Frank Friday Fletcher as admiral.



As an introduction to a description of the latest improvements in torpedoes, a summary will be given of what the torpedo has actually accomplished in war.

This mode of warfare first came into practical use during the American Civil War. At that time and subsequently the spar torpedo was in use on a boat of 6 to 12 knots speed and the boat had to come into actual contact with the ship firing 9 to 10 shots a minute, besides small arm fire. Under these conditions fifteen attacks were made with the spar torpedo. Three attacks were made in broad daylight and twelve at night.

In these attacks six ships were sunk and three were seriously injured.

  • Thirty-three boats were engaged in these fifteen attacks, and the crews amounted to a total of 350 men.
  • Three boats, or about 9%, were lost, and twelve men, or about 3% of those engaged, lost their lives. The casualties to both boats and crews were due almost entirely to causes incident to the attack and only one man was killed by gun fire.

These torpedo attacks with the spar torpedo were made during the American Civil War, the Russian-Turkish War, the Peruvian-Chilean War, and in the war between France and China in Tonquin:- four wars, in which seven nations were engaged.

The success of the spar torpedo brought out the automobile torpedo, and the defense followed with the R. F. (rapid fire) gun.

Exclusive of the war now in progress between Russia and Japan, there have been eight attacks in which the boats came under fire and discharged automobile torpedoes.

  • In these eight attacks eight vessels were sunk:- three armored vessels, three cruisers, a tug boat, and a schooner.
  • Twenty-seven torpedo boats were employed in these eight attacks, but only sixteen discharged their torpedoes and were under fire.
  • Twelve per cent of the boats engaged were lost—one boat by gun fire and one by running aground. Of the crews, making a total of about 500 men, only about 2% were lost.

These torpedo attacks with the automobile torpedo were made during the Russian-Turkish War, the Chilean Revolution,  the Brazilian Revolution, and in the war between China and Japan:- four wars, in which six nations were engaged. Thirty-two torpedoes were discharged and nine of them made hits, sinking eight vessels. This is a percentage of hits of more than 28.

« « «
The histories of the spar torpedo and of the automobile are much alike in results. The percentage of successes or of failures and of losses of boats or of personnel are about the same. With each weapon about thirty boats have been employed in attack, of which about 6% were lost with 2% to 3% of the personnel. The spar torpedo sank six ships and the automobile eight,- in all, fourteen ships sunk and three injured. This is exclusive of the ships recently sunk or injured by torpedoes at Port Arthur.

Since the above results of torpedo fire in action took place, the automobile torpedo has gradually improved in range, speed and accuracy. The torpedo as heretofore issued to our service has a speed of from 26 to 28 knots over a range of 800 yards. Those which are fitted with the Obry steering device are required upon test to pass within 12 yards of the point aimed at 800 yards distance. Those without the steering device are required to come within 24 yards.


From the Quarterly Reports of Torpedo Practice in service, there has been compiled the record of practice runs covering a period of several years.

Of runs without the Obry gear, 46% are recorded as coming within less than 10 yards of the point aimed at; 37% came between 10 yards and 50 yards of [the] point aimed at; 17% were more than 50 yards away.

This does not include the number of cases where the torpedo failed to run. The record shows that only about one-half the practice runs come up to requirements and that about one torpedo in two without a steering device can be counted upon for a hit.

Of runs with the Obry gear, 78% are recorded as coming within less than 10 yards; 14% came between 10 yards and 50 yards; 8% were more than 50 yards away.

This would indicate that with the Obry steering device about three torpedoes out of four could be counted upon for a hit at a distance of between 500 and 800 yards when the torpedo makes a run. But this record does not include failures to run or lost torpedoes. Since January 1903, to date there have been lost twenty-one (21) torpedoes. This is no doubt due to lack of training of the personnel. These results do not show a very high percentage of efficiency as compared to what the weapon is capable [of], particularly when it is considered that in nearly all cases recorded above, either the boat or the target was stationary. It is believed that when the torpedo practice is placed upon the same competitive basis as target practice with guns, the percentage of hits will be . . .

The Whitehead Torpedo started about 30 years ago with a speed of six or seven knots and with fairly reliable runs for 200 yards. Its  speed, range, and accuracy have made successive strides, adding four or five knots of speed and several hundred yards to the range with each new type developed.

At the present time the Bureau of Ordnance is about to introduce two new types into the service,- an 18-inch 5 metre and a 21-inch 5 metre.

The new 18-inch torpedo will be required to give a speed of 35 knots for 1,200 yards, and the 21-inch torpedo will give a speed of 24 knots for 4,000 yards.

The range of the torpedo can be increased by running them at slower speeds. Experiments recently made with the 5 metre, Mark II, torpedo gave a speed of 18 knots for 2,000 yards and of 12 knots for 4,000 yards. Experiments for speed at long ranges are being conducted with the same torpedo fitted with the superheater, which should increase the range by about 50%.

The following ranges and speeds should be obtained with the different types of torpedoes:-

Speed (Knots)

Range (Yards)

18", 5 metre, Mk. II





18", 5 metre, Mk. II with superheater





18" Turbine (superheater and 2250 lbs.)







21" Turbine






5000 to 6000

These increased ranges and speeds with the 18-inch torpedo are obtained by two additions to the source of powder:- one is by increasing the air pressure in the flask from about 1,000 lbs. to 2,250 lbs., or 50% more power,- the other is by the use of the superheater, or burning alcohol in the air flask during the run. This also adds about 50% more power. Mechanically the distinctive features of the new torpedoes will be a turbine engine and the absence of distance gears, locking and other devices, which will tend to make a much simpler mechanism and therefore more reliable. The gyroscope is so fitted that it can be set to make the torpedo turn through any predetermined angle with a hard over helm, thus introducing curved fire with the torpedo. The power for operating the vertical rudders is taken from the main shaft by means of pawl and ratchet under mechanical control of the gyroscope, thus avoiding the delicate air valves of the present steering engine and the possibility of their clogging by grit. These and many other ingenious devices which make this practically a new American torpedo are the creation of Mr. Frank M. Leavitt, the consulting engineer of the Bliss Mfg. Co.

The late improvement of the torpedo is such that we may soon expect to be in possession of weapons that can be effectively used at a range of 3,000 or 4,000 yards, or possibly, with 20 knot speed, 5,000 to 6,000 yards.

With this in view, efforts have been made during the last year to develop a gyroscope that would maintain its directive force for the time necessary to steer the torpedo straight for these greater distances. The Obry steering device and several types of the modified gyroscope give, upon shop test, when in good condition, a directive force of from two to five minutes’ duration. The directive force for this length of time does not change more than but the time would probably be considerably reduced when the gyroscope is subjected to the shock of firing the torpedo and to the vibration during a run.

The Obry gear as issued to the services is a very simple mechanical design, but defects of manufacture have brought it somewhat into disrepute. By correcting these defects and fitting the wheel with ball bearings similar to those that have been in use abroad for some years and correcting other defects the results are more satisfactory. Also, by introducing electric control, a directive force is maintained suitable for a long-range torpedo. The strengthened Obry gear with electrical control as tested at the Torpedo Station will spin for from 30 to 40 minutes and maintain its directive force for fourteen or fifteen minutes with a variation not exceeding 1/2°. As the rotation of the earth during this time is more than 2°, the wheel has to be over-compensated accordingly.

The gyroscopes that will be fitted in torpedoes hereafter will be capable of causing the torpedo to run in a circle upon entering the water, and after turning through a predetermined angle from the line of fire will then run in a straight course. By this means a torpedo may be fired from each beam and after turning through 90° or more, it may strike an object directly ahead or astern or in any intermediate position, according to the angle for which it is set to run. A torpedo boat can thus stand directly for the enemy and without change of course discharge the torpedoes from both bows so that they will run ahead parallel to the keel or converge at some point ahead.

In making use of curved fire with torpedoes, certain conditions have to receive consideration.

Curved fire with the torpedo is obtained by setting the gyroscope valve at an angle before the torpedo is placed in the tube, and this angle cannot be altered until the torpedo is hauled out of the tube. Mechanical means have been devised for setting the valve from the outside of the tube and is being tried and perfected, but this method of adjusting the angle of fire has not yet been perfected for submerged tubes. Experiments were conducted with four torpedoes fitted for curved fire. There was no difficulty in making them turn through any angle with the modified gyroscope and then run straight. No 1 shoes the trajectories of curved fire.

Tactical Diameter.

The radius in which these torpedoes turned with hard over helm varied from 500 feet to 800 feet. This makes the tactical diameter of torpedoes with curved fire greater than that of a torpedo boat and about equal to that of torpedo boat destroyers, which vary from 330 to over 600 yards. Also, if the torpedo is not in adjustment, the turning radius will differ to the right or left.

Loss of Speed.

When turning through 90°, the distance the torpedo travels over an 800 yard course is increased by about 120 yards, or 15%, which is equivalent to a similar loss of speed.

In order to strike an object directly ahead or astern from a fixed broadside tube, the torpedo must be made to turn through an angle varying from 105° to 118°, according to its tactical diameter. In this case, the distance traveled over an 800 yard range will be increased by an amount varying from 126 yards to 218 yards. This is equivalent to reducing the speed of the torpedo from28 knots to 22 knots. In turning through smaller angles, these losses are reduced accordingly, and at 45° would amount to only 1.26 knots (mean).


If a torpedo fitted with a gyroscope is deflected upon striking the water, in a straight run the gyroscope quickly brings it back parallel to the line of fire with a loss of but a few feet of transfer. The deflection from the submerged tube of a battleship at speed is about 12°. This deflection is recovered by the gyroscope, reducing the error at the target to about 16 feet. In curved fire, however, the effect of a deflection upon entering the water is more serious. In turning through 90°, the torpedo, during the time it is recovering from the deflection, is traveling at high speed in a direction at right angles to the target, so that the error of 16 feet in straight fire becomes an error if 155 feet when transferred through 90° curved fire. In a 45° turn, this error of deflection would be increased from 16 feet to only about 120 feet. This deflection is shown in Fig. 1-(b).

While the torpedo is turning under the influence of a hard over helm in curved fire through as much as 90°, it travels from 250 to 350 yards, or more than one-third of its full run, before it comes under the directive influence of the gyroscope.

The Use of Angle Fire.

This will be considered, first, in its connection with training tubes on a torpedo boat, and, second, in connection with fixed submerged tubes of a battleship.

The general method of mounting torpedo tubes upon modern torpedo boats and destroyers is to place two tubes on the central line, permitting and angle of fire from 45° forward of the beam to 45° abaft the beam, and, in many cases, at an even greater train than this. If the enemy appears at any position within this angle of fire, one or both torpedoes may be discharged at once without any more loss of time than that due to training the tube. If, however, the enemy appears within an arc of 45° measured from either side of a direct line ahead, it then becomes necessary to manoeuver the boat by swinging off from the target until the tube points directly at it.

It is contended by some that a preferable method of attack with torpedo boats is to steer directly for the enemy, with the torpedoes so arranged that they can be discharged directly ahead or the torpedoes made to turn through an angle after discharge and then run directly ahead. This method is equivalent to having fixed tubes and commends itself for simplicity. It has the advantage of requiring from the commanding officer a singleness of thought, which has its value.

By reference to No. 2 it will be seen that the chances are even, that the enemy will be sighted more or less than 30° on the bow, and the mean probable position of discovering the enemy will be about 33°. The effect of this upon manoeuvering the boat with the two methods of attack is as follows:- With tubes fitted to train as far forward as 33° on the bow, the chances are even that the boat will not have to be manoeuvered before the torpedoes can be discharged. The chances are also even that the boat will have to be turned through some portion of an angle of 33° away from the enemy in order to discharge the torpedoes. The mean angle of turn in this case will be approximately one-half of 33°, and the probable mean of all angles though which you would have to manoeuver a boat with training tubes before discharging the torpedoes would be about eight degrees.

If, on the other hand, the torpedoes are set for angle fire so as to run directly ahead, there is a dead angle of fire of practically 360°, and the boat will have to be always turned toward the enemy through a greater or less angle in order to discharge the torpedoes. If the enemy is discovered within torpedo range, the boat will have to turn toward him through a probable mean angle of something more than 33° in order to discharge the torpedoes ahead. This method of discharging torpedoes requires the same manoeuvering of the boat as if she were fitted with two fixed bow tubes, and if the enemy is at close range the torpedoes could pass on each side of him.

The effect of these two methods of discharging the torpedoes—straight fire or curved fire—are illustrated by Nos. 3 and 4. In one case the enemy is sighted within torpedo range, and in the other case the enemy is sighted beyond torpedo range.

In the first case, when the enemy is discovered within torpedo range, say 800 yards, it is presumed that both boats will discharge their torpedoes as soon as possible. If the enemy is sighted at a mean position of 33° on the bow, the boat with training tubes would at once turn from the enemy and the torpedoes be discharged direct as soon as the tubes bear. The torpedoes would run at full speed, with no loss of time, and the minimum distance at which the boat approaches would be 738 yards, while the time the boat would be under fire (at a speed of 18 knots) within 800 yards would be 30 seconds. The boat with curved ahead fire would have to turn toward the enemy and approach through a distance of 165 yards before being able to discharge the torpedoes. The torpedoes would run in a curve, with a reduction of speed from 28 knots to 26.1 knots. The boat would have to continue to approach the enemy after discharging the torpedoes and would reach a minimum distance of 498 yards or less, and would be under fire of the guns within 800 yards through a period of 78 seconds, instead of 30 seconds as in case 1.

When the enemy is sighted beyond torpedo range, the conditions are slightly altered. In this case, it may be fair to assume that the boats would head directly for the enemy and discharge their torpedoes the instant they got within range, which, for illustration, we will presume to be on the 800 yard circle. It may be readily conceived that this is the ideal and most favorable condition for the employment of angle fire on torpedo boats.

Under this condition of sighting the enemy, the boat with training tubes would have to put her helm over before reaching the800 yard circle of the enemy and turn through an angle of 40 degrees before the torpedo is fired. The effect as to speed of torpedo; distance from enemy; and time under fire are illustrated by Fig. 4. In each case the curved fire shows a disadvantage.

By having sights fitted on our tubes, so that the torpedo tube may be pointed at any desired angle to the right or left of the enemy, curved fire may be employed by training the tubes. But it will be recognized that this method necessitates the introduction of a number of corrections, such as the distance of the enemy, tactical radius of the torpedo, deflection for speed of boat, loss of speed of the torpedo, %c., all sources of inaccuracy of fire.

Correction for Transfer from Line of Sight.

It is seen from the diagram, Fig. 1, that the transfer of the torpedo from the line of sight varies according to the peculiarities of each torpedo, the angle through which it turns, and the speed of the boat. This can be allowed for by setting the gyroscope at a different angle in each torpedo, or by setting the torpedo tubes at varying angles. Thus, No. 88 torpedo would have to have its tube trained at an angle of 40.5° from the bow, and No. 30 at an angle of 36° from the bow, in order to come back to the line of sight at a point 800 yards ahead with the gyroscope set at 45°. The above condition holds when [a] torpedo is fired from a stationary boat. If, however, the boat is at a speed of 18 knots, one torpedo would have to have its tube set at 37.5° and the other at 35.5deg;, if the boat permitted such a sharp train.

Another method may be employed in compensating for this transfer, and that is, by fixing the torpedo tubes at an angle of, say, 45°, and then setting the valve of the gyroscope in order to make the torpedo turn through the correct angle. In this case, with both tubes fixed at 45°, Fig. 1, one torpedo would  require the gyroscope set for a turn through 50° and the other for a turn through 57°, in order to strike an object 800 yards ahead. If at a speed of 18 knots, the angle for one gyroscope would be 54° and for the other 59°, instead of the above.

Correction for Distance of the Enemy.

One great advantage claimed for torpedo fire lies in the fact that its accuracy does not depend upon the distance of the enemy. When, however, the torpedo is made to run in a curve, the distance of the enemy must be known, particularly when turning through a large angle like 90°. This is illustrated by reference to Fig. 1. This curved run of a torpedo as compared with its straight run is somewhat analogous to mortar fire as compared to [the] flat trajectory of a high power gun.

Avoiding Your Own Torpedo in Firing Ahead.

The tactical diameter of the torpedo in curved fire is about the same as that of a torpedo boat destroyer. If therefore the torpedo is discharged so as to curve and strike an object ahead and at the same time the helm of the boat is put over, both boat and torpedo are turning in a similar circle in opposite directions, starting from the same point, and would be in danger of collision in the vicinity of a point marked “C”, No. 4.

Use on Battleships.

On board a battleship using submerged tubes, the conditions are somewhat different, owing to the fact that the tube is in a fixed position, pointing abeam, without ability to train. Also, a battleship may be in formation where it is not desirable to change course. Conditions might arise where it would be advantageous to be able to fire at the enemy when he is ahead or astern or at some bearing not within range of a fixed tube. This requirement can be met by the use of angle fire; by making the proper corrections referred to above, and by providing facilities for quickly and accurately adjusting the gyroscope to the desired angle after it is in the tube. In firing at the enemy’s fleet, considering the whole fleet as a target and depending upon the probability of a hit within the extremities of the fleet, these refinements for accuracy would not have to be considered.

Effect of Striking Angle of the Torpedo.

The question is often asked as to what is the least angle at which the war nose is effective. For the 5 metre, Mark I torpedo, having a pointed head, the war nose become effective when striking at an angle of 18° or more with the surface; and the 5 metre, Mark II, and 3.55 metre, both having blunt heads, become effective at an angle of 30° or more. This question comes up in connection with the probability of a torpedo being effective in case the ship’s bow should be headed directly toward it.

Training Gear and Night Sights on Torpedo Tubes.

Training gear and night sights will be fitted on the tubes of torpedo boats and destroyers.


Changes have been made in the Naval Defense Mine, with a view to greater simplicity. Fig. 1 shows the general arrangement of the Naval defense mine as described in the publication and issued to ships until recently. This shows the mine case and automatic anchor; the quadruple conductor cable with battery box, splice insulators, safety break, releasing detonator, mooring bolt, and stray line for reaching the anchor cable. The changes relate more particularly to the electric cables and method of handling. The first type requires a quadruple conductor cable; two cables leading to the battery box at a distance of about 300 feet from the mines, and from there to a safety break 100 feet further; the remaining two conductors being utilized to explode a detonator with released the mine from the anchor cable and allowed it to come to the surface preparatory to raising the anchor. In exercising with these mines it was found on one or two occasions that the detonators in them had exploded, and on one could tell exactly how it happened. It is seen that with the old arrangement of quadruple conductor, the detonators are always in the electrical circuit, and if at any time the circuit should accidentally become closed while handling the mine an explosion would occur. It was therefore thought better to devise the plan shown in Fig. 2, in which only one single conductor cable is used, and the mine is not in the circuit at any time while handling it. When the mine is dropped, the single conductor cable is paid out as the boat pulls away, and when near the end of the cable the battery box is plugged in and thrown overboard. This should insure safety.

The method of using a releasing detonator to permit the mine case to rise to the surface before raising the anchor has been found unnecessary. A piece of wire rope is tailed on to the upper end of the anchor cable and is stopped along to the electric cable. In raising the mine, the electric cable is grappled for and when raised to the surface is under-run. If while under-running it you come across the end of this wire tail rope you should turn around and under-run the electric cable in the opposite direction, pick up the battery and disconnect it. This will throw the mine entirely out of circuit and it can then be handled with safety. The electric cable is then under-run until you again come to the end of the wire tail rope. This is taken in the boat and the mine and anchor are hoisted by it until the mine comes to the surface, when it is cut adrift from the mooring belt and floats alongside or is hauled into the boat while the anchor is being raised.

In planting mines, the book of instructions required them to be planted in two lines, so that the distance between mines shall not be closer than 128 feet. This requires rather careful work in order to properly close a channel so that a . . .


One set of detonators fired at 110 feet because the circuit closer was torn from its fastening and fell on its side in the bottom of the mine, hence closing the circuit.

With the present mine and circuit closer, 125 feet should be allowed between mines. By strengthening the circuit closer, this distance can be reduced with success to 80 feet. Below this distance, many parts of the mine begin to give way and the whole mine would have to be redesigned to stand closer approach.

The following method of planting Naval defense mines has been the one most successfully tried at the Torpedo Station. A distance line is marked by means of small floats at intervals of every 60 feet, with a weight or anchor on each end and a float or buoy placed near the anchor at about the depth of the water in which it will be dropped. This distance line is taken by a light boat and dropped in the positon where it is intended to lay the mines. The floats will show each position for dropping the mine, and one or more boats may be used as convenient. As each mine is dropped, a small boat with electric batteries in it should take the electric cable and row off at right angles to the line of the mines at a distance of about 400 feet or to the end of the cable, plug in the cable and drop it overboard. This is for convenience in recovering the mines. If it is desired to place the mines in a double row, thus increasing the lateral distance between them so as to insure greater safety, the line with distance floats may still be used, planting a row of mines on each side and opposite every other float at a distance of about one boat’s length.

Source: Papers of Frank Friday Fletcher (1873-1928), University of Virginia Library, Charlottesville, Virginia.