the action became a general mêlée. The Austrian vessels, without any general order, devoted themselves principally to attempts to ram, some of them even securing their broadside batteries. From lack of speed and an impression that it was necessary to slow down or stop the engines before collision, most of these attempts were failures, the sinking of the Re d'Italia being the only marked success.

Out of this victory of the Austrians, however, has grown up a strong sentiment in favor of attacking in line abreast, some writers favoring the single line, others the double line, still more, perhaps, the double line indented (the formation prescribed in our tactics), while some propose even three lines, as in the" Carré Naval" of Admiral Bouet Willaumez, where a squadron of nine vessels is formed in three lines, three abreast, and this square is considered the tactical unit, certainly the perfection of concentration; but it seems to me impossible to imagine that such a fleet can be carried into action without being completely broken up and disorganized by the first shock. An accident to any one of the vessels of the first and second lines must cause those in rear to run into her, or, in their endeavors to avoid collision, to break up the square. To reform it in action would seem impossible. Remembering that the modern ironclad is generally a vessel of from six to nine thousand tons displacement, with ram bow, and a maintained speed of from ten to twelve knots, that ships in close order are only one hundred fathoms apart, it would seem a fundamental rule for all orders of battle that no vessel of a fleet should be in the water of another. Certainly, in the moment of going into action a commanding officer should not be required to keep his eyes glued to the speed-ball of the next ahead, and his nerves at a high tension to avoid a collision with his leader or being rammed by his next

astern.

At the present time there are two distinct systems of tactics in use: Ist, that in which the single vessel is made the tactical unit, as it is with us; 2d, the group system. The latter originated with the French, and has been adopted in part and with modifications by the English. The idea is the subdivision of the fleet into groups, generally of three, which shall be manoeuvred as units, the vessels of each group maintaining the same relations toward each other through all evolutions, Nos. 2 and 3 confining their attention simply to maintaining their positions relative to the group leader, who is responsible for the group. It is claimed that this gives greater mobility to a fleet, greater capacity for concentrating and dispersing, and especially greater facility for reforming during the confusion of battle. It seems to me such claims are well founded. At the same time there has been much difficulty experienced in handling these groups as single vessels, at least by the English, for, according to Captain Freemantle (Naval Prize Essay, 1880), it is prescribed in their manual that "when ships are in group formation, any signal which entails an alteration of course is to be acted upon by all the ships in each group together, unless otherwise ordered, and group formation is not to be resumed until signal to that effect is made." This certainly would seem to show that the groups, as arranged by the English, are not handily manœuvred. As far as I know, there has been little if any similar complaint from the French; this is due probably to the difference in formation of groups adopted, and their systems of manoeuvring. The English group is in the form of a scalene triangle (Fig. 1), the French in the form of an equilateral triangle with the apex toward the enemy (Fig. 2).

In manoeuvring their groups, the English have endeavored to handle Nos. 2 and 3 of each group by helm and variations of speed, so that throughout all evolutions they would maintain their position relative to No. 1. The difficulty of this is apparent; to accomplish it

would require the most careful management and the nicest calculations throughout The French on the contrary have gone to the opposite extreme; Nos. 2 and 3 of their groups are required simply to gain their positions at the close of each evolution as best they can. As these orders apply to two-thirds of the vessels, it would seem as if a fleet must be very thoroughly exercised before evolutions can be performed without indecision, hesitation, and delay; unless well accustomed to work together, one commanding officer will naturally wait to see in what direction his consort will move before putting over his own helm. The French group formation or peloton d'attaque is criticized on the ground that Nos. 2 and 3 mutually mask each other's fire, while the scalene triangle, it is claimed, gives free use to each vessel of the group of her guns, ram, and torpedoes. In endeavoring to obtain all this, however, the English have sacrificed the mobility of the group, and consequently the very principle of the system, which makes it imperative that the groups should be handy and capable of being manœuvred as single vessels.

Firmly believing in the fundamental principle at the bottom of the group system, I would suggest that the groups consist of two vessels; in other words, that the fleet should be divided into pairs, mutually supporting and depending upon one another, the No. 1 of each pair in charge of, and responsible for, the movements of the two. There would be required only two original formations for the pairs as follows:

In either case the position of No. 2 would be four points abaft the beam, distant two cable's-lengths. All evolutions should be performed by the pairs without change of form unless otherwise signalled; No. 1 of each pair manoeuvring in accordance with present tactics for single vessels, No. 2 placing herself in the proper position from No. 1 by the shortest possible route and without reference to any other vessel. This it seems to me will secure the greatest possible mobility, and at the same time, as the groups are entirely distinct, and there is no

provocation for No. 2 of one group to get in the way of No. 2 of another, this would obviate the indecision and danger likely to occur in following the same rule with the French groups of three. At the same time it seems to me as contrary to the true principle of the group system to require No. 2, in all evolutions, to move on a prescribed curve, as it would be to the true principles of military tactics to require a skirmish line in broken country to march with the regularity and precision of a battalion on review.

The value of this subdivision of a fleet into pairs seems to me capable of proof in many ways. It simplifies manoeuvres, in that onehalf of the fleet have no other duties to perform than to maintain their position in relation each to his own leader. Each leader of a pair will advance to the charge and use his ram or torpedo with more confidence, knowing that his consort is close at his back to support him if disabled, or aid him in overwhelming his adversary if successful. The reforming a fleet after once passing through the enemy's line will certainly be much simpler and shorter work if the vessels first group themselves in pairs.

Let us suppose two fleets of twelve vessels each advancing to the attack (Fig. 3), A in line abreast by pairs, B in two lines. It will be perceived that A's formation is virtually the same as the indented line or the "order of battle in two lines" of our own tactics; but to my mind the strength of the formation and the mobility of the fleet are greatly increased by its subdivision into pairs.