that, who cares? If we win, we can take care of our ships, and if we lose what does it matter? We would lose the ships too. Remember, we are speaking of a fight to the finish.

The same reasoning as before holds here too. A little more or a little less normal fuel does not affect fighting power seriously. It might even be said that, since our present normal supply is of the order of five per cent of the displacement, there is no danger of loss of fighting power, no matter what is done, so long as excessive increase is not contemplated.

Until the battle becomes imminent, there are various ways of maintaining the battleships, but they do not affect fighting power. We have contemplated fueling at sea, and we have prepared to carry emergency fuel. Either can fulfill the object, and neither affects fighting power.

Having fixed on the policy that in the battleship a definite, large percentage, as large as possible, of the displacement, should be given to fighting power, and no reduction, for secondary purposes, tolerated, we might at some length discuss the proper division of the displacement, or cost, among the elements of fighting power.

Before undertaking this it will be well again to state that there is no limit to the future development in size, nor in speed, nor in any other quality, but these qualities should retain their proper relative magnitudes. There is bound to be increased size. It is just natural development of the species. We now believe that the Panama Canal is a final limit. It is not, nor will any other present or future hindrance be. When we come to the point where the canal is too small for the ships we want to build, we will ignore the canal, or we will enlarge it. Development is as sure as fate. And with development of size goes development of speed, of gun-power, of armor, of everything. Increased size always means increase of some one factor in the ship, and as long as the different factors remain at about their proper relative values the fighting power increases.

Fighting power includes offence and defence. It might be stretched to include most of the general characteristics of the hull, such as length, beam, metacentric height, etc. The length and metacentric height are the most important of these, from the point of view of fighting.

A battleship should not be longer than necessary, and we should even be prepared to accept some reduction in the efficiency of propulsion, to get a shorter ship. Great length means heavy hull, extra armor and decks, greater target, higher center of gravity, and wetter decks in a head sea. It also requires more weight for torpedo protection. The present tendency, due to the canal limitations, is to increase displacement by increasing length. This has the effect of reducing fighting efficiency.

Metacentric height has an important effect on the efficiency of the ship. There should be enough to prevent great heel when firing a salvo, or turning at speed, but not so much as unduly to cut down the period of rolling. A short rolling period always means much rolling. If a period of about twenty seconds for a complete roll could be obtained, rolling would be almost eliminated, except in the severest weather. Metacentric height is also a factor in the range of stability.

Coming now to the factors more directly concerned in fighting, we will start with torpedo protection. A battleship should be so protected that she can stand a torpedo explosion in any part of the hull without disablement. There is no present difficulty, with our present type, in nearly accomplishing this, and the ship will stand torpedoing several times, in different locations, without great loss of fighting efficiency. The only important parts that cannot now be protected are the rudder, propellers, and shafting, and a hit in their vicinity may mean loss of motive power. It might be possible to improve this situation by a different design of hull, but probably with a loss of propelling efficiency, and whether it is worth while is open to question. There is no doubt, however, that torpedo protection should be thoroughly carried out, and large size is for this object a distinct advantage.

Considering now the armor protection, we can establish it at once as a fact that if the vessel is intended to fight only at long range, as seems to be a favorite idea, there is no need for the side armor, since it becomes a vanishingly small part of the target. This is, however, far from the actual fact. The vessel may have to do some fighting at long range, but how much of her scanty allowance of ammunition can she "afford to throw away at long range? Remember that when her ammunition is gone her fighting power disappears.

History shows that never yet has a decisive action been fought between nearly equal forces at long range, and the term is here used in a relative sense, each period having its own definition of long range. The reason for fighting a decisive battle at close range is not hard to find. The offence is a matter of how many effective hits can be made. Therefore we may expect the same thing to come to pass. Why should we attempt to decide a battle at long range? Cannot much greater damage be done at short range? Of course we will suffer ourselves, but we cannot expect to fight without receiving damage. A successful fighter always takes punishment, and is able to withstand it better than his opponent. That is the whole thing in a nutshell.

The battleship must therefore be designed primarily to fight at short range, and secondarily at long range. High spotting stations, kite balloons, elaborate fire control instruments, are all good in their way, and probably useful, and should be developed. The ship may need them at long range. But these things will all be swept away when we come to the really serious part of the fighting. Then we will want thick armor, powerful guns, simple and rugged turret machinery, and plenty of ammunition.

Our ships are not now armored in such a way that they can stand up for a long time at short range. They may do very well at 12,000 yards, but that is not short range. The range can be considered short when say more than fifty per cent of hits can be made, by the average ship. It will also be necessary to be somewhat independent of range-finders and range keeping. These things we cannot expect to survive in a decisive battle. Perhaps we might, for the present, define short range as 5000 yards. If, at this distance, our ships were fighting broadside to broadside, or, as they used to say, yardarm to yardarm, the fifty per cent of hits would probably all be effective, and the damage. would be great. We should fit armor that will prevent this.

Inclined armor will go a long way to accomplish this. This is not a new idea. It was used in the Merrimac of the sixties. It is used now in our turret fronts. We can incline all our side armor inward at say 45 degrees, and it would be proof at very short range against the heaviest modern projectile, if of the thicknesses now used. It is a delicate problem in stability for the naval architect, but it can be done.

Our barbettes will need greater thickness than they have at present, and should be made considerably heavier. It might be possible to compensate somewhat for this by reducing their heights. This will be touched on later.

Protective decks, of the thickness we now use, are good enough at short range, but as the ship must be exposed to gun fire at long range, provision should be made for it. There is no way of escaping serious damage at long range, if the enemy obtains hits. If the protective deck were made narrower, as a result of inclining the side armor inward, we could afford to make it thicker, but not otherwise. The decks now are so heavy, and their effect on stability so marked, that we could hardly afford to put any more weight into them. The protective plating on our decks now amounts to about 10 per cent of the displacement.

The secondary battery is intended only for protection against torpedo craft. Yet we see a great number of men on our battleships that are used only to man these small guns. The number is quite comparable to that used for the main battery. This does not seem logical. If so many fighting men are really necessary for an object that is quite secondary, the natural inference is that the existence of the type is rather precarious. There is need for depending more on passive torpedo protection, just as against gun fire we depend more on passive armor protection, and less on "a well-directed fire from our own guns." That this is quite logical is indicated by the fact that in our latest ships the torpedo protection is quite as efficient against torpedoes as the armor is against heavy projectiles. A weapon of destruction is intended for the offensive, and not primarily for the defensive. There is of course no objection to smashing a destroyer if we get a chance, though, as has been said before, it is cheaper to have smaller types to look out for that. The secondary battery and its crews should be much reduced.

It is hardly necessary to mention torpedo outfits. It is coming more and more to be recognized that these have no place. on a battleship, and the outfits we now fit are considerably reduced from what we at one time fitted. It is to be hoped that at a date not very distant they will be entirely eliminated.

Now, last of all, we come to a consideration of the main battery, the primary weapon of the battleship. There has been so much argument about this matter as to leave us entirely con

fused; there have been so many different arrangements, numbers, and types, that it is quite certain no logic has governed.

The British have come to the belief that not more than eight heavy guns should be mounted on one ship, but they have been building moderate sized ships, and for a special field of action. It is doubtful whether we could take their example for our own guidance, especially when we build larger ships.

We all have a feeling, which is founded on subconscious conviction, as we have not yet been able to justify it by logic, that a larger ship should have a greater battery; in fact, that we are not warranted in building a larger ship unless we increase the battery in proportion. We can easily justify this idea on the basis of economy. There should certainly be at least a standard percentage, for the offensive, in every ton of displacement, and progress in development would seem to call for an ever increasing percentage. Should not every successive phase in development increase the fighting power of the men on board? Hence the distrust with which we regard a design which doubles the displacement, for purposes of speed, and leaves the battery without such increase.

We may therefore feel justified if, on our larger displacements, we carry a heavier battery than the British have placed on their standard battleships. The difficulty is, however, in making proper use of the larger weight. Shall we use larger caliber guns, or more guns, or longer and more powerful guns?

We have mounted up to six turrets on one ship, and have lost some of our fear of interference between the different turrets. A large number of turrets requires, however, a heavy weight of armor for the individual barbettes, and more is used up in the turrets than if we used fewer turrets with a greater number of guns per turret. The need for a greater amount of armor protection on our ships, so as to enable them to fight at closer range, may force our hand in this respect, and compel the adoption or retention of multiple gun turrets, much as we might prefer not to do so, for reasons of putting all our eggs in one basket. This latter point becomes more serious at short range, and will require more careful consideration.

There would be a certain amount of armor weight available for turrets and barbettes. If we used this weight for two-gun turrets, we would have thinner armor than if we used it for three