There seems to be some probability of a large fleet being assembled this summer to carry out a program of maneuvers on our coast, and this may be an opportune moment to call attention to certain deficiencies in our means of communicating between different vessels and between the different divisions of a fleet.
Very little advance has been made in recent years in the navy in the matter of signalling, except in the development of various systems utilizing the electric light. Certainly, in the wig-wag code, we have taken a step backwards in the adoption of the American Morse alphabet. The writer is unaware of the reasons which led originally to changing the superb Meyer one, two, wig-wag code for the English Morse, but it is commonly credited in the service that the English Morse was superseded by the even less desirable American Morse, now in use, because the signal corps of our army use the last-named, and it seemed desirable for both services to have the same code. If that is the case, we have crippled ourselves in consideration of a very remote advantage. The American code is used by all operators in the United States, excepting for most exceptional purposes, and there is an evident advantage in the army signal corps adopting it. With us, the space in the letters c, r, o, y and z, and in the conjunctive "and," is undesirable in the wig-wag, awkward in night flashing, and very difficult in fog-whistle signals. However, we must make the most of it. We have the American Morse code, and we must try to overcome the difficulties it presents in actual service, although a return to the Meyer code would not be unwelcome to the navy at large.
The qualities to be sought in any method or device for transmitting signals of any kind between ships are simplicity, rapidity, reliability and distinctness. As we increase the distance between the sender and the observer, we find it necessary to exaggerate the quality of distinctness, at the expense not only of rapidity, but usually of simplicity. Indeed, as between short-distance and distant signalling, different methods of transmission are found to be necessary. For instance, our largest signal-flags are eleven feet, and, even with them, in from three and one-half to four miles we have the limits of distinctness, as the colors blend; hence we must resort to some other method of long-distance day-signalling. Similarly, where hoists of lights are used at night, the lights blend at not very great distances, and we must resort to some other means. It will, indeed, be found necessary to have for all day, night, and even fog signals, a different device for long-distance signalling than that used at more or less short distances.
Any system which is proposed must take into consideration the immense differences in the rig and type of the various ships likely to compose a squadron. For instance, as between the Miantonomoh, Enterprise, Philadelphia, Vesuvius, and Gushing, the differences in the rigs and sizes of the vessels must necessarily limit the usefulness of any system apparently of great merit, judged in its applicability to one type of vessel alone.
DAY SIGNALS.
There are four conditions of service under which a hoist system of flags fails. One is where the flags fly edge-on to the observer; another is in a dead calm; a third is where the sender is in the glare of the sun between the receiver and the sun; and a fourth is at long distances where the colors blend. Now, with the eleven-foot flags it is possible at great distances to make out the shapes of the flags in a hoist (as between square flags and pennants) without being at all able to distinguish the colors. It therefore appears possible, by the use of various shapes in a hoist, to increase the range of visibility very materially over the present limit of flags, which limit is from three and one-half to four miles. This is a much-talked-of and frequently suggested method of overcoming these difficulties. The objections to using shapes are: 1. They take up too much room for stowage; 2. they are rather unwieldy and heavy; and 3. to be light enough they are seldom strong enough to retain their shape in service. They possess, however, the supreme advantage of visibility at very long distances, and there are conditions in time of war which seem to demand a system of long-range day-signals such as in one or more vessels as scouts or pickets for a fleet, or in blockading a port or an enemy with long distances between vessels of the blockading squadron, etc. In Plate I is given an arrangement of shapes from which a code may be selected. Figs, 1, 2, 3, 4 and 5 are suggested as alternates, or for code indicators, although to prevent the multiplication of shapes it might be well to indicate a telegraphic or
geographical signal by the 1st, 2d or 3d repeater hoisted over a signal, or displayed at a yard-arm or mast-head accordingly as the hoist is at the mast-head or yard-arm respectively. Just as satisfactory a way to indicate code signals would be to have numbers added to the general signal-book to indicate "the code to be used in reading the signals which follow." The necessity for the use of shapes for distant signalling might be rare, but it is an open question whether or not they should (in smaller sizes) replace for all purposes flags as hoist signals. In a calm, or head to a fresh breeze where the flags appear edge-on, or in the glare of the sun, they would be visible where the flags fail. As regards the materials of which shapes could be made, that is a matter for experiment. Were aluminum cheaper the solution would be easy. Paper squeezes might be made hard enough to offer sufficient resistance to abrasion or destruction in ordinary service. As a matter of fact, steel wire frames and canvas coverings are most probably the best materials. The shapes should be painted black or green, and each should be a figure formed by the revolution of a symmetrical body about that axis which is to be vertical in the hoist. In this way the shape presents the same appearance to observers at all points of the horizon.
A revised set of hoist signal-flags is being experimented with in the Squadron of Evolution. A yellow flag with a black ball in the center is substituted for the old No. 2 (white), in the interests of visibility; the repeaters are changed slightly; the geographical pennant (blue) becomes an assent flag; a negative is added; the despatch flag is used to replace the geographical pennant, and a new danger flag is substituted. These changes are, however, not so important in themselves as is the definite adoption of almost any set of flags for all ships in the navy. Once adopted, there is little need for change, unless to discard the whole system in favor of the use of shapes. There is much to be said on both sides, but the balance seems to be in favor of the shapes.
FOG SIGNALS.
To illustrate the disadvantages of the space in some of the letters of the American Morse code as applied to the fog-whistle, if the syllables ele, Hi, eli, ile, sle or els occur, there is no way to distinguish them from the simple letters o, y, r, c, z or and. Furthermore, as far as there is any official order in the matter, the general call is a sixty-second blast; a dot lasts less than five seconds; a dash, from five to ten seconds; and a long dash, front, or space, fifteen seconds. At this rate, with expert signalmen, five minutes would be good time in transmitting a change of compass course. It is drawing too fine a distinction, and above all consumes too much time to grade the lengths of the blasts as above in order to distinguish between a dot, a dash, a long dash, a space, and a call. The following changes are needed : Reduce the general call from sixty to thirty seconds; reduce the dot to one full second ; make the dash with two toots, each a full second, but with only a half-second between them; use a four-second blast for a space, and a ten-second blast for the letter l or the numeral 0. Care must be taken to distinguish between the letter i, which is two dots (toots), and t, which is two dots or toots with a half-second interval. To facilitate using the whistle, make fast the whistle-cord about one foot from the end of a squeeze handle, and work the handle as a lever, securing one end as the fulcrum, and applying the hand as the power at the other. In this way one can manipulate the whistle systematically. It has been suggested in the service that squadrons should practice as in a fog at sea, by enclosing the captain, officer of deck, and man at the wheel in a canvas screen, so that they cannot see around at all, and exercise with whistle, etc., as in actually steaming in a fog. It would be excellent practice, and is needed to familiarize people with the new code and its difficulties. The North Atlantic squadron has been about the only one, up to recently, where much squadron cruising has been carried on, and some years ago. with the Meyer code, no difficulties were experienced in the least in certain forms of tactical drill in foggy weather.
Distant signalling in a fog beyond the limits of audibility of the steam whistle can only be carried on by gun-fire. In this case it is impossible to make a dash by a prolonged sound, and it should be represented by two blasts with a full f-second interval, and a space by three successive fires with a full |-second interval between each. The revolving cannon seems to offer the best solution. As the continued firing of a single gun is a generally recognized international signal of distress, the general call should be a dot, dash arrangement, say five dots and five dashes, alternating, first one dot then a dash, etc.
NIGHT SIGNALS.
In the matter of visibility the Very's system of night signals leaves nothing to be desired. It can be used with accuracy up to ten miles or more under favorable conditions, and is unexcelled. There are certain modifications that experience calls for. In the first place, the pistols are miserably poor affairs. It has been suggested that short double-barreled breech -loading shot-guns be issued as projectors. They have been used in the service and been found to work admirably. In the next place, the original code, involving the bracket in certain numerals, should be changed for a four-element code. The bracket leads to too much embarrassment, in that, if one of the cartridges fails to go off, it involves repeating from the beginning. With a four-element code, if a cartridge fails to go off, one can keep on trying at least for two minutes to get another one off, as the ammunition would have to be almost worthless to cause a repetition of the message on account of over two minutes' delay between fires.
The four-element code now used in the Squadron of Evolution and the North Atlantic squadron should be issued for general service. It is as follows:
General Call, G followed by rocket.
Message Call, G. (The message call is to be habitually used as a general call when the ships are within ordinary signal distance.)
1RRRR
2GGGG
3RRRG
4GGGR
5RRGG
6GGRR
7RGGG
8GRRR
9RGGR
0GRRG
Answering, G Repeat, R
Divisional point, date pennant and designating flag… G G R G
Interrogatory pennant… RGRR
Affirmative or "yes" pennant… R G R G
Negative or "no" pennant… G R G R
Numeral pennant… GRGG
Annulling pennant… R R G R
Danger or distress... R repeated.
Telegraph flag… (R G) bracketed.
Geographical pennant… (R G) bracketed followed by rocket.
Use Navy List ...(R G) (R G) bracketed in pairs.
While this code involves longer time to work than the original three-element one, in the end it saves time through the non-liability of having to repeat. The telegraphic, geographical, and navy list designations contain brackets, but as these precede a numeral signal, the correction of the failure of one color in a bracket is quickly accomplished by beginning over. With a better projector than the present pistol, and, using the four-element code, the Very's system possesses for distant signalling the ideal qualities of visibility and certainty. It is, however, too slow for tactical and routine squadron signalling at moderate distances. The Squadron of Evolution is using at present the Ardois system of night signals, and a brief description of it may not here be out of place. The method of slinging the cable containing the wires leading to the lamps, and of suspending the lamps aloft, is shown in Plate II. The cable is seized at intervals to a backstay, or a special wire-stay, to take the strain, and the lanterns themselves are suspended on Scotchmen, which are seized to the cable, and further supported by a distance-line from the lantern next above. There are five lanterns, each double, as shown in section in Plate III, and each containing two 32-candle incandescent lamps. The outer globe, or lens, of the lantern is in two colors, the upper half being white and the lower red, separated on the inside by a brass diaphragm. In any display of lights only one light in a lantern is shown, that is, the light is either red or white, but never both red and white in the same lantern. Practically, the Ardois code admits of a display at any one time of from one to five lights, any one of which may be red or white. The wires of the cable lead to a box, circular in shape, and divided on its upper face into sixty-four segments. A top view of the box is shown in Plate III. Each segment corresponds to a certain display of lights which is indicated on the face of the box in that segment by dots, colored red or white to correspond. In the center of the box, and in the same plane as the whole top disc, is the circular turn-table, a, carrying with it the pointer, c, and the vertical cylinder, d. When it is desired to display a letter, say A, the pointer, c, is brought opposite the segment marked A. When the catch, b, slips into a space abreast d, the only thing that remains to be done to make the display (in this case red, white, red) is to turn the handle, e, through 90°. The act of turning the handle, e, is to revolve the central vertical axis in the cylinder, d, which thrusts out ten little pistons, which make contact with such terminals as will light up the proper lamps, so that, in this case, the lamps would show red, white, red. Displays are read from the bottom upwards. The general call is the so-called cornet, a red and two whites. There are five code-calls which indicate which code to use. "Gen" means the general signal-book; "Letters," the Ardois alphabet; "Compass," a compass signal; "Cypher," a special code, etc. Every display of lights is answered by the ship receiving turning on the same display, each keeping it on until the sender turns off, which is not done by the sender until all repeat. In case a ship is so situated as not to be able to see the display or one or more of the lights of the sending ship, this observer takes the signal from some other ship and repeats it. In this way absolute certainty that a message has been accurately received is attained. It is quite a rapid system of signalling when the fact is taken into consideration that, through repeating back the signal, certainty is assured, and the method can be adapted to other codes and systems. The signals are visible, under favorable circumstances, up to about three miles. As a summary, it may Be said that the Ardois theoretically possesses the advantages of certainty, mobility, rapidity, and visibility for night signalling in squadron. Practically, there are certain mechanical defects, which, however, admit of correction. Owing to sparking and the consequent fusing or burning of the pistons or contact-studs which are thrust out from the cylinder, d, when the handle, e, is turned, a good deal of overhauling of the signal-box is necessary. The use of platinum contacts would obviate this. There is a serious defect, also, in the lanterns, in the inability to shift a lamp readily in case a filament is destroyed by burning out or otherwise. Some change could be made in the lantern to meet this objection, and the quality of certainty be practically as well as theoretically secured.
There are, however, some serious objections to the use of the Ardois system in our service, i. It is ever so much too complicated mechanically, and the same advantages which it offers can be obtained much more simply. 2. Each outfit costs about $1800, where a much less expensive device will accomplish all that it will, and admit of being readily overhauled in case of faulty circuits, whereas the Ardois cable is difficult to test out or repair. 3. It introduces a new alphabet, numerals, and signal code generally, whereas a similar but much simpler device will admit of the use of the American Morse alphabet and the Very's code.
The device here proposed is shown in Plates II and III, and may be described in general terms as follows : It consists of five groups of three lights each, the groups being spaced about five yards apart, and each group consisting of a red, a white, and a green light. The wires running to the lights lead through a long tube, put together in sections, which will admit of shortening the distances between groups for special ships, and also admit of readily overhauling the leading wire circuits in case of the development of faults. There are, in all, fifteen lights, which, with a common return, would give sixteen wires in the tube. The insulation of the leading wires should be colored red, white, or green, according to the color of the light it operates, and marked with the proper number on a brass tag. The inside of the tube should be coated with shellac to prevent the grounding of any wire that might by accident be bared of its insulation.
The wires should lead to a key-board, conveniently located. A rough sketch of the plan for this is shown in Plate III. The keys should be those of the ordinary kind used in turning on and off an incandescent light or a branch circuit, and should be arranged in five sets of three each to correspond with the lights aloft, so that the keys for the top lights should be on the left of the key-board and in the same order as they are aloft, viz. red, white, and green. A switch controls the current, and it is not thrown on until the proper keys have been turned on the board to make the desired display aloft. Then the switch is turned and all the desired lights appear simultaneously, and should be kept on until answered by the same display from all ships receiving. The display should be read from the top downwards, just as a hoist of flags would be read, and not from the bottom up, as in the Ardois. It must be remembered that in the last-named system the central turn-table must be revolved each time until the pointer comes opposite the desired letter or character, then the catch, b, must be sprung, and finally the handle, e, turned through 90°. Certainly, in the time which is taken by this operation, a display could be made by the proposed system, for it only requires rapidly turning certain keys and then throwing on a switch. Above the key-board, over each set of keys, should be a group of very small low-resistance lamps, colored the same as the lights to which they correspond aloft, and so arranged that, as a key is turned, its tell-tale miniature light would instantly show with its proper color. The display above the key-board would be the index of the correctness of the signal which is about to be displayed aloft when the switch is thrown on. This will obviate mistakes due to accidentally turning a wrong key, and admit of placing the key-board well under cover where the real display is not visible to the sender or operator.
The proposed lamp is shown in Plate III. The longer the arm, a, the less the tube, e, however large it may have to be, will cut off the light from an observer who has the tube in line with the light. The arm, a, also admits of having the lamp vertical, which helps it to shed water. The globe, c, is red, white, or green, accordingly as we wish, and it is open below to admit of readily shifting the lamp in case of a broken filament. The globe is held in place by a hinged collar, d, which tightly clamps it in place. It is a simple matter to hoist a man in a boatswain's chair. to shift a light or replace a broken globe, and the proposed arrangement of lamp and globe admits of both being cleaned readily.
To make a signal according to the Very four-element code: Say the signal is 1239. Make the message-call by showing the upper green light until answered. Display four reds, and when all answer with four reds, turn off; then four greens, etc.; then three reds and a green, etc. The advantages of this system are that a signal can thus be sent in one-fourth the time that it can be by firing the colored lights; and, by each ship repeating, no mistake in receiving is possible. Hence for squadron purposes at ordinary distances the gain is an immense one. To make a bracket, display the upper red and the second green; to make two brackets, display the upper red, second green, fourth red, and lowest green. This will leave a space between the brackets. To make a bracket and a rocket, use the upper bracket, followed by white, green, and red; in other words, make a rocket by a display of white, green, red. It will be observed that in this device for signalling with the Very's four-element numeral code, we have to use only four lights at one time, and if for any reason a light fails, and it happens to be either a red or green of either the uppermost or the lowest group of lights, we still have four reds or greens to work with. In other words, there are two chances in five of not being blocked temporarily by a failure of a light. In the cases cited of uppermost or lowest red or green failing, it does not block us in the double bracket or the bracket and rocket, for we can always make the alphabetical signal, G. L. U. (geographical list use), or N. L. U. (navy list use).
To use the American Morse alphabet: This grouping of lights proposed answers admirably for the service wig-wag code and overcomes all difficulties. No letter has more than five elements. Call a white light a dot, a green a space, and a red a dash. It will be observed that in the alphabet there are spaces only between dots, hence the green light will only be displayed with whites. For instance, y would be two whites, green, and two whites; k would be red, white, and red. Now the numeral code of the American Morse contains one numeral, 6, with six dots. We have provided already for the numeral code in the Very's signals. To use it for wig-wag numerals, display the uppermost white light over the Very four element numeral, and this will signify that the Very numeral, which appears under it, is to be read as a simple numeral. An "error" is seven dots, and this should be changed to five reds. There yet remains one additional feature of this proposed device which must commend itself. If any light fails so as to apparently block a signal, we have always recourse to the other two lights of the set in which the break occurs. For instance, if the red light fails, display both the green and white; if the white fails, the green and red; if the green fails, the red and white. This is the only case in which two lights of the same set are displayed simultaneously. At long distances the two lights will blend, as they are only a foot apart, but an adjacent ship seeing the break will interpret correctly. In tactical signals speed of signalling is everything, next to certainty, and in any case we always have the pistols or firing code to fail back on, but in a squadron at tactical distance there would be no difficulty up to half a mile in seeing the two lights distinctly. It must be borne in mind that the failures of lights, here provided against, do not on the average occur oftener than once in two or three months, but so much importance is given to that consideration here, because the failure is bound to occur at a critical juncture, and the remedy must be immediate. It will be observed that the Ardois code can be as readily transmitted by this device as by the $1800 machine furnished with it, with the additional provision, as above, in case of a failure of a light. Furthermore, by cypher codes, by changing the color of globes, by almost any combination, we can adapt this device to almost any desired system. It will also be observed that the time gained in displaying all the elements of a letter in one display as against flashing the elements successively is further increased by non-liability to having to repeat, and certainty that the message is received as sent.
There are many details which are omitted on account of lack of time and space. If, in the coming season, it is desired to fit all the ships of any squadron with this device, it can be done in a very short time. All the materials can be purchased in the open market. Any seaman-gunner ought to be able to run the wires and arrange the lamps and key-board, as the lamps proposed are very simple, and the wire is No. 16 (Birmingham gauge), .05 inch in diameter, carrying seventeen amperes to each lamp. As only two amperes are required, it leaves a sufficient margin of safety for weathering.
This paper has been somewhat too hastily prepared, and the illustrations are only rough sketches. Working drawings can be furnished, but it has been the hope that the requirements of the device are so simple that no detailed explanation would be needed.
It seems not out of place to here again call attention to the need for increased pay, and the rating of signalmen on board ship for those called upon to stand signal watch.
It would be a good thing just now to undertake a thorough overhauling of orders relating to signals and a correction of defects in the systems in use. Anything that can be proposed has objections that can be cited against it, but anything is better than confusion and lack of uniformity. Doubtless the reports from the various squadrons throw much light on the experiments now being conducted. It can at least do no harm to propose the following scheme:
For 1. Day, ordinary: Small shapes in a hoist.
2. Day, distant : Large shapes in a hoist.
3. Fog, ordinary : Steam whistle.
4. Fog, distant: Gun-fire with revolving cannon.
5. Night, ordinary : Five groups of lights of three colors each.
6. Night, distant : Very's night code of four elements.
We have already the international code, and the above, or anything else to take its place, contains enough work to make it worth while for us to rate our signalmen and pay them better.