It is not proposed to here touch upon the history of signaling, nor to consider the question of interior communication aboard ship, viz., by annunciators, voice tubes, bells, etc. It is the plan to first point out the limitations which affect and control the methods of communication which are available under different conditions, and from a consideration of these to show wherein naval signaling differs in principle from army signaling. We will next take up the consideration of signal codes in use in different military organizations at home and abroad, and from them deduce the theoretical principles which should determine the choice of such codes. This involves a discussion of the concessions which theory must in certain cases, and for certain good reasons, make to practice as brought out by actual trial in service. The methods of army signaling are touched upon only to illustrate the difference in principle. The different methods of signaling at sea in use in our own and in foreign navies are then taken up and discussed in detail. However meagre the data on which are based the conclusions here set forth, enough is presented to show that there are certain definite principles which must be adhered to in the formation of a code, or the design of any sort of an apparatus or method of transmitting it.
In the navy we have in a measure been dominated by the army. This influence dates back to the Civil War. The Myer signal code was invented by the late Brigadier-General A. J. Myer in 1856, and was first reported on favorably by a board of army officers in 1858. In 1860 Myer was appointed chief signal officer of the army, and ordered West on Indian campaign duty. Many of the officers and men whom he then instructed and organized into a signal corps went South on the breaking out of the Civil War, This resulted in the Myer code being used with conspicuous success on both sides. It was also adopted in the navy, but even in such important operations as those in Mobile Bay we had army signalmen aboard all our ships to receive and transmit messages. It remained the army and navy code until April 3, 1886, when it was unwisely superseded by the Continental Morse, which was in January, 1890, itself more unwisely succeeded by the American Morse code. It was the genius of Myer which first organized and perfected the signal corps of our army, and which first demonstrated the value of rapid and accurate communication in extended operations on shore. Since his death we have gained nothing from the army in the matter of signaling.
In recent years the addition of the duties of the weather bureau to those of the signal corps has done much to overshadow the importance of signaling as an art in itself. Since the Civil War the navy, in squadron cruising, has had daily practice in signaling of all sorts, and we now have sufficient experience to qualify us to judge for ourselves of our own needs. It will appear further on that some of the conditions of signaling in the navy are widely different from those governing army practice. We have, therefore, ample data upon which to strike out for ourselves in this important matter.
We start first with the limitations of signaling in general.
Rapid and accurate communication between the different parts of organized military forces, both afloat and on shore, is of vital importance. The limitations of military signaling in general are, roughly speaking.
1. Distance.—At sea the dip of the horizon limits the range of signaling, and mechanical difficulties add greatly to the problem. On shore modern inventive genius has in a measure solved most of the problems of distance by the use of telegraphy and heliographing for military signal purposes, while both afloat and ashore carrier pigeons offer a restricted field of usefulness.
2. Meteorology.—Fog, mist, rain, falling snow, gales of wind, and the conditions of the atmosphere as regards refraction, etc., affect materially the methods which are available for communicating or signaling. Under this head we may also classify daylight and darkness, that is day and night.
3. Topography.—The character of the land or water, hills, valleys, mountain ranges, deserts, woods, etc., determines also the character of the methods we may use in communicating. At sea the chief limitation of topography is the dip of the horizon, which is in itself one of the restrictions imposed by distance.
4. The enemy.—In time of war the presence of the enemy in force between signal stations, the smoke of action, and the exigencies of battle may require the use of special methods of communication; for instance, in the recent labor troubles at Homestead, Pa., the militia surrounded the steel works, one division being encamped on the opposite side of the river. Owing to the smoke from the mills and the vigilance of the strikers, ordinary communication with wig-wag flags or telegraph was cut off" between the two main divisions of the militia. The heliograph was brought into use and the reflected sunlight readily pierced the smoke, and messages were read without difficulty.
On shore to communicate rapidly and accurately, four methods are available under some one of the above limitations, viz.:
1. Messengers.—Messenger service is here taken to include all means of sending a verbal or written message by a bearer ; for instance, by an orderly mounted on horseback, or bicycle, or in some vehicle of transportation, such as a wagon, railway or balloon. It also includes pigeon service, which is a valuable adjunct to any means of day communication.
2. Visual signaling.—This includes all methods of transmitting a prearranged code by the motion of some object, by the display of certain symbols, shapes, or lights, or by the flashing of light.
3. Phonetic signaling.—In this the same object is accomplished by the emission of sounds of different characters; as, for instance, by gunfire, bugle, bell, drum, gong, whistle or siren.
4. Telegraphy.—With a. a well organized field corps, and an outfit including wires, transmitters, poles, wagons, batteries, telephones, balloons, etc., we have in telegraphy the highest development of military signaling.
At sea the conditions are somewhat different. Dispatch vessels and the other auxiliaries of a fleet carry messages, and the pigeon service has a more limited application than on shore. Visual and phonetic signaling are, however, admirably adapted to naval uses, but telegraphy between vessels is as yet not feasible, although telegraphy by induction at sea has passed its earliest experimental stages and may be perfected in the course of a few years. However, at the best its range will be short and its usefulness restricted to less than long visual distances. In other words, telegraphy and heliographing from high stations, combined with balloon and pigeon service, give a very long range of communication on shore; at sea the curvature of the earth, the absence of elevated stations except by captive balloons, the impracticability of telegraphy and distant heliographing, all combine to restrict the range of signaling, and, as before stated, make the problem of distant signaling at sea one of the hardest to solve.
It would seem that army and navy signaling should involve the same underlying principles, but, curiously enough, the development has been along such radically different lines that the fundamental principles are surprisingly different, at least as between the methods used in our own army and navy. In the army the principal methods used are the wig-wag, the telegraph and the heliograph. These cover nearly all the conditions of strict signaling, as at night the torch is used with the wig-wag and a flash light with the usual type of lantern and shutter. These methods all aim at communication with only one station at a time, that is to say, to wig-wag, telegraph, or heliograph from one station to another. All around visibility or audibility is not a factor. In the navy we have sought and are seeking the development of methods which give visibility all around the horizon, so that a squadron can see the signals from any direction. Later on, in the consideration of the theoretical principles of naval signaling, it will be shown that the objections to flags and to the semaphore are largely that they do not possess this virtue of all around visibility. It is very important, therefore, at the very beginning of the consideration of naval signaling to point out the capital error of adopting army codes and methods in our navy.
Let us first take the heliograph. In our western territories, during several Indian campaigns, the heliograph was used most successfully. In the military department of Arizona from November, 1889, to May, 1890, a chain of heliograph stations 2000 miles long was maintained in Arizona and New Mexico, completely networking the country in which the hostile Indians usually operated. The main stations were on mountain peaks, some 8500 feet above the sea level, and "were from 50 to 100 miles apart. Connected with these were numerous secondary stations, forming a complete network over the territory embraced in the main lines of communication, so that any movement of the hostiles was rapidly reported to headquarters. The entire scheme was under the management of and developed by Col. W. J. Volkmar, U. S. Army, of the Adjutant General's Department. Two sizes of mirrors were used, 4 ½ and 8 inches square, for short and long ranges respectively, although the smaller size was successfully used on the long ranges. These were mounted on one end of a sighting arm that was itself mounted at its middle point on an ordinary tripod. Through a small imsilvered spot in the center of the mirror, this arm could by the eye be pointed at any station and clamped in position. On the end of the sighting arm opposite the mirror, sliding on a vertical sight-bar or rod, was a small round disc. To adjust the instrument on the distant station, the disc was moved up or down so as to just mask the view of the station from the eye at the unsilvered spot on the mirror. The mirror was capable of a slight lateral motion on its horizontal axis by means of a tangent screw, and of a revolving motion around the said horizontal axis. The object of the disc and of the unsilvered spot on the mirror was that when the unsilvered spot cast its shadow, so to speak, on the disc in front of it, the full beam of reflected light from the mirror was on the distant station. This enabled an assistant operator, by adjusting the axis and turning the mirror gradually, to follow the sun in its change of altitude and azimuth, and to keep the beam of light constantly on the distant station; for all he had to do was to keep the disc unilluminated by the unsilvered spot. Some three feet in front of the mirror was another tripod carrying a vertical shutter or screen, operated by a lever so as to fall flat or rise to a vertical position at the will of the operator. This shutter, when vertical, masked the beam of light thrown by the mirror towards the distant station, and by working it the beam could be flashed on the dot and dash system, using the Morse code. To call the distant station the beam was turned on steadily till answered. If at any time the sender's mirror got out of adjustment so that the distant station failed to see it during the course of a message, the distant station turned on its beam of light to stop the message till re-adjustment was effected. So successfully was this means of communication operated that, on May 14, 1890, partial communication was established between the stations on Mt. Graham and Mt. Reno, a distance of 125 miles. Two stations with one intermediate covered 195 miles. Constant communication was kept up between Mt. Graham and Lookout Peak (Sierra Anchas), 105 miles distant, and between Mt. Graham and Fort Huachuca, a distance of 90 miles. The altitude and the remarkable clearness of the atmosphere were, of course, favorable to such remarkably long ranges. As previously noted, the heliograph was used this summer at Homestead under peculiar circumstances, illustrating its value in smoky weather and with stations separated by the enemy.
The navy has blindly followed the army in the matter of signal codes ever since the Civil War. In all the three codes which we have gotten from this source, viz., the Myer, the Continental Morse, and the American Morse, we have copied the instructions for heliographing as if we seriously contemplated using this method. Heliographing is not possible from one ship to another, nor from a ship to the shore, nor from the shore to a ship. Absolute immobility is required in the receiving and transmitting stations. This we do not get in either a ship at anchor or underway. Heliographing from its name implies the use of the sun's rays. It is a method of visual day signaling. It is adapted to high altitudes where the disturbing effects of refraction and the convection due to heat rays is inconsiderable. Along the coast at the sea level it has a most limited range of usefulness. It is essentially not adapted to naval purposes. What we use and what we want is a flash light or lantern, which is a very different sort of an affair, as it gives visibility all around the horizon. Of course it is well enough to know how to use a heliograph in the navy and how to signal with one, for they are useful in triangulating work, such as is done on the coast of lower California and Mexico, but, as a method of naval signaling, heliographing is not especially useful in cruising.
The army, in June, 1889, adopted for signaling the American Morse code in place of the Continental Morse.
The reasons which led up to this change were good and sufficient ones, considering the problem solely from an American standpoint. We have no military telegraph lines to speak of, and in any operations in our own country, as, for instance, in recent Indian wars, messages must be sent over commercial lines. The operators are skilled in using only the commercial code, and, besides the vast number of operators offers an inexhaustible supply on which to draw in time of war. The many theoretical and practical defects of this code are gotten around the best way possible in the wig-wag and by the heliograph for the sake of the advantage the code offers in telegraphy. Now, from a naval standpoint, no advantage accrues to us from adopting such a code. We do not want telegraph operators on board ship; we have no signal corps to recruit, and we have no telegraph lines.
This code was invented before the principles of signaling were at all understood, and has been in use so long that it is now impossible to either correct its defects or substitute any other for it. The best military codes have only two elements. The American Morse has four, viz., the dot, the dash, the long dash, and the space. When this is transmitted by a wig-wag flag or a torch, the dot is a motion to the right, the dash is a motion to the left, the space is a motion to the front, the long dash is made by pausing in the motion to the left and holding down the flag or torch for a preceptible interval; the "front" or space between words is made by introducing a slight pause before and at the end of the front motion to distinguish between it and the space in the letters c, o, r, y, z and &. In other words, in addition to the elements of motion right, left and front, a pause or element of duration is introduced. This is objectionable, as it brings tension on the eye to catch the pauses, and in long range signaling, when the receiver uses a glass to aid the eye, it is very trying. An alphabet and numeral code require only thirty-six characters. With only five elements in each character, that is a maximum of five and a minimum of one, sixty-four permutations and combinations of two elements are admissible, yet this code goes to all the trouble to introduce characters containing six and seven elements, and that, too, with four fundamental or unit elements to build upon. Using the four units, the theoretical number of possible permutations and combinations, with one as a minimum and six as a maximum, is 5460. However, as the space can only come between the other elements, the practical number is much smaller, but still the three or four thousand possible, and the need for only some forty or fifty characters at the most, show how little theory entered into the considerations when the American Morse code was invented. The army may be able to put up with the many defects of this code in consideration of the other advantages which come with it, but for naval purposes the needs of a code cover a much wider range and impose most trying conditions. For instance, in the wig-wag, unless the sender is facing the receiver squarely, the space motion in the letters will invariably be taken for a dot or dash (according to circumstances). On shore, in communicating between stations, there is no difficulty in facing squarely. Afloat, on board a rapidly moving torpedo-boat or a vessel turning, the sender cannot always face the receiver squarely, particularly if sending to two or more vessels or receiving stations which are not close together. With a telegraph key and sounder which works quickly, the space is rather short, and rapid signaling is possible. Using a fog-whistle or an electric light to transmit this code, the slowness and deliberateness which must be used makes it necessary to prolong the element, as will be hereafter explained. The time elements of space and long dash add much to making this the slowest code invented. Practically on the fog-whistle this code is transmitted as follow : the dot is a short toot, the dash is a blast of at least two or three seconds, the space is a blast of five to ten seconds, the long dash is a blast of ten to fifteen seconds. Failure to distinguish between the long dash and the space letters leads to confusion in distinguishing the space letters, c, o, r, y, z, and &, from syllables containing the letter "l," such as ele, eli, ili, ile, sle and els. The time consumed is enormous. To make signal for a simple change of course requires some five minutes, during which time the blasts are almost continuous, approaching vessels cannot be heard, the people on the bridge are deafened, and the vessel itself is positively a danger to navigation. In some such system of night signaling as the Ardois or similar apparatus used abroad, there is no way to transmit the space letters, and a hoist of six lanterns is required to transmit the numeral six. As we are the only navy that uses as many as five lanterns, six would, indeed, be a step backwards. As a matter of fact, the American Morse code cannot be used on the Ardois, and in our service we have a special code called the "Ardois" to use with the apparatus. These criticisms are only meant to bring out strongly the important fact that it is not safe to violate the fundamental principles of signaling in the construction of a code. There are principles in signaling and it is now proposed that we take up the theory of signaling and work out those principles, testing them afterwards by the conditions imposed in actual service. It is certainly unsafe for us to be guided by the conditions which determine the selection of a code by the army, when the objects to be sought and the topographical and meteorological conditions are radically different. The simpler the whole question of signaling can be made, and the fewer the codes which men have to learn, the better we will get along, as we have no corps of trained signalmen and we require practically all our officers and men of the active combatant force to be up on signaling.
The Theory of Signaling.
Signaling is the transmission of a prearranged code as accurately and rapidly as mechanical or other means will admit. All signals or codes are prearranged or preconcerted. The simplest are those of one element where the display of a certain motion, sound or symbol, a certain number of times or under certain circumstances is taken by previous agreement to have a special meaning. For instance, a lantern, a handkerchief, a hat, or anything by which the attention is attracted. In codes of more than one element the. Elements are color, motion, sound, time, form or position, grouped to represent special characters, letters or numerals, or all three. The characters, numerals or letters may or may not in their arrangement make direct sense. The key is furnished usually, in some book, or else the message is spelled out by means of an alphabet. Signals are of two kinds, transient and permanent. For instance, motions and sounds are transient because they disappear as soon as completed. Symbols, such as flags, shapes or lanterns, are kept hoisted or displayed till read, and are therefore classed as permanent. It will be found that symbols which answer for short distances generally fail at long distances, so that the question of signaling naturally divides itself into ordinary signaling and distant signaling. From a naval standpoint ordinary signaling is what may be called squadron or fleet signaling, that is, signaling within the limits imposed by squadron or fleet cruising. Under any circumstances a good code is of the first importance. Expertness comes only with long practice. It is important, therefore, to limit the number of codes which a signalman has to learn to receive and transmit under varying distance, by day or night, or in fog or smoke. The observation of fundamental principles in the formation of a code gives flexibility and reliability.
The simplest military codes in general use are those based on the ten numeral characters. One element may be displayed from one to ten times, or groups of two or more elements, each group representing a numeral character, may be used, or ten different elements maybe adopted, each representing a numeral character. One element signaling is too slow. Using ten elements, represented, for instance, by ten flags, gives rapidity, but is more complicated. As a rule, the greater the number of elements the greater the speed in signaling, but the more complicated it becomes, and the greater the mechanical difficulties. Instead of ten flags, we may use, for instance, ten different positions of a semaphore arm, or ten different shapes, or ten different displays of lights, etc. With a semaphore or wig-wag, or display of lights, the numerals of a group or combination are displayed successively, and in some forms of distant signaling the elements of each numeral character are also displayed successively. This is very slow, but the mechanical difficulties and considerations of visibility limit both the number of elements and speed in signaling. With the Very's night code, in which stars are fired from a pistol, the two elements used are red and green. The height to which they go and their visibility makes it a most excellent night code for distant signaling. The use of only two elements, the desirability of groups of four, and the requirement of high firing, restrict the speed of signaling. The code is as follows:
1. RRRR 2. GGGG
3. RRRG 4. GGGR
5. RRGG 6. GGRR
7. RGGG 8. GRRR
9. RGGR 10.GRRG
Another form of a night numeral code of two elements, in this case white and red, is that used in the German navy with the Conz electric night signaling apparatus. It is as follows:
1. W 6. R R
2. R 7. W R R
3. W R 8. R W W
4. R W 9. W W R
5. W W 10. R R W
Four other signals, W R W, R W R, W W W, and R R R are used for code purposes.
In flag hoists the signal composed of a group of numerals is made in one hoist, and this gives rapidity in signaling. Limiting the hoist to four, the permutations and combinations, using one, two, thee or four in a hoist is 11,110, which using code flags to five different meanings to the same display, gives ample range for a general signal book. It has, however, been found by experience that four-flag hoists are difficult to manage in a strong wind, and take time to bend on and hoist. If we limit the hoist to three, the possibilities are only 1110. The Italian naval authorities have adopted the consonants of the alphabet as represented by the eighteen International Signal Code flags. With three as maximum and one as minimum in each hoist, the total number of displays admissible, using the eighteen international consonants is 6174. The step beyond this is to take all the letters of the alphabet and the ten numerals besides. In English this would give us thirty-six characters, and enable us to spell out any message. The scope of such a code is limitless, but to transmit it by thirty-six different flags is impracticable. In the French Mediterranean Squadron an experimental night code is used which involves two peculiar innovations. The two elements are red and white electric lights. By a make-and-break key these lights may be called flashing. This gives four elements in reality, a fixed white, a flashing white, a fixed red and a flashing red. As in our Very night code, the groups are each of four elements. If any display contains less than four elements, there is a mistake and the signal is disregarded.
In a semaphore code the elements are those of position, and each letter is displayed transiently in one motion. This gives great speed, but limits the code to visual signaling. There is no way of transmitting it phonetically or by flash. The night semaphore, using arms illuminated with lights or lighted up by reflection is not regarded as at all satisfactory. Many experiments have been made in the English navy, but the result is that the Continental Morse code is used with a flashing or winker light, transmitting the dot and the dash.
To construct an ordinary alphabet and numeral code two elements are used, and these are combined into thirty-six or more different characters. Two elements, limited to four as a maximum in any combination and with one as a minimum, give a possibility of only thirty different characters. With five as a maximum the limit is sixty-four, which gives great scope, as only thirty-six are absolutely needed. In a theoretically perfect code of this character, the following conditions should obtain:
1. No elements of elapsed time, or duration, and no elements requiring a pause should be used, as the tension on the eye, ear, and brain is very trying.
2. Transient elements, such as those of motion, give speed in signaling, and are better than time elements.
3. Two elements, limited to a maximum of four in each combination to represent a letter, give the best results.
4. The vowels and letters which are on the average most often used should be selected from the simple elements or combinations having the fewest elements, in order to give speed in signaling.
5. Groups of symbols, representing words or complete phrases and sentences, should be separated by a positive character or symbol, instead of by a pause or time element of duration.
6. The numerals should all have the same number of elements in the combinations representing them.
7. No numeral signal should be made as such, without a distinctive signal "Numerals follow," and, on the completion, one implying "Numerals end."
These principles here enunciated may not, on grounds of expediency, be best in some particular method of signaling, but, for a general code adapted to all conditions, they will be found fundamentally sound and safe to build on. So many mechanical means of transmitting signals have, from time to time, been invented and tried, and each new invention has seemed to treat the theoretical requirements of a code with less and less consideration, that the result is theoretical considerations are either little understood or never heeded. Many of these mechanical devices are extremely valuable, but it should not be lost sight of that a good code is of the first importance, and the mechanical means of transmitting it is in a sense secondary. No change in a good code should be made simply for the sake of adopting a good mechanical means of transmitting it, unless the reasons therefor are of the utmost importance.
The simpler the question of signals can be made, and the fewer the number of codes that have to be learned, the less training signalmen will require, and the greater the number of officers and men that can become proficient. There is, of course, such a thing as sacrificing too much to simplicity, but if one code can be made to answer for all the devices required for squadron and distant day, night, and fog signaling, then it is reasonable to expect that every officer and man in the combatant force shall know that one code thoroughly.
The two best known and most thoroughly tried military signal codes in the world are the Continental Morse and the Myer.
In this code the distinction between the elements dot and dash is one of duration. The interval between the words is a pause. In transmitting it by a wig-wag or a torch, the dot is a motion to the right, the dash is a motion to the left, and the pause or time interval representing the space between words is a motion to the front. In other words, time elements of different values are transmitted by motions of equal value in point of time of making them. With the heliograph or winker light or on a fog-whistle, the dot and dash are distinguished by the greater duration of the dash. On the Ardois or Kasolowski, or similar system, the dot is a red and the dash is a white light. By gun-fire the dot is transmitted as one gun a dot, two successive guns a dash. There is no positive character or “front” separating words and sentences, or groups of numerals, but a pause or element of elapsed time is used instead, which is one of the few defects of the code.
The Myer code is as follows:
Myer Code
A 2 2 H 1 2 2 O 2 1 V 1 2 2 2
B 2 1 1 2 I 1 P 1 2 1 2 W 1 1 2 1
C 1 2 1 J 1 1 2 2 Q 1 2 1 1 X 2 1 2 2
D 2 2 2 K 2 1 2 1 R 2 1 1 Y 1 1 1
E 1 2 L 2 2 1 S 2 1 2 Z 2 2 2 2
F 2 2 2 1 M 1 2 2 1 T 2 End of word 3
G 2 2 1 1 N 1 1 U 1 1 2 End of sentence 3 3
End of message 3 3 3
Numerals
1. 2 1 1 1 2 6. 1 2 2 2 2
2. 1 2 2 2 1 7. 1 1 2 2 2
3. 2 2 1 2 2 8. 1 1 1 1 2
4. 2 2 2 1 2 9. 1 1 2 1 1
5. 2 2 2 2 1 10. 2 2 2 2 2
In transmitting the Myer code by means of a telegraph key and sounder, the 1 is a dot, the 2 is two successive dots, and the 3, which is the “front” or interval separating words and sentences, is three successive dots. With the wig-wag, a motion to the right of the sender is a 1, to the left a 2, and to the front a 3. With the flash or winker light by night, or the heliograph by day or night, the 1 is a short flash, the 22 is two successive short flashes, and the 3 three successive short flashes. On the fog-whistle the 1 is one toot, the 2 is two toots and the 3 is three toots. Both the Continental Morse and the Myer codes have really three elements, for in the former the pause or elapsed time between the words is as much a third element as the three of the latter, but the two codes differ radically in their character. The Continental Morse is founded upon three elements, or elements of duration, viz., the dot, dash and elapsed time interval, whereas the Myer elements are those of motion or distant duration. Hence the latter is essentially a wig-wag code, because the motions to the right, left and front correspond to the motions 1, 2, and 3. As to which elements, those of time or those of motion, answer best for naval purposes, the following comparison of the Continental Morse and Myer, as representing the best types of each, would seem to indicate that those of motion are superior to those of time. In the wig-wag both codes are transmitted with the same ease and rapidity, and no advantage rests with either. Indeed, as far as construction goes, we might convert one code into the other by making a dot a 1 and a dash a 2. On the fog-whistle it seems more easy and certain to transmit toots than to transmit blasts of varying lengths. A steam whistle, particularly in craft where the steam supply is not great, is not always reliable as to the emission of sounds. In the Myer code the sounds are taken as of equal value, although they may accidently vary in length, and the use of the "front" signal is an advantage, as it relieves the mind of all doubt and embarrassment as to the end of the word, as to abbreviations, as to end of sentence, and as to end of message. It also gives speed in signaling, as there is no need of pausing till the message is ended. Motions or short sounds, in groups of one, two or three, seem to lend themselves more readily to the eye and ear than dots and dashes. This may seem too subtle a distinction, but there are several things to be kept in view from the American naval standpoint. We have no trained signal corps. All the combatant force should know something of signaling. That which appeals to or strikes most readily the untrained eye and mind is best. Simplicity is the first consideration. For fog-whistle and heliographing the advantage in speed of signaling both in theory and practice rests with the Myer. By gun-fire signaling with the Morse code it is necessary to call the dot a 1 and the dash a 2 in order to transmit it, whereas the Myer code is all ready for use and needs no such modification. In naval signaling there can be no question but that the advantage rests with the Myer code as against the Continental Morse. It would be unfair however, not to point out one rather serious defect in the Myer. The front signal 3 is three toots on the fog-whistle and three short flashes by heliograph, which is the same as the letter Y. This leads to confusion, so when this code was in use in the navy the front was made by a blast on the whistle and by a prolonged flash with lights. The Continental Morse has no such "front." To introduce one would require the use of a long dash. This is objectionable on the fog-whistle and makes signaling slow. If we convert the Continental Morse into a Myer code by substituting a 2 for a dash, we would still meet with the same difficulty, as the letter s is three dots, and we would have to use a dash for a "front."
The general adoption in the principal navies of the world of some such system of night signaling as the Ardois or Kasolowski apparatus by means of permanent hoists of red and white electric lanterns would seem to call for certain changes in the construction of the Continental Morse and Myer numeral codes, because the lanterns are for certain practical reasons hereafter stated now generally limited in number to four, whereas the Continental Morse and the Myer numerals contain five elements each.
It will be noted that there are only thirty possible combinations of the dot and dash or 1 and 2 limited to four elements as a maximum. On the other hand, there are twenty-six letters and ten numerals or thirty-six characters in all, to be provided for. This means that some of the numeral characters must duplicate the letters. In the above arrangements the consonants h, v, z, j, b, p and x are duplicated in the Morse, and z, f, j, g, v, m and b in the Myer numerals. The intervals . - • - and 2212 occur in neither the alphabet nor in the numeral codes proposed. They separate words, sentences, groups of numerals, etc., and are in other words positive ''front" signals. The cornet is the general call to all vessels or stations in sight. When followed by an initial letter it calls a particular ship or station. The ''letters call'' indicates that the characters which follow are to be read as letters ; in other words, as a spelled out message or as consonant signals, as in the International Code. The "numeral call" indicates that the characters which follow are to be read as numerals. It is the "numerals follow" signal, and no character is to be taken as a numeral unless preceded by this numeral call. This is exactly what is done in the English semaphore code. The numerals duplicate letters. The code call ''numeral" is displayed when they are meant to represent numerals. It may be thought that in limiting the code to thirty characters too little latitude is allowed for call or code signals, such as telegraph, compass, international, action, etc., and for such special designating characters as interrogatory, affirmative, negative, annulling, preparatory, danger, etc. The following considerations, however, will show that. this limitation is largely imaginary. In the wig-wag code all the special significations are provided for in our navy, as follows: The code calls are T. D. U. (telegraphic dictionary use), A. S. U. (action signals use), I. C. U. (international code use), S. B. U. (signal book use), N. L. U. (navy list use), G. L. U. (geographical use), and F. D. U. (fleet drill-book use). These cover all the code calls for all other methods of signaling, except the Very's night code, which is more limited in its scope for reasons hereafter stated, and any number of code calls can be added to cover future contingencies, such as C. A. U. (cipher "A" use), C. B. U. (cipher "B" use), etc. As all other methods of signaling, other than the Very and flag hoists, are alphabetical, the interrogatory, annulling, affirmative, preparatory, etc., can be made by abbreviations, such as "prep." "interrog." or int." etc. This question will be discussed under each method of signaling, which is hereafter described, as will be also the question as to four lanterns in the Ardois or other night electric code. To summarize, we have in the Myer elements of 1 and 2 two almost perfect elements. Certain modern conditions would seem to demand the use of four elements in the numeral characters. If those conditions are imperative, we can easily accomplish this desired end. If a five element code is thought best or desired most, there is none better, both in theory and as demonstrated in practice than the original Myer code which we gave up in April, 1886.
METHODS OF SIGNALING AT SEA.
The patents taken out in the various countries of the world for devices for signaling at sea would, in their published specifications, fill volumes. Those which have survived the crucial test of actual trial have not been many, but have been sound in principle. Methods which conform to certain conditions are inapplicable to others, so that in naval signaling we have numerous methods. Where one code is utilized for all purposes, expertness is as much a question of ordinary intelligence as it is of practice.
As previously stated, all naval signaling divides itself as to character into two classes, squadron and distant ; as to limitations imposed by nature, into day, night, and fog signaling ; and as to method, into messenger, visual and phonetic signaling. Distinction is also made as to the relative permanency of any signal displayed—that is, whether transient (disappearing as soon as made, such as a sound) or permanent (displayed until read and understood, such as hoist flags). Bad weather somewhat limits the application and range of any form of signaling. Fog, mist, or snow prevent visual signaling and also some forms of messenger service. Night and day reverse the forms of visual signaling, and distance by day destroys the value of color and increases that of form, whereas by night it increases the value of color and destroys that of form. Experience has demonstrated the value of certain forms and colors for visual signaling and of certain methods for phonetic and flash signaling. Unfortunately the data on which certain recognized standards have been adopted is not available, but an enumeration of the methods of signaling at sea will at least show that it has developed along certain well defined lines.
I.—Messenger Service.
By Dispatch Vessels.—Written messages are sent by dispatch vessels or by the other auxiliaries of a fleet. There are no limitations as to distance. A distinguishing or dispatch flag is usually displayed by such vessel to indicate the character of her mission, and as a notice to others that she is not to be interfered with. Speed and sea-going qualities are, of course, requisite under ordinary conditions of dispatch service, and for distant cruising coal endurance is a prime requisite. The dispatch vessel has been developed as a special type in most navies, and, as an auxiliary of a fleet, when not employed in carrying dispatches, it is most valuable as a scout.
By Pigeon.—During daylight and for distances not over three hundred miles, communication by pigeon between vessels and the land is quite practicable. Between vessels at sea it has a much more limited application. Such communication is impracticable bynight and in a fog, but is not affected by cold, rain or snow, except in the matter of speed. In France satisfactory experiments have been made in communication between ships of a squadron out of sight of each other, and out of sight of land. In the French navy a regular pigeon service has been established and has been used in squadron manoeuvres. The birds are taught to alight or set out whilst the guns are being fired, so that they will be available for use in a general action. They know their own ship and fly with as much certainty at sea as on land. Abroad the whole question has passed the experimental stage, and pigeons are used, under certain circumstances, in the French, Italian and Austrian navies. In the United States, satisfactory experiments have been recently made from stations at Newport, R. I., Annapolis, Md., and Washington, D. C. The homing pigeon service, from experiments on the recent cruise of the Constellation, are said to have justified the hope that the service will be officially established in our navy. Several very important messages were sent this summer, and out of twenty-seven birds liberated all but two turned up, and these bore duplicate messages. About September 1, 1892, birds liberated from Fort Monroe covered the distance of 125 miles in excellent time, and all reached their destination. The steamer Waesland liberated a bird 315 miles from Sandy Hook at I P. M., and it was in its loft on shore the same evening. When released at sea, pigeons fly direct to land and then take their flight from shore bearings. On shore the longest record of one-distance flight is about 1100 miles, but in these cases numerous birds are let loose, and only a few reach their destination. Forty-six kilometers, or 28 ¾ miles per hour, is an average long flight, while 1370 yards per minute, or 46 miles per hour, has been recorded in exceptional cases. In Italy regular lines have been established between Rome and Madalena, and between Naples and Cagliari. The former is 150 miles long and averages five hours in flight, and the latter is 280 miles and averages nine hours. Experience has demonstrated that, to keep up frequent communication between such distant stations, from 200 to 250 birds are needed for each line of flight. By pairing birds in one place and feeding them in another, what is known as "there and back flight" is obtained, and with this it is possible to keep up a continuous to and fro flight between two stations. At sea this finds its best application with light-ships and off-lying stations in communicating constantly with the shore where the distance is not over 50 miles. In October, 1883, a lightship broke adrift from her moorings in heavy weather twenty-two miles from Tornung, off the mouth of the Eider. Four pigeons were liberated from the ship and brought the news in 58 minutes. The messages are usually carried in a section of goose quill, about 1 ¾ inches in length and hermetically sealed at the ends. Light wooden cases are also sometimes used. They are fastened securely to one of the underneath tail feathers. The birds are marked either on a light aluminum band around each leg, or by a stencil mark on the underneath side of the wing, or both. They are also sometimes spotted with certain dots of different colored paints on the back and wings. In the Austrian navy the only station is at Pola, where they have 120 pigeons. Several flights from ships down the Adriatic to the home loft have exceeded 250 miles. In Belgium the government has no regular lines of communication, but 600,000 birds are owned by private individuals or clubs for sporting purposes. France has some 100,000 pigeons. It is estimated that 25,000 pigeons would be required in war time for Paris and its outlying circle of forts. Switzerland and Austria are developing their systems. Berlin has two lofts of 500 pigeons each, and Ham, Metz and Strasburg each 1000 birds. In 1888, Germany had 52,240 birds available. Spain has 18 stations and Russia has numerous stations on her western frontier. Messenger service by pigeon is a most valuable method of distant signaling at sea, and should receive the development and encouragement which its great value merits.
By Balloon.—Under certain very peculiar circumstances balloon messenger service may be utilized for naval purposes, as, for instance, off a blockaded port, with a favorable wind to make a free ascent and descent either from shore to seaward, trusting to being picked up by friends, or from seaward to shore to communicate with friends in the blockaded port. The use of a sea anchor in making a descent, and drifting at sea till picked up, makes one phase of this hazardous service at least practicable under some circumstances.
Going from seaward to shore is not nearly so hazardous, and under some circumstances might be of great importance. Pigeon service would, in this case, serve most purposes. The general subject of ballooning, as applicable to naval signaling, is considered under the head of Visual Signaling, where captive balloons are utilized. This form is both feasible and valuable, and is described in detail. Of the recent French manoeuvres (1892) in the Mediterranean, the Army and Navy Gazette of August 20th says:
"A chief purpose of these manoeuvres was to test the working of the semaphore and carrier pigeon service, as well as the captive balloon. The semaphore stations proved their efficiency, and Admiral Brissonby, in command at Toulon, was kept au courant with the progress of affairs, and was able to direct the whole of the defence (against the supposed attack on that harbor). Doubtless these stations and their telegraphic apparatus would be the object of an early attack in case of war. The despatches carried by pigeons reached the commander irregularly, some being promptly brought while others were long delayed. As to the balloon, it burst and was useless." (This refers to a captive balloon.)
II.—Day and Night Squadron and Distant Visual Signaling.
Day Visual Signaling.—Day visual signals are, as to character, either permanent or transient. Bad weather limits the application and range of either form. Fog, mist, or snow destroys their usefulness for the time being, and phonetic signaling must then come into play. The elements of visual signaling are form, color, duration, position and motion. With flags we have both form and color; with shapes we have form only, although color may come into play at short ranges; with the semaphore we have position; with the wig-wag we have elements of motion, and with the heliograph we have elements of duration. Heliographing is not here regarded as a method of naval day visual signaling, as its application is strictly limited to use on shore. In messenger service we have seen that the same general means exist afloat as on shore, although more restricted at sea both as to range and rapidity of communication. As regards telegraphy there is no limitation as to distance, and the speed and capacity of signaling have been in recent years most marvelously developed, but from a naval standpoint the telegraph wire and cable are only indirect means of occasional communication from one base on shore to another. As previously stated, the development afloat has been in the direction of all around visibility. This points to the use of shapes for general squadron signaling. Mechanical difficulties are largely a drawback to the development of this method. Experiment is necessary and should be entered into seriously in our service. It promises much in the simplification of the code, as with the use of shapes the same code as used in the wig-wag and Ardois could also be used.
Flags.—Flags possess the great advantage that in one hoist three or four elements can be displayed at once, and for general signal purposes thousands of permutations and combinations can be made. The objections to the use of flags are that in a calm they hang limp; in an unfavorable breeze they fly on ; in haze and smoke the colors cannot be distinguished; the colors get dull and soiled through the effects of smoke, powder gases, sunlight and rain; and in action the halliards are so liable to get shot away. The greater the number of flags in a hoist the greater the chance of making a mistake in reading the signal ; also the longer it takes to bend on and hoist a signal. Using ID flags and no repeaters, the chance of mistake in a one-flag hoist to the chance of mistake in a two-flag hoist is as 10 is to 10x9, or using 20 flags and no repeaters as 20 is to 20x19. The simpler or fewer the colors in a flag the less liability to make a mistake in reading, particularly if the breeze is light. The smaller the flags the quicker they can be handled, but the more limited the range of visibility. Mechanical devices and practice increase the rapidity of flag signaling. Snatching the halliards and running away with them on deck; devices for bending on and detaching quickly; putting lead weights on to bring them down quickly; all these add to smart signaling. Using different shaped flags to indicate the character of the signal from the shape of the uppermost and the number in the hoist, leads to quick reading. Experience has demonstrated that the buntings which possess greatest visibility are of the colors red, canary yellow, black, white and blue. Much, however, depends upon the way they are combined, upon the character of the light, and upon the colors not fading easily. As between white and canary yellow, the former soils easily, and on the fly end of a flag lacks visibility. Red with blue is also a successful combination. White is a good intermediate color with red or blue in horizontal or vertical stripes, or in a square patch on a red or blue field, but it is poor as the fly or hoist of a flag or as a single color. Bunting should be rigidly tested in respect to fastness of color and non-liability to fade. No more than two colors should be used in any one flag, as the chances are increased of mistakes where the flag droops and only a patch of color shows. The five colors red, white, blue, black and yellow give a wide range of two-colored flags and pennants.
I. International Flags. By international agreement eighteen flags, representing eighteen consonants of the alphabet, are used in hoists of two, three or four flags to communicate at sea between ships, or between ships and shore stations, using a code book in which the meaning of each hoist is given. The total number of permutations possible, using more than one flag and not more than four, in which no flag appears more than once, is 78,642. Two flag signals having the letter B uppermost are attention signals; with a pennant uppermost, compass signals; and with a square flag uppermost, urgent signals. The three flag hoists are universal and express latitude, longitude, time, weights, numerals, and all ordinary sea signals. Four flag hoists, having the B flag uppermost, are geographical signals; with the pennant uppermost, are spelling and vocabulary signals and names of men-of-war; and with the square flag uppermost, they are the names of merchant vessels. Of these combinations, 1440, from GQBC to GWVT, are allowed for distinguishing signals for men-of-war, and 53,040, from H B C D to W V T S, are for merchant ships. The assignment of these combinations to the names of ships is left to the government of each nation within the limits prescribed. The French Government has recently proposed to the United States and Great Britain certain changes, which are numerous in detail as to the additional signals required, and also as to the modifications and corrections of signals now in use; for instance, to give each flag a special significance, as F, "end of word," K, "repeat," G, " do not understand," etc., etc. The main proposition, however, is to add two letters, X and Z, which are to be represented by two yellow and black pennants. This would increase the permutations from 78,642 to 123,500. In the main, the suggestions are in the direction of widening the scope of the code and increasing its efficiency. The increased speed of vessels and the rate at which they now pass one another at sea render quick signaling imperative, and the French are more keenly alive in these matters than any other nation. The British Government has submitted to France and to the United States a proposition to consider the extension of the international code to night signals and to adopt a phonetic code for certain purposes in foggy weather. No action has as yet been taken. In the Italian navy the international flags have been adopted for general signaling. The advantage to be derived from this is that using two and three flag hoists a total of 5,202 permutations are possible, which answers for all ordinary signaling; using four flags will add 73,440. Special code books are used, and only a few special and additional flags are needed for distinctive and code purposes. In the International Code no signal has ever more than one significance. The character of the upper flag and the number of flags determine its character on sight. Admiral Gherardi has submitted to the Navy Department a most excellent scheme for using alphabetical flags. As the international possess most of the virtues as to colors, etc., it would seem advisable to follow the example of Italy. It would simplify the study of signaling, and there is no advantage in making it difficult.
2. Naval Flags. Each nation has its code of flags which, when hoisted, convey special meanings interpreted in books specially devised and intended to be kept secret. As far as the secrecy of the alphabetical or numerical significance of each flag goes, it is of little importance. The key to the meaning is the important thing. The usual basis of the flag codes are the ten numerals, each represented by a special flag. Using 1st, 2d and 3d repeaters, the permutations and combinations using ten numerals and limiting the hoist to four flags gives 11,110. The English use a fourth repeater or substitute, which gives a five-flag hoist. They use altogether 58 different flags. The French have the most elaborate system of flag signals in the world. They have four sets of numerals. The first series are square flags, the second are "trapeze" shape, and the third are pennants. Separately each flag has a special meaning. The fourth is a telegraph series with nine square flags and five pennants. They can tell the character of signal by the upper flag. The total is 51 numeral flags. Their flag signaling is so complicated that there is an official movement on foot to abolish their use altogether and substitute shapes. In our navy our signal books are very antiquated and few signals have been added. We have nothing on torpedo-boats and little or nothing on any modern appliances adapted to new ships. The modern demand is for one, two and three flag-hoists. We can certainly get in the numbers from 1 to 999 all the important signals. In any revision of the signal book, now so urgently demanded, the index columns should be arranged for either numerals or consonants (that is for both). In case shapes supersede flags they will probably be on the basis of numerals, although not necessarily.
The Semaphore.—Many officers in our navy, struck by the efficiency of this method of signaling as used aboard British vessels-of-war, have suggested its adoption in our service. Its supreme virtue is rapidity of signaling. Each character is made by practically one motion of one or more arms, which assume a definite position at which they are read. It is a three element code, it has forty- four characters, and it duplicates, with its ten numerals, ten letters of the alphabet. It possesses the same defect as the wig-wag, that it can be read only from one direction. An attempt has been made in the British navy to use the semaphore for night signaling. To this end the arms have been illuminated by reflected light or by lights on the arms. On the Nile and Trafalgar, and also on the Chilian armored cruiser Captain Pratt, light steel semaphore arms (perforated for lightness) have been fitted at the mast-head for signal purposes. There are two sets, the upper one for abeam and the lower for ahead. They are worked by endless chains running up the hollow steel mast operated from the berth deck. This is the development of the semaphore in the direction of increased visibility, both in increased size of arms and in the altitude at which mounted, and represents really the semaphore for distant signaling. It is not regarded as a pronounced success. One reason is that the lost motion in the endless chains leads to wrong positions of the arms. For day signaling the arms are painted black. Where the semaphore has been used at night, the arms have been painted white so as to reflect the light thrown on them. The defects of the semaphore are that it is a complicated code, hard to learn and to remember, that it is only adapted to day signaling, that it is only visible in one direction, and that it cannot be used as a code for any other than the one purpose. In other words, it lacks simplicity and is limited in its range of usefulness. In the British navy, with its highly trained and well organized signal corps aboard each ship, and in the French navy, with the sea-coast defenses entirely in the hands of the navy, the question of simplicity is not so important as with us, where we expect so much from our combatant force. The code used has been explained. Using small hand-flags the code is transmitted as to method as follows, which is, however, the French semaphore code.
The French mechanical semaphore uses the same code as is used with the flag signals.
Wig-wag.—The elements are those of motion; the display of elements is successive and transient, and the method is slow. It possesses the defect that it is only visible in one direction. It is a most excellent method for our purposes because it can also be used at night by means of a torch; its elements are simple, it is easily learned, and with some such code as the Myer, it is adapted to all purposes of signaling. The code calls are initials such as T. D. U., G. L. U., etc.; the annulling, affirmative, repeat, error, etc., are special characters, and all- other significations are spelled out. It is slow, but it is very simple, and we would make a great mistake to give it up for a semaphore code. In the German navy, using the Continental Morse, a dot is made by holding one flag out, a dash by holding two flags (one in each hand), and a front by crossing both flags over the head.
Shapes.—In squadron day visual signaling the use of shapes is only a matter of experiment. No navy has finally adopted them. The French are conducting some experiments, but the only information given out is that collapsible shapes are used, and that these are mounted on three arms projecting in the same plane from a common pivot. They are turned towards the observer just as the semaphore is. This destroys or neutralizes all the advantages of such a scheme, because, if we use for shapes figures of revolution, then from every point of the horizon each shape looks the same, and if we mount them on arms and confine their visibility to any particular direction, we destroy the one great advantage shapes possess over flags. If Ave turn them so as to display them successively to all points of the horizon, then we get visibility, but at the expense of speed of signaling. Unquestionably, the development of signaling by means of shapes is not along the lines practiced by the French. Collapsible shapes should open by one motion and close by reversing it. They may be made of canvas fitted to steel or wooden frames, and the collapsing motion may be made by a spring. The two most satisfactory shapes, from a mechanical standpoint, are cylinders and cones. If only one shape is displayed at a time the elements are made successively and the method is too slow for squadron purposes. The writer holds that the Myer code, using with it the four element numeral code proposed, is the ideal naval signal code, adaptable to all modern conditions. Also that the code is of the first importance, and that the means of transmitting it somewhat secondary. To transmit this code by shapes we must have one shape to represent a 1, and another to represent a 2. To display all the elements of any character at one time we must be able to exhibit four shapes in one hoist or one display. The writer has unofficially submitted to Commander Sigsbee, U. S. Navy, whose inventive talent is so well known, a device for displaying shapes to conform to the Myer code, and it is his opinion that with several important modifications the scheme of shapes is feasible and practical mechanically. The device as modified by his valuable suggestions consists roughly of the following parts:
1. A signal mast of steel or wood with four grooves running its entire length, of which the positions are shown in B, which represents a section of the mast. This mast to be at least forty feet long, fitted with ordinary stays and backstays and with shrouds spread out or sprung out near the masthead to give clearance for making shapes on the mast.
2. Four shapes consisting of two light steel circular discs joined by a circular diaphragm of rubber, similar in general construction to a Japanese folding paper lantern. The lower disc is fixed to the mast, and the upper one can be raised by means of a cord which runs up a groove in the mast and over a sheave in the truck. When this is done the figure presented is that of a cylinder, as in C. When collapsed it takes the shape shown in D. Traveling on the mast is a loose collar inside of each shape fitted with steel ribs, somewhat like those in an umbrella. When the collar is held and the shape expanded, the umbrella frame opens out and gives the shape the appearance shown in Fig. E. The cords or wires which operate the display and collapse of the figures travel in the grooves in the mast, and are worked by levers either in the tops or on a signal platform, or from deck. [It would be practicable, as in the English semaphore apparatus, to run the halliards and other signal gear inside a hollow steel mast and work the shapes from the berth deck.] Calling Fig. C a 1 of the Myer code, and Fig. E a 2, such a display as Fig. F would read 1 22 1, that is, the letter M or the numeral 9 of the modified code. The discs might be made from 2 ½ to 3 ½ feet in diameter, and the shapes given a hoist of from 4 ½ to 6 feet. They would be easily visible from four to five miles from any direction of the compass.
If the shape representing 2 should be found by experiment to offer mechanical difficulties in operating the umbrella-like frame, we might substitute for it a collapsing ball, which would ordinarily house around the mast, inside of the space covered by the shape I when it is hoisted. This ball, when collapsed, would consist of a closely packed bundle of flat steel ribs surrounding the mast. When the upper collar is forced down towards the lower one, the steel ribs would take the form of a globe, which would represent a 2 of the Myer code, or a dash of the Continental Morse. The greatest difficulty in working shapes is experienced in blowy weather. The effect of the wind is to bind the steel discs, in the shapes proposed, to the mast, and prevent their being hoisted or collapsed. The effect of the mast is to steady and guide the shapes, and the form proposed would seem to offer less objection in bad weather than any other. British vessels-of-war are fitted with collapsing drums, which are, rigged some 20 feet or more above the deck to a gaff of some kind, and are opened and collapsed on the dot and dash principle of the Morse code. , They are barrel-shaped and, when expanded, the dimensions are as follows: height, 4 feet; greatest diameter, 3 feet 6 inches; diameter of heads, 2 feet 8 inches. They are rigged as shown in sketch. The drum is steadied by the guys D, and raised and collapsed by the ropes C. and B. A is weight of 10 pounds to collapse the upper half by gravity. As shown in the sketch the drum is only partly collapsed. When completely so the thickness is only 8 inches. This is a device for fleet signaling at as great a distance as has yet been found practicable in the British navy. During the manoeuvres recently, these drums were found to be much too small. The subject will be discussed further under the head of "Distant Day Visual Signaling." It is a very slow method of signaling, but it represents one step in the direction of the use of shapes, and is a further movement on the part of the British navy away from the semaphore for anything like distant signaling. If used for distant signaling, powerful tesescopes are needed, and the height above deck must be increased.
The shapes proposed by the writer, operated on a mast specially constructed for the purpose and of sufficient size to be visible five miles, would seem theoretically to possess the following advantages: 1. Visibility from every point of the compass; 2. Ability to transmit an alphabet or numeral code; 3. Simplicity and uniformity in that the same code can be used for this as for all other purposes. If flags are to go, some such system must take their place.
Distant Day Visual Signaling.
Distance destroys the value of color and increases that of form. In our navy our largest numeral signal flags are eleven feet long. At a distance of between four and five miles, under the most favorable circumstances, with strong glasses, the limit of visibility as to color is reached. The shapes of the flags can be made out at five miles or more. The development of methods of distant signaling must, therefore, be in the direction of shapes of some kind. The dip of the horizon, of course, limits observation, so that any method must fail at a certain distance. Before discussing the question of apparatus, it may be well to outline briefly some experiments made abroad in naval ballooning for signal purposes. A captive balloon carrying a signalman gives, of course, an increased range of visibility for observation and signaling.
Firsts as to some recent experiments: On Sept. 6th, 1890, the torpedo-boat L'Audacieux, with a captive balloon, attached by a short cable, steamed from Toulon harbor to Hyeres roads, a distance of twenty-one miles, in two hours. The balloon was transferred to the St. Louis, the gunnery ship, and the captain ascended and gave orders to his ship by telephone at a height of 250 meters. Lieutenant Serpette ascended, cast off the cable, and made a free ascent of 1800 meters. He came down in the open sea, and by means of a sea anchor, hereafter described, he waited till he was picked up by a torpedo-boat, which towed the balloon in. Afterwards the balloon was taken on board the Formidable, and was filled behind the armored turret on the after deck, hauled to the mizzen-top, and several officers successfully made a captive ascent. They ascertained that in clear weather all the details of the coast from Marseilles to the extreme point of the island of Hyeres were plainly visible, and that no building or ship for 30 or 40 kilometers around could escape the notice of an observer in a balloon. The bottom of the sea to a depth of 25 meters was clearly distinguishable, and the movements of a shark were watched with interest. On the high seas no special value may be assigned for balloon service for reconnaissance, but for the operations of a fleet near a coast, in attack or defence, it may be made to play an important part. From a signal standpoint an observer in a balloon could telephone to the deck any signals or movements displayed ashore, or could, by signal flags, or the wig-wag, communicate with other vessels of the fleet at distances much greater than could the ship itself.
For night signaling, as hereafter described, the balloon offers several valuable methods. Balloons for French naval purposes are made of Pongee silk, with several coats of varnish. They are spherical in form, and have a cubical contents of about 320 cubic meters, with a diameter of 8.5 meters. They carry a wicker basket and a steel cable with a telephone wire in the heart. [English military balloons have cables 1- in. in circumference; breaking strain, 1 ton; weight, 10 oz. per foot. The English telephone is a patent and does not require a battery.] They are filled with hydrogen gas, which is stowed in steel cylinders holding four cubic meters of hydrogen, under a pressure of 120 atmospheres. A tube filled weighs 30 kilogrammes. [English military balloons, hydrogen tubes, 70 lbs., 120 cwt. gas.] The sea anchor consists of a large water-tight bag, stiffened above and kept open by an iron hoop to which the anchor line is attached by means of a linen band. In the bottom of the bag is a valve which can be opened from the car of the balloon by means of a cord. The balloon is usually anchored with 40 or 50 meters of line. The weight of the water in the bag prevents the balloon rising, and keeps it captive unless the valve is opened to free it. Ballooning, except by experienced observers, offers little for an immense outlay of money, time and storage space. Wind is the great enemy of the balloon, and sea-sickness, escaping gas, excitement and inexperience are apt to neutralize most efforts, unless a good deal of practice is had in time of peace. Accidents are not infrequent, and the expense is great, but at critical times they fill an important place.
Distant Day Visual Signaling.
The French use, as their distant code, large flags and a black ball or shape. Form of flag and not color is what counts.
French ships, in communicating with distant semaphore stations, use a cone in place of a ball. In communicating with one another they use the ball, but can use the semaphore code.
The International code of distant signals is as follows: There are three symbols, (1) a ball, (2) square or rectangular flag, and (3) a triangular pennant. The alphabetical characters have each three symbols. For instance, B. With the coast semaphore, used by all nations by international agreement, an arm pointing downward represents a pennant; an arm horizontal, a ball; and an arm upwards, a flag. The disc on top at horizontal is the answering signal, and the disc vertical indicates that distant signals of the International code are being used.
In the British squadron manoeuvres of 1891 the collapsing drum issued for service, as described, was found too small for distant day visual signaling. Aboard different ships different devices were tried. On some vessels this drum was used in conjunction with large national flags folded as pennants or as narrow flags, some 22 feet long and about 4 feet broad. With these some prearranged code was used. The difficulty was that the flags did not show except in certain directions, being thick and only held out in a stiff breeze, and then were only visible in certain positions. On other vessels immense collapsing drums were built of timbers and. canvas, which required the entire watch on deck to hoist and collapse, using the Morse code. The general drift of all the experimenting was that it is impossible to communicate accurately or with any satisfaction at greater distances than those at which shapes of about 6 feet in height are plainly visible. In other words, shapes larger than 6 feet are too difficult to work and collapse in any kind of a breeze. The question is yet in an experimental stage. It has been proposed that sails be used—for instance, as in Figs. A, B, and C—to transmit the Myer or Continental Morse code. Fig. A, with the heads hauled out, would, for instance, represent a 1 or a dot ; Fig. B a 2 or a dash; and Fig. C a 3 or a ''front." Of course the sketch is only a rough outline, and such a scheme appears fanciful and unseamanlike. Consider, however, with a fleet undertaking actual operations, how necessary it must be at times to communicate with the scouts or with vessels at a distance. Some means must be devised. However much we may respect the traditions of the service, we all know that most of the spars in such ships as the Chicago and Newark would go ashore in time of war. In ships which are now building we ought to design the spars with due regard to this vital question of signaling—not, of course, to rig such an apparatus as shown in Figs. A, B and C, which would simply be used in an emergency and rigged with the resources on board, but to seriously get rid of top hamper, to consider questions of all-around visibility of signals, and to consult the demands of modern conditions in general. As has been said, the question of distant day signalingis in the experimental stage. Nothing satisfactory has emerged from it all. The chief defect which is likely to develop itself in some such system of shapes as proposed by the writer is that it will probably be found that refraction will .distort the shapes at long distances and render distinction between them a matter of considerable uncertainty. Any such defect is, of course, fatal to distant signaling. Experiment can alone settle these points.
Night Visual Signaling.
The utilization of the electric light has added vastly to the rapidity and range of night signaling for squadron purposes. Night increases the value of color as an element in signaling, and destroys that of form. With powerful lights and with increased altitude to overcome the effects of the dip of the horizon, the range of night signaling may be made to far exceed that of distant day signaling.
Wig-wag.—Night signaling with the wig-wag code by means of a torch finds now only a limited application, and that for special purposes, as the introduction of some such system as the Ardois has practically superseded it, owing to the advantage the latter offers in rapidity and accuracy of signaling. With the electric apparatus all the elements of a letter or character are made in one display, and, being answered by the same display by the receiver, mistakes in receiving are eliminated. In the army several improved types of torches for night wig-wag have come into use, and have also been adopted in the navy. With the electric light, however, an incandescent lamp on a staff offers considerable advantages over the torch, especially in windy weather. In transmitting the wig-wag by night, a lantern is placed at the feet of the sender to mark his position.
Semaphore.—Experiments have been made in the British navy, as previously stated, with several devices for lighting up the semaphore arms. One plan has been to paint the arms white and throw on them a reflected light from a group of incandescent lamps. A semaphore for night work has also been tested, in which the arms carry a number of incandescent lamps to outline their positions. Neither plan is regarded as entirely successful, and the winker light is generally relied on for night squadron signaling.
Winker or Flash Lights.—In both the British and Italian navies electric lamps are used with some device or shutter to flash dots and dashes, using the Continental Morse code. To get increased range of signaling, lamps of fifty and a hundred candle-power have been experimented with. The great defect has been found to be that with such high-powered lamps the carbon does not flash or die down quickly enough. This consumes time, as the dots and dashes have to be made much longer than in telegraphy, but the Italians have overcome the difficulty somewhat by using a globe lantern in which a nest or group of six or eight 12 or 16 c. p. lamps are used, all worked with one key in the same circuit. These are mounted on the trucks of torpedo-boats and very high on other vessels. The smaller carbon filaments are found to work successfully, and the fixing of the lamps rigidly on the trucks of smaller vessels has been found to be a great advantage in a sea way. In the British service one device to overcome the question of the slowness of the carbon has been to have a multifiber filament, that is to say, one made up of a bundle of smaller ones. This has reduced the defect very considerably. Another device has been to keep the lamp burning and operate shades to obscure the light. On some vessels this shade is simply a bag fitting over the lamp, operated by halliards on deck, which are attached to a lever which is worked up and down. A very successful device, invented by Captain Scott, R. N. , has been tried but not adopted. It consists of a high-powered lamp, surrounded by a number of vertical shutters worked by an electro-magnet and a spring. The shutters revolve through 90° when a current is sent through the electro-magnets by the signal key, and the light is exposed all around until the current is broken, when a spring snaps the shutters to and obscures the light. The lamp has a resistance in it which is thrown in when the shutters are in an obscuring position, so that the lamp does not burn at full power. This saves wear on the filament. In our navy no regular device of this kind has been adopted, although some experimenting was done on the Trenton and since carried on in the newer ships. As long as we have the American Morse as the service code, no serious attempt is likely to be made to perfect one. The Ardois or some such system is so much more rapid and certain than winker lights that the latter can only have a limited application. In a series of experiments off Montevideo in March, 1892, it is recorded that a 32 c. p. white light, at the main truck of the Bennington, was visible on a clear night over eight nautical miles, which shows the value of a truck lantern for distant wig-wag, where the lower lights of the Ardois would be hull down.
Coston Signals.—In the absence of any better system the Coston lights were at one time widely used for naval and commercial purposes. Recently an improved percussion arrangement was added to these signals, and the Navy Department was prevailed upon to order a test. In March, 1892, a board was appointed on the South Atlantic station to make a series of trials of the Coston, Ardois and Very systems on board the Chicago and Bennington. The night selected was a dark, star-light one, with atmosphere unusually clear and free from haze or mist near the horizon. Of 50 Coston lights burned, at distances varying from three to eight miles, and at a mean height of 32 ½ feet on each ship above the water, 29 were correctly read, 20 incorrectly read, and 1 was not seen. Eight miles was the limit at which the dip began to interfere seriously. Nine signals were marked doubtful on account of the confusion of white and green, as the shade of difference was very slight. A blue light was read as green. The signals averaged 30 seconds in burning each color of which it was composed. The finding of the board was that "The Improved Coston Night Signals are entirely unsuited for naval use. Forty-two per cent, of them failed to be correctly read, the color distinction being very poor. This is a cardinal defect, and would alone be sufficient to cause the rejection of a signal for military purposes; but in addition thereto the Coston signals show the following minor objections:
1. They require a long time to burn, and signaling with them would be much slower than with the Very signals.
2. They make much dirt, the products of combustion flying about the deck and defacing paint work.
3. They make a blinding glare, rendering it impossible for those standing about to see anything inside or outside of the ship.
4. They are made only by the manufacturers from whom renewed supplies must be ordered, and cannot, like the Very's, be reloaded on board ship."
Electric Night Signaling Apparatus.—An experimental apparatus and night code was used in the North Atlantic squadron in the summer of 1890 and until the latter part of 1891, the invention of Lieut. H. McL. P. Huse, U. S. Navy. It was intended for night signals with colored lights, and day signals with collapsing shapes; for red light using a cone, and for green light a cylinder. The following is the description as issued for trial:
In this system, one red and one green light are necessary. They must be placed sufficiently far apart that each may be distinctly visible.
The Very code for night signals may be used in connection with any of the signal books, or a message maybe spelled out by the American Morse code.
The general call is made by exposing a green light until it is answered by all the vessels within signaling distance.
The special call is made by signaling the initial letter of the vessel called until answered.
The answer to a call, or I understand, is made by showing a red light.
The repeat, or, I do not understand, is made by showing a green light.
Note—This change is made to preclude any mistake as to what vessel is making the call. If the answering were green, attention might be directed to an answering vessel instead of to the calling vessel.
Attention may be drawn by firing a red Very signal.
The lights will be made in succession, according to the Very code accompanying this circular.
Use telegraphic book: Show both lights together twice, followed by the designator G G R G.
Note.—Lights should be exposed about two seconds and be separated by an interval of about two seconds. It is not intended that measured time elements shall enter. Times of exposure and of darkness are given simply as a guide. A little practice will enable a signalman to make his lights and intervals regular; this is all that is necessary. Signals are separated by the designator.
The completion of a signal is indicated by making two designators.
Use General Signal Book, make three doubles.
Use Morse code. Expose the red continuously; while green is shown intermittently twice; then proceed with the message, a dash being green and a dot red. Words are separated by a double.
Two dots separated by a space are made thus: the red is exposed continuously for about six seconds; during the third and fourth seconds, the green is also exposed.
"Error, I have made." With any code, expose the green continuously while the red is shown intermittently three times. If using the Very code, begin again with the designator preceding the error. If using the Morse code, begin with the last word correctly sent.
To break, show a green.
To start after breaking, show a red.
Instead of the red and green lights, any other recognized means may be used in answering, as a torch, or two lanterns, or Very signals. In using Very signals for this purpose, the signals fired must conform in color with the lights called for.
The objections to this system are that it uses a green transmitted light, it is very slow, and with such apparatus as the Sellner, Ardois, or Kasolowski, it compares most unfavorably both in rapidity and certainty of signaling.
A brief discussion of the theoretical requirements of an electrical night signal apparatus will throw much light upon' the lines along which these modern systems have been developed. In the first place, as to colors: Three have been more or less used, green, red and white. Green has now been entirely discarded. As a transmitted light it lacks visibility and is confused with white. Red and white have come into general use, but in the most recent apparatus a pulsating white light has been substituted for the red. In a French experimental apparatus, as previously stated, both red and white pulsating lights have been introduced. In the face of the exhaustive experiments abroad we are compelled to accept the dictum that a white light is visible further than a red (which also conforms to theory) and that a pulsating white light is visible further than a fixed white. As to types of lanterns used: The range of visibility of a light at sea in clear weather depends upon the height of the lamp above the water, its candle power, and the character of the lense or reflector used. Some apparatus limit the visibility to a certain direction by putting in shades and reflectors. The requirements of a lantern are as follows:
1. It should be double, and provided with a diaphragm to exclude the light of one lamp from the other half of the lantern.
2. But one lamp in each lantern should be displayed.
3. The lantern should be air-tight to exclude moisture and powder gases which erode the terminals and contacts.
4. It should be fitted with a Fresnel lense, which concentrates the rays in a horizontal plane.
5. It should give visibility all around the horizon.
6. It should be so suspended as to always take a vertical position in order that the rays of light may be thrown out horizontally, but not so freely as to strain the leading wires or chafe the main cable.
7. It should be simple in construction and easily opened to replace a lamp burnt out.
As to candle power of the lamps: We generally use 32 c. p., which gives excellent results. The French use 30 c. p., and the Italians 25 c. p. High power, of course, gives increased visibility and distinctness. As to cable: In the wiring of signal lamps the leads are generally made up into a cable and protected from chafe by a covering. A common return is generally used and separate leads to each lamp. In connecting with the signal box or key-board the attachment should be made by a coupling and not by separate wires to separate binding posts. The coupling facilitates testing out the circuits. The cable should be more or less permanently suspended from a yard-arm or working gaff, and secured at intervals to a taut guy or backstay to take its weight. In the Italian navy the cable is sometimes suspended to the fore and aft horizontal stay between the military mastheads. There should be a switch in the common return, working automatically, so that the current will not be thrown on till after the contacts are made for signaling. This prevents sparking, and consequently burring and wearing of contacts in the signal box. As to signal box or key board: Contacts which are made against the face of a spring are not reliable, as the springs weaken in service. The box should be water-tight, and the movement of the key arm should be as simple as possible to give speed in signaling. Boxes are sometimes set at an angle to facilitate operating the lever or arm. The face of the disc should be marked opposite each sector as to the display of lights and the significance thereof. As to the number of lights in the permanent hoist: The French at one time with the De Meritens apparatus, used eleven. The two upper and the two lower lights were red, and the intermediate were white. On account of the blending of the lights, owing to their close spacing, the range of visibility was very limited. This system was superseded by one of four single lamps. Experiments are now being conducted with four double lanterns. The Italians and Austrians have adopted four; Spain is experimenting with four; and Germany has adopted three (all double lanterns). The British Government has ordered a three-lamp apparatus for experiment, but the system meets with but little favor. We have practically adopted a five double-lantern hoist, in that we have committed ourselves by contract in the ships now building to the use of the Ardois or some similar apparatus. The question involved is this: The board organized on the South Atlantic Station to test the Ardois and other systems report that "Under ordinary circumstances the Ardois signals can be read with certainty at a distance of four miles. By actual experiment the Ardois signals of the Chicago were read from the Bennington five miles distant with night glasses, the lights being distinct both as to number and color: "But the Chicago lights are 32 c. p., spaced 5 meters, or 16 feet 5 inches, apart, and the lower lantern is not less than 35 feet above the water line. We fail, however, to realize that our more recent ships have no such spars, and that a modern squadron is made up of all sorts and conditions of craft. The auxiliary vessels of a fleet, the despatch vessels, the torpedo gunboats, the torpedo-boats and repair vessels, the swift commerce destroyers, and also the low free-board battle-ships offer no such hoist for signal lanterns as will allow the upper lantern a height of 100 feet above the water line. If our service code requires it, and if we want five lanterns, we must, of course, spar all our vessels with some regard to the requirements of space between lanterns. With five lanterns, the scouts of a fleet, the vessels on which devolve the reconnaisance, and on which depend the rapid and accurate transmission of important intelligence to the main body, will be just the ones whose hoist of lights would be most limited in space and hence in visibility. With four lanterns instead of five, the maximum distance would be greatly increased.
Devices Used Abroad for Night Signaling.
Berg' s.—This has been experimented with in the German navy. It consists of three oil or electric lanterns with a white shade fixed in front of them, and with a movable red and a movable green shade so operated by electro-magnets as to throw at will a white, red, or green light. Where electric lights are used, a small hand dynamo gives the current. With this apparatus twenty-seven three light signals can be made. The defects are that it lacks range; it gives visibility only in one direction ; and it uses the green transmitted light, which is limited in visibility and leads to confusion with the white.
Conz.—This apparatus is patented in England, the United States and Germany. It is a German invention. The last named government has ordered loo sets for installation, which practically amounts to adopting the system. Three double lanterns are used with the colors red and white. The white light has an opal shade, but no device or lense is used for intensifying the power of the lamps. Resistances are used to keep the required current constant. Fourteen characters can be transmitted by this apparatus.
De Meritens.—This is used in the French navy. It has eleven lamps on a backstay, each lamp having a separate key. There is a tell-tale box to show what lights are on. The two upper and two lower lights are red, and the seven intermediate are white. This device lacks range, and is in a fair way to be superseded.
Ducretet.—This was used in the French navy. It consists of four single lanterns transmitting fifteen different signals, showing, however, only in one direction. It is out of date. The Sautter and Harle apparatus uses the same code as the Ducretet but is simpler in design, and has a switch in the circuit to prevent sparking.
French Experimental.—This is being tried in the French Mediterranean fleet. It consists of four double lanterns, half red and half white; it has eight separate leads and a common return; and it uses a pulsating or fixed current at will. The pulsations are given by a hand key. It uses a four-element code: (1) a fixed red, (2) a pulsating red, (3) a fixed white, and (4), a pulsating white. Each display has four lights. If less are shown, an error is evident and the signal is disregarded. This apparatus is quite up to date and gives a very flexible code.
Kasolowski.—This apparatus is used in the Italian navy. It has four double lamps of 25 c. p., and for torpedo-boats has a space of 5 feet between lanterns. Each lantern is double, the lamps being carried in a bronze frame which is held rigid between two guys which are slung from a gaff or yard-arm and set up taut on deck. For vessels not provided with an incandescent plant a hand dynamo operates the circuit.
Sellner.—This is unquestionably the simplest and most widely used night apparatus yet invented. In one form or another it is found in both the Austrian and Italian navies, as it is practically the same as the Kasolowski, and similar to that used in the German navy. The latest model is, however, a great improvement on the older forms. One of these was purchased by our government and mounted on board the U. S. flagship Chicago in December, 1891, for trial in comparison with the Ardois. As it was fitted with four double lanterns, which is largely the custom abroad, instead of five, it was for convenience marked with the characters of the International code, using a few special and code calls. No lenses are used in this type in the lanterns, but instead plain red and ground white glass shades are fitted. With the apparatus sent two screens or blinds were attached to each lantern so as to make the display in certain directions, this being the custom in the Austrian navy. As these blinds can be easily and readily removed, they form no real part of the apparatus unless desired. They are simply prescribed by the Austrian naval authorities in their contracts. There is a device fitted to the transmitter or key box by which small resistance lamps may be thrown in or out of circuit to keep the current constant. These are intended for use with a hand-power dynamo, but can also be used with a regular dynamo as a tell-tale showing the signal displayed. The official report was made by Lieut. S. A. Staunton, U. S. Navy, as to the comparative merits of the Sellner and Ardois apparatus, and he objected to the Sellner as follows: 1st, to the marking of the instrument with the International code as not giving a fair test with the Ardois, which uses a full alphabet; 2d, to there not being thirty displays provided for, which is the maximum for four lights; 3d, to the use of the screens or blinds, as destroying the greatest value of the apparatus, viz., all around visibility; 4th, to the lanterns not being provided with lenses for strengthening the lights; 5th, to the awkward motion of the switch or lever on the transmitter for signaling; 6th, to the use of the resistance lamps as adding complications; 7th, to the cable not having a coupling for ready detachment for testing out the circuits; and 8th, that in point of general workmanship it is inferior to the Ardois. This represents, of course, an individual opinion. Against it must be set the fact that one of our best authorities in naval matters pronounces the Sellner apparatus the simplest and most endurable in service of all those in use in the various navies. It may not compare in workmanship, but it costs 5700, delivered in this country, as compared with $1100 for the Ardois. The transmitter can easily be arranged for displaying thirty characters, and, using the Myer, Very, or similar code, can transmit an alphabet or numeral code without difficulty. Lieut. Sellner, of the Austrian navy, the inventor of the apparatus, replies to the remaining objections of Lieut. Staunton, as follows: The blinds form no necessary part of the apparatus. The signal switch or lever is purposely made of not too easy manipulation to assure the attention of the operator. After numerous trials the use of lenses with the lanterns was abandoned on the score that such arrangement failed to increase the range of the signal. This range depends upon the distance the lanterns are spread from the others in their hoist, and if this distance is 4 meters, the signals are well visible for about 5 miles. Where the intensity of the incandescent lamps reaches 32 candle power there is no use of increasing this intensity, since no longer range of the signal is thereby attained. The use of a coupling in the cable is objectionable as introducing a break in the conductor, and hence an element of unreliability. The principle of plug contacts in the transmitter is where this apparatus is superior to any other yet devised. In large ships there will always be two complete night signaling apparatus, thus avoiding difficulties arising from masking of lights, failure of lamps, etc. These arguments on the Sellner apparatus are here given in detail to show that this question is not to be settled off hand. The original Ardois apparatus has many faults, which in new designs in this country are being remedied. The field is an open one. The problem is merely to display certain combinations of lights. Experience abroad cannot be ignored. Where we are making a mistake in our navy is, on the one hand, in not fully recognizing the immense value of such apparatus, and on the other, binding ourselves to the use of five lanterns instead of four, and to the use of an extremely objectionable and unnecessary code.
Massari.—This is a new Italian apparatus on trial in the Italian navy, similar in some ways but simpler than the Kasolowski. The transmitter or key box has a hinged lever which lifts and sets down at the proper sector, making contact against the face of a spring, which is an objectionable feature. A Morse printing apparatus receives the signal made on slip of paper. The same types of lanterns and the same code is used as in the Kasolowski.
Ardois.—This apparatus has been adopted in our navy and, in all, fourteen sets have been ordered from France. For the ships building at navy-yards the contracts are let in this country, and many improvements are being introduced by the contractors, and by suggestions from the office of Electric Lighting, Bureau of Equipment. The foreign type, as mounted first on the Chicago for trial, was described by the writer in No. 58, Proceedings Naval Institute, 1891. The new features are the lantern and key box or transmitter. The new lantern is air-tight and water-tight. A special feature is the readiness with which a burnt-out lamp may be replaced. Another feature is in the lead of the wires through the center of the top and bottom, enabling the lamp to be suspended at its middle point to swing freely and remain vertical, thus insuring the transmission of light in the horizontal plane. If a lantern with a Fresnel lense is canted, the light is not visible as far as if the lantern is vertical, and hence the rays of light horizontal. A further feature of the mount of the lanterns is the use of two parallel stays to which the lanterns are secured. The contacts in the key box are the same as in the old, but the box is water-tight. There are sixty-four sectors, but the marks radiate from the center, and the lever works around the circumference of the box. The pistons which make contact with the bosses or stops are beveled and rounded to give good contacts. It is, however, a grave mistake to use the so-called Ardois alphabet. It is inferior in every particular, save one, to the old Myer code. As to the folly of adopting five lanterns instead of four, it need only be pointed out that when we go on paying $1000 a piece for such apparatus for vessels like the Vesuvius that cannot mount five lanterns, and also for the short-masted new ships (even like the Philadelphia), and when we are committing ourselves indefinitely to the use of codes that violate every principle of signaling, one voice of protest is not enough to arouse the least official interest in the subject. When, however, at our own invitation, we assemble a fleet in our harbors next May, we will make an exhibition as to signals that will give us no reason to congratulate ourselves. The Ardois apparatus is an excellent one; but the field is an open one, and we want the best. There is, however, no excuse for the existence of the Ardois code, and there is no hope of the Myer code being re-adopted.
Distant Night Visual Signaling.
As previously stated, night destroys the value of form and increases that of color. To increase the range of night signaling, we must overcome the dip of the horizon, and must increase the power of the lights used. All around visibility is a prime requisite, and the conformity of the code with that used in squadron cruising is very desirable.
By balloon.—Balloons give increased altitude, and hence increase the range of signaling. Mr. Eric Stuart's signaling balloon, tried recently at Fulham, England, is made of varnished cambric, and is therefore translucent. It is i8 ft. in diameter, has a capacity of 3200 cub. ft., and is controlled by a cable 500 ft: long, weighing 35 ½ pounds. The balloon contains six incandescent lamps, each of 10 candle-power, and the electricity is supplied from 26 accumulators by means of wires running up the cable. The signaling key has carbon contacts. England and Belgium have purchased for army purposes the Bruce Electric War Balloon, which is made of a translucent material, and is operated similarly to the above. At Heligoland, in 1891, a German squadron used, with great success, a captive balloon for distant and squadron night signaling. Incandescent lamps attached to the underneath side of the car were operated by a key using the Morse code. Arc lights were also used, and, besides lighting up the surrounding harbor and shore line, the beam was used for distant signaling.
Search lights.—The beam of search lights thrown on low, heavy banks of clouds offers a feasible and useful method of distant night signaling. With vessels hull down and practically below the horizon to each other, it is the only method possibly available. Search lights are fitted with shutters for this purpose, and such signaling occupies a definite place in any scheme of night signaling. In August, 1892, a search light used on Mt. Washington, N. H., transmitted a message which was read on the clouds overhead at Portland, Me., 85 miles distant. The beam was thrown upwards at an angle of 45°, and it has been estimated that to reach Portland it must have been no miles long, and was reflected from clouds 80 miles overhead in the last named city.
Very’s signals.—Red and green stars are projected from a pistol to a height of 100 feet or more. This height is limited by the character of the stars, which cannot be made tough enough to stand the increased velocity and pressure of larger charges. Exhaustive trials have been conducted under the supervision of Lieut. R. T. Mulligan, U. S. Navy, whose report has not yet been adopted as to the recommendations made. As to colors, all shade into either red or white at any distance, so that only two are practicable. Green was recommended and is used, but white is quite as good. All cartridges deteriorate in service. The original bracket code has not been formally abolished. Lieut. Mulligan's four-element code is issued experimentally, and is very successful. The firing pistol is still issued. It is inferior to the short double-barrelled shot guns recommended. The green star should be changed to a white. This is not particularly important, except for the sake of uniformity with the Ardois. It is not feasible to transmit the alphabet by the Very code, as it takes too many cartridges. In the experiments made by the board of officers in the South Atlantic, previously referred to, 13.7 nautical miles must be taken as the practical limit of the present Very outfits furnished ships. As eight miles is given as the limit of deck signals, the five and seven-tenths miles of further visibility must be due to the projected height of the stars. The official report of the board states that "The Very's signal meets all requirements for long distances. It is slow but certain, the color distinction being, excellent; and with the lights grouped in fours to avoid bracketing, repetitions are seldom necessary. It is not suited for tactical signals because it is slow, and because it is not absolutely certain that the signal has been correctly read, and an error in tactical signals may produce collision. Certainty of correct transmission can only be assured in a night signal by repeating each display back to the sender. The scope of the Very is limited by the number of cartridges supplied. A long telegraphic signal would quickly exhaust the allowance. These cartridges can be reloaded on board ship, red and green stars being supplied for the purpose."
This excellent system of signals is capable of much improvement in detail. With brass reloading cartridge cases, excellent pistols or guns, and a four-element code, we would have the best distant night method in the world. The French use Coston signals, but supplement it by a code of lanterns hoisted in groups similar to flag signaling.
III.—Phonetic Signaling.
In a fog visual signals are, of course, useless. The only resort is to sound, using some phonetic code transmitted by drum, bell, gong, horn, bugle, whistle, siren, or gun-fire.
Squadron Phonetic Signaling.
Fog-whistle.—The steam whistles of vessels-of-war should be fitted with elbows in their pipes and drain-cocks for running off the condensed steam. Hand levers should be fitted to take the place of whistle cords. These should be capable of unshipping when not in use. Fog-whistle drills should be instituted in squadrons to perfect officers and men in the use of a code. Canvas screens should be put up around those on the bridge in charge, and at the wheel, and the conditions should be as nearly as possible those of a fog. Such drills are most important and valuable. At present in our navy we have no authorized method of transmitting our service code on the fog-whistle. As this American Morse code is dangerous to navigation, it is probably just as well that no method is prescribed. All vessels should be fitted with audiphones or other mechanical means of locating the direction of sounds in a fog. The importance of a squadron keeping in touch in foggy weather, and the danger thereof, unless an efficient phonetic code and apparatus are used, would seem to warrant some official interest in the matter.
Using a bell, gong or drum, a dot is one sound and a dash two successive sounds. The superiority of a four-element Myer code over this or any other dot dash numeral code is most apparent.
Distant Phonetic Signaling.
Rapid fire guns offer a good means of transmitting a code at a distance in a fog. It is best to limit the code to numerals, although with an interval character like 2 2 1 2 it is possible to transmit an alphabet. A numeral code would, with a telegraphic dictionary, effect the same purpose. The French officially prescribe the above code for gun-fire purposes, using two successive fires for a dash.
Conclusion.
In discussing the various methods of signaling here presented, nothing has been said of the relation of signaling to tactics. We have discussed methods only, and have endeavored to simply show wherein and in what measure naval signaling differs from army signaling.
Tactics demand of methods of communication three things: reliability, simplicity and rapidity.
Reliability involves mechanical perfection as to not failing at critical times, and certainty as to the message being received as sent. This is usually insured only by repeating back, which is, however, at the expense of rapidity.
Simplicity implies the use of a code of few elements covering a wide range of usefulness. The fewer the elements, the slower the signaling, but the greater the range of visibility, or the more reliable the signal as to distinctness.
Rapidity is, therefore, limited by the consideration of simplicity, and increased by mechanical perfection and reliability, which last restricts the too rapid methods of signaling. Tactics demand rapidity, but the crowning virtue is reliability.
Therefore, that method is best which is most reliable and which is as simple and rapid as is consistent with absolute reliability.
The question of interior communication on board ship affects tactics in the extent to which the whole ship is subordinated as a unit to the guiding mind which is working out the problem before it. The accurate signaling of the range, the constant touch necessary between the bridge and the engine room as to speed and keeping position, the net-work of communications to the torpedo tubes, batteries, ammunition supplies, water-tight subdivisions—all this is a problem in itself, but it is entirely a matter of methods and not of codes.
We must bear in mind that in our own service we are in a transition stage, so that any apparent criticism of methods is really only the application of experience to the modernizing of our methods. From a tactical standpoint the signal books need a general revision, and in view of the possible use and wide range of alphabetical characters (as in the international code), the index for entering the books and each signal itself should be given both as a numeral and as a group of letters. As to the use of action or battle signals, there are many who hold that in a general action no signals can be read or will be heeded. Single flags are, however, advocated for battle signals, and these should have an alphabetical significance as in the international code. This would enable a vessel to transmit the signal by her fog whistle when obscured by the smoke of battle. It may be readily imagined that some sort of communication between vessels for tactical purposes would be needed in battle under some circumstances, and if flags and shapes are obscured, the fog-whistle offers another means. If shapes are to take the place of flags, then the battle signals should consist of simple characters capable of display in one hoist.
We have in our navy three service codes: (1) a numeral flag code often elements; (2) the American Morse alphabet and numeral code, and (3) the original Very three-element or bracket code. We have, besides, three experimental codes in use in the North Atlantic squadron: (1) a new flag numeral code, using yellow in several flags in place of white; (2) the so-called Ardois alphabet and numeral code, and (3) the four-element Very code (originally proposed by Lieut. R. T. Mulligan, U. S. navy). There is as yet no official order prescribing the method of using the American Morse code on the fog-whistle or by gun-fire. Our general signal book contains almost no recent additions of new signals demanded by modern conditions, and is filled with many that are out of date. Surely there are enough questions awaiting some decision or official action to warrant the appointment of a board to consider what is to be done. The new flags possess the merit of great visibility and distinctness, but it should be definitely settled as to three-flag hoists being the maximum, and as to battle or action signals being one-flag hoists. Shapes should be experimented with; the Very four-element code should replace the old bracket one; homing pigeons should be regularly adopted as a means of communication; and, above all, the American Morse code should be withdrawn from active service. We have no method of distant day signaling, and distant phonetic signaling is only vaguely alluded to in the Revised Instructions in Signaling, which are issued experimentally.
The old Myer code is the most nearly perfect one in existence. It has been shown, however, that certain practical considerations as to four lights in a permanent hoist have made it desirable to limit the group of elements in the numeral code to four. Happily, the four-element Very night code offers just the numeral code needed to take the place of the one of five elements used in the old Myer. As applied with the Myer alphabet, it gives us thirty characters in what we may call the Myer-Very code or the Modified Myer code. These thirty characters are capable of transmitting any possible message under every possible condition of service. It is a perfectly well recognized principle of modern signaling that one group of elements (that is, a character) must be allowed more than one meaning or signification. For instance, in the English semaphore code of three elements, the ten numerals duplicate ten letters of the alphabet. In our general flag code, by the use of such flags as the geographical or telegraph, we can give to a hoist of flags several different significations; also, in our wig-wag code, as specified in the instructions in the General Signal Book, we may, by the use of such code calls as G. L. U. or T. D. U., etc., give a special meaning to any signal used.
We may, however, in place of the Myer code use the Continental Morse in the same way, either by calling the dot a i and the dash a 2, or by changing the Very numerals to dot and dash, as shown on page 448. This would give us a modified Continental Morse code of thirty characters.
If there is any value in theoretical considerations ; if the experience of foreign navies is in any way a guide to us ; if what has been here outlined is correct in principle, and of the least practical value, then we should adopt the Modified Myer code for all purposes of signaling. We can use its thirty characters as a wig-wag code for hand flag, torch, winker light, or search light; as a squadron distant code with shapes on a signal mast; as a night code with four double lanterns in a permanent hoist; with the Very stars for a distant night code, and with the fog-whistle and gun-fire in a fog. Abolishing the musical signals in the infantry and artillery instructions, the letters of this code transmitted by bugle could be used, and much confusion and misunderstanding avoided in brigade operations on shore. With only thirty characters to learn for all purposes, signaling in general would be much simplified. The same remarks apply to the present night speed-signals authorized by the Fleet Drill Book. With a double lantern (half white and half red, as in the Ardois) for squadron cruising purposes, on the truck of all our vessels-of-war, there would be no need for using the elaborate yard-arm, truck, and stern lights now carried. No matter how well drilled the look-outs may be, or what rare presence of mind an officer-of-the-deck may develop in any emergency, it is too much to expect that the speed-signals now in use can be properly made in the excitement of a moment. Assume that in steaming in open column (which is the safest formation for. night cruising) the leading vessel suddenly develops a danger close aboard. The helm must be put over ; both engines reversed or stopped ; the steam whistle sounded as to helm signal; if the danger is very imminent, the water-tight door signal must be made ; the two white lanterns must be hauled down from the yard arm, and the red light displayed astern. If the halliards jamb, or if the signal is imperfectly made and the vessel next astern rams the leading vessel, then the officer-of-the-deck of the last named vessel is responsible for the signal not being properly made, and yet he is not in a position to control the signals in an emergency and keep his attention on the danger ahead. The lights as now authorized are a real source of danger to the vessel carrying them, and they confuse merchant vessels, as every one will testify from experience. In time of war, no squadron is going to advertise itself with any lights. It would, therefore, be best in time of peace to practice cruising as would be done in time of war. No permanent display of lights is really needed. Vessels can keep in touch without it, and should be required to do so. A board of officers in the Squadron of Evolution, of which Captain Philip, U. S. Navy, was senior member, submitted a report to the Commander-in-Chief in November, 1891, on the subject of speed signals, as follows:
"1. With regard to day speed-signals in squadron, it would be in the direction of simplicity to abolish the pennant and use only a ball of an improved type, which ball should have four positions. Up at yard-arm, 'going ahead at full speed last indicated by signal;' half-way down, 'backing or going astern;' just above rail, 'steerage way or slow.' Out of sight, 'stopped.' The ball should be larger and heavier than that now used. It could be pierced through one axis by a steel tube working on a wire back stay, the tube working as a traveler. It would come down of its own weight. The lower end of the back-stay could be shifted and set up when the yard was braced (if the yard is movable). The question is the usefulness of the pennant. It is supposed to show relative speed, but it is a deceptive and inadequate device for accomplishing a purpose which can better be accomplished by signaling the exact speed. In the manoeuvres off Bar Harbor this summer (1S91), if the real speed had been signaled, then there would have been no doubt as to the 'fast speed' intended. The pennant required one or two men to run it. On a bridge of a vessel like the Concord it adds extra men. It can easily be dispensed with.
"2. With regard to night speed-signals, no lights other than the international lights should be permanently displayed. Each vessel should have in position, astern, a red globed lantern containing a lamp, controlled by the officer-of-the-deck from a switch on the bridge. This light should be displayed only in the emergency of stopping and backing. It should signify to the vessels astern 'look out for yourselves!'
"The yard-arm speed-lights (white) are a source of danger, on account of the liability of the halliards jamming, and they do not adequately indicate relative speed. They throw a glare on the bridge, they confuse merchant ships, and require from two to four men to manipulate them quickly. In case of emergency they are one more complication for the officer-of-the-deck . . . The masthead (truck) light is not useful, as at night the sextant is rarely used to judge distances ... In time of war, ships would have to keep their position, and keep together without lights . . . (We recommend that) there be placed on the fore and main truck of each ship in commission a double lantern, similar to that used in the Ardois apparatus, and connect the lamps by wires with a key board. The white light should be used as a winker to transmit the Myer-Very code, using the red light for the 'front' to mark intervals between words and sentences and end of message. These lanterns are useful in case the Ardois or other apparatus is masked or out of order; they can never be masked; and they would be in position to use as a white or red truck-light for squadron purposes such as for guard ship, etc.
''We recommend (in place of the white yard-arm lights), in case it is desired to signal, 'we are going ahead a little faster,' flash dots from the white light at the mast-head (truck). In case a vessel slows down for any reason, flash long dashes with the white lantern which will indicate 'look out, we are slowing a little.' (The red lantern should be used similarly for stopping or backing.) The points aimed at are that (1) no permanent lights should be displayed other than the regular ones, and (2) all other speed and emergency signals should be under the immediate control of the officer-of-the-deck. This point is vital." (Oil lanterns should be kept at hand, lighted and ready for signaling in case of accident to electric plant or truck lantern.)
Let us take the Armored Cruiser New York and imagine her fitted with the latest appliances for signaling, and using the methods of signaling h-ere advocated. She would have two military masts with signal topmasts, each fitted with a set of four collapsible shapes, one set for squadron cruising, about 2 ¾ ft. in diameter of discs, and 4 ½ ft. hoist, and the other, 4 ft. in diameter and 7 ft. hoist, for distant signaling. There would be halliards for international flags (since all ships are required to have them), and two signal yards, one on each mast, for hoisting flag signals (international) for such routine signals as might not seem to require the use and wear and tear on the shapes. There would be two sets of permanent hoists of four lanterns, one on each mast, on a stay from the masthead, so as to show from the port or starboard bow, and from the starboard or port quarter. Each hoist should have its own key board and circuit. On the truck of each mast would be a double lantern (red and white) for a winker light and speed-signals. Each truck lantern would contain two groups of six 12 c. p. lamps, worked from a key on the flying bridge for the officer-of-the-deck and on the signal bridge for ordinary signaling. The steam whistle would be specially fitted with levers, cocks and drains for fog-whistling purposes, and there would be on the bridge an audiphone for locating the bearing of sound signals. Arrangements would be made for homing pigeons for use in squadron manoeuvres to be utilized in time of war, and for Very' s signals short barrelled shot-guns would be provided for projecting the stars. The signal men would be rated men, petty officers, ranking next after coxswains, and their assistants would be men or boys training to become quartermasters or signal men, receiving extra compensation for such service. They should all be required to know the American Morse code, so as to be able to communicate with army stations when necessary; but it would not be the code used on board the New York or in the naval service. Thus equipped, this ship would be unquestionably better off in respect to signal code, apparatus and methods than any vessel afloat.
It would be too confusing to go into a detailed description of the method of using the Modified Myer code for all purposes of signaling. It will, however, concisely illustrate the simplicity of the scheme to give here a tabulated outline of it. Every detail is capable of ready explanation, and, while the synopsis in no way explains itself fully, the scheme is all there. It has been submitted officially and awaits the action of the Navy Department.
Code Calls.
1. A. S. U. Action or Battle signals use. 7. C. A. U. Cipher "A" use.
2. I. C. U. International code use. 8. G. B. U. Cipher "B" use, etc.
3. T. D. U. Telegraphic dictionary use. 9. O.N.U. Ordinary numerals use.
4. G. L. U. Geographical list use. 10. C. S. U. Compass signals use.
5. S. B. U. General Signal Book use. 11. V. N. U. Vessel's numbers use.
6. F. D. U. Fleet Drill Book use. 12. N. L. U. Navy list use.
DISCUSSION.
Lieutenant R. T. Mulligan, U. S. Navy.—It is not my intention to enter into any criticism of Lieutenant Niblack's very comprehensive article on Naval Signaling, or to discuss the merits or demerits of the various systems and codes used by other nations, which he has so fully described. I shall confine myself to the systems and codes now authorized by the Navy Department for use in the United States Navy, and, if possible, offer certain suggestions as to changes or modifications in them.
I fully endorse Lieutenant Niblack in all he has said in condemnation of the American Morse Code for naval signaling, and concur with him in believing that the naval service almost demands an immediate return to the old Myer Code, modified in the direction of simplicity.
The disadvantages of the present "Army and Navy Code for Visual and Telegraphic Signaling" can only be fully appreciated by those who have had practical experience in its application to fog-signals. The only instructions, authorized by the Department, for a system of fog-signals will be found upon pages 13 to 15, inclusive, of the General Signal Book, and experience has long since proved that these instructions, even when applied to the old and almost perfect Myer Code, have failed and have fallen into disuse, being impracticable and cumbersome. Following the instructions on page 15, General Signal Book, it would take an expert signalman 6 minutes and 15 seconds to signal, "Course S. W."
In consequence of this, signal officers have, from time to time, been called upon to improvise a system of fog-signals which depended upon the ear and not the watch for its successful operation. A time element is dangerous and should be avoided in any system. As it is impossible to use the American Morse Code without the introduction of a time element, this is a sufficiently good reason for its being abandoned. When you attempt to apply it to the fog-whistle confusion becomes worse confounded. The fourth element of this code has two meanings : it may either represent a space or a front.
The advantage of the Myer Code is its extreme simplicity; it is a code of but three elements, and is constructed upon sound and simple principles. The time element is eliminated. It can be applied to the wig-wag (flag or torch), heliograph or flash lantern, fog-horn, steam whistle, bell, gun-fires, Very's Night Signals, or the "Ardois Alphabet," and to all with equal facility and certainty. The present "Ardois Alphabet" is cumbersome, and it is an additional code to be learned, there being no connection between the displays and the dot and dash of the American Morse Code.
The system of numerals proposed for the Myer Code is excellent; it fits in with a modified Very's Code.
My idea of flag signals is that no tactical or general signal should consist of more than three flags. By making use of the zero and three repeaters in the numeral code we have 13 one-flag signals, 156 two-flag signals, and 1716 three-flag signals, making in all a total of 1885 displays or permutations. I am of the belief that the present tactical and general signal books can be reduced to this number of signals. It requires but a glance at any page in the General Signal Book to see what a large percentage of these signals can be stricken out without reducing its efficiency. These signals should be arranged and grouped so that those of urgency and importance would contain the least number of flags or shapes, and should be bound in one small book. The Telegraphic Dictionary and the Geographical List can be expanded to any extent; they should be as full and complete as possible, and should be bound in a separate volume. These signals need never go beyond a four-flag hoist.
I do not believe in a flag alphabet. By increasing the number of flags or shapes (elements) you increase the chance of making an error in each signal. Our present code (proper) consists of thirteen (13) elements, and I am firmly of the belief that with it we can make all the signals that will be necessary for the handling of a fleet and its dependencies in time of war.
If we give up this code and adopt one having 26 elements (a flag for each letter of the alphabet), we will have, in accordance with the theory of chances, in every one-flag hoist twice as great a chance of error, in every two-flag hoist four times as great a chance of error, in every three-flag hoist nine times as great a chance of error, and in every four-flag hoist twenty-one times as great a chance of error.
The navy needs a simple code that can be easily learned, and one in which the chance of personal error is reduced to a minimum. The proposed flag alphabet does not fulfill these requirements. Let us simplify the present system and not complicate it.
Very's system for distant signaling is excellent. The code adopted by the Navy Department is a bad one. It can at once be improved by representing the numerals by groups of four (4) stars, elimination of all brackets, and the introduction of an interval or space to separate words in a telegraphic message, etc.
If the stars were loaded in metallic shells and fired from a short double-barrelled breach-loading shot-gun, a great deal of time could be saved. The one great objection to Very's system is that it is slow.
In conclusion, I will simply state that I am of the belief that a modified Myer Code, which can be applied to the wig-wag, a modified Ardois Alphabet, a modified Very's Code, and fog-signals, would give satisfaction to the entire service.
Lieutenant H. P. Huse, U. S. Navy.—I have been very much interested in Lieutenant Niblack's excellent paper on signaling, and what remarks I have to make on it are supplementary rather than critical, and are confined to visual signaling.
Two systems of visual signals are necessary for the service, a day system and a night system; and two codes are necessary, a long-distance code and a short- distance code. This gives us four classes of visual signals; but the matter can be simplified if we use the same code for day and night short-distance work, and a second code for day and night long-distance work. If we consider apart all signals which can be projected to a great altitude, like rockets and Very's signals, it is not difficult to reduce the question to this form, and then, I think, that most officers who have had much practical experience in signaling on board ships at sea will agree with me that the short-distance code should be based on what Lieutenant Niblack calls transient signals, while the long-distance code would better come under the head of permanent signals. If this is admitted the problem is simplified.
Short-distance Signaling by Day.—The objections raised by Lieutenant Niblack against the wig-wag are only too familiar to us. Whoever has had to send a message to several ships, no two of which could read the same signal, has felt the want of an all-around system. In fact, the short-comings of the wig-wag are so great that the flag and the torch will, it is hoped, soon be relegated to the class of only possible accessories in ship-signaling.
In 1889 Lieutenant Benson suggested to me the use of collapsing shapes, and experiments were carried on in the N. A. Squadron, first on board the Baltimore, and afterwards on board the Philadelphia, with a view to discovering the possibilities of such a system. A cone and a cylinder, each about two feet high and made with barrel hoops and bunting, were the primitive means at my command. The shapes were exposed and collapsed by means of endless signal lines and small blocks, so rigged that when the shapes were triced up aloft, a downward motion on the lines exposed the shape, and an upward motion collapsed it. The signal was sent by one man, who worked one shape with each hand. The rapidity of work was about equal to what could be done with the small wig-wag flag. The signalman had complete control of the situation, and could make a signal visible all around the horizon. Signals made to the Petrel, about half a mile off, were read without difficulty by apprentices, who had had no practice except with flags. It is to be observed that with two collapsing shapes five elements can be conveniently used. The result of the work was to satisfy me that with carefully constructed apparatus the system would prove successful. The difficulties I encountered were that the shapes were too small, and that the bunting of which they were made jammed in the blocks. Other duties and my detachment from the Philadelphia interfered with further work in this direction at the time. The code used was that embodied in Lieutenant Niblack's paper on page 471.
Short-distance Signaling by Night.—I think Lieutenant Niblack is right in his criticisms of the method used for night signaling in the N. A. Squadron in 1890 and 1891. It was found to be too slow, though very reliable. The green light is not an essential part of the system, however, and I still think that, in the event of the temporary or permanent failure of the more complex arrangements of lights required in such systems as the Ardois, the ship whose signaling outfit was thus disabled might find this simple system a great convenience. A fully equipped key-board costs only 130; the lights are the ordinary electric deck lanterns furnished. The system was first suggested to me by Lieutenant Benson.
Long-distance Signaling by Day.—I agree entirely with Lieutenant Niblack in nearly all that he says about the use of flags for signaling, but I think his suggestion, that we follow the example of Italy and adopt the international code flags, is open to criticism. In a complete system of signals, submitted to the department by Admiral Gherardi and briefly referred to by Lieutenant Niblack, a whole system of flags was designed on the following principles:
1. The colors best adapted for the purpose are blue, red and yellow.
2. The design of the flag and not the colors should distinguish the flag, i. e., the colors should serve only to mark the design.
3. Consonants should replace numbers. It was found necessary in the development of the system to introduce one additional flag (A) as answering and affirmative pennant.
4. There should be no top and bottom to a signal flag. That is, the flag should be symmetrical in construction and design with respect to a line perpendicular to and at the middle point of the hoist.
The object of not having the color a distinguishing feature of the flag was not that signals might be read by the color-blind, but because there are conditions of light and distance when it becomes impossible to distinguish between even red and blue, and both appear simply dark, although the line of demarcation between two colors can clearly be seen. Thus at sea at midday I have seen a signal in which it was absolutely impossible with the best glass in the ship to tell whether a certain flag in the hoist was five or six. Finally the signal quartermaster was positive that it was five; it proved afterwards to have been six. The ship was the Galena, and I was the signal officer.
Anyone who has tried to do fast signaling knows the annoyance of getting hold of the downhaul of a big signal flag when he wants the hoist to hook on. It is essential to quick work that there should be no difference between the two; the top should be the end the signalman gets hold of first. Both ends should, of course, be clearly tagged.
The advantages of a system of flags designed on these principles should outweigh any small advantage to be derived from adopting the international code flags. It would, of course, be convenient to have only signal flags belonging to one code for all purposes, but the difficulty of making out some of the I. C. flags at a long distance, and the danger of confusing such letters as P and W, when the flag is not blown out clear, should of themselves be sufficient to condemn them.
The size of signal flags now issued to ships for general signaling is too large for ordinary squadron work and too small for distances above three to five miles. Two sizes might better be used, the smaller size for squadron evolutions, etc., when speed is sometimes of great importance; the larger for long-distance work, as communicating with outlying vessels on scouting duty. In this connection I may say that the very important question of sufficient altitude of masts and signal yards to admit of sufficient hoist for signals seems to have been overlooked in the design of some of our new ships. In the Baltimore this had to be remedied as soon as she became a flagship. On none of the new ships that I have seen is there any evidence that in sparring and rigging them the question of signaling facilities has been considered as a matter of primary importance, not a peace importance either, bat a war importance.
Lieutenant Niblack has so completely covered the subject of long-distance night signaling that there would appear to be little left unsaid. I can only add that the Very system, using the four-star code, always worked satisfactorily in the N. A. Squadron in 1888-1889, provided the pistols were not worn out, the cardboard cartridge cases not too much swollen to enter the chamber, or the stars not so defective that occasionally one failed to ignite at all, or did so a few feet above the water, leaving the sender in doubt whether he should fire it over or not. But with a simpler pistol on something like the old Remington model, with metallic cartridge cases, and with careful laboratory work, the Very system, with the four-star code, leaves little to be desired in the branch of long-distance signaling for which it is best adapted.
The subject of a code and the much broader subject of the general arrangement of the signal book are very tempting, and I should like to discuss them in this paper, but as an appendage to another paper it is already long enough. I must, however, call attention to the fact that much greater speed than is given by the Myer code can be obtained by using five elements instead of three, and that five elements are at our command with two lights of different colors or with collapsing shapes. A return to the Myer code would, in my opinion, be a wrong step, as it is distinctively slower than the American Morse now in use. The American Morse has never, to my knowledge, given trouble which has not been due to the inherent defects of the wig-wag system.
Lieutenant W. F. Fullam, U. S. Navy.—Lieutenant Niblack treats the signal question so exhaustively and so practically, that little remains to be said except to comment upon the facts that are cited and the conclusions arrived at in his paper. It is conclusively proven, in this searching examination of the subject, that, in passing from the Myer to the Continental and then to the American Morse code of wig-wag signals, the navy has progressed steadily backwards during the past six years. That the army had good reasons for making these changes may be admitted, but that the navy, in following the army, grasped the shadow and lost the substance of this important matter, is clearly demonstrated.
There is, perhaps, a certain advantage in having the same code of signals in the navy as in the army, but this principle is greatly outweighed by several others, each of which is of far greater importance. The result is, therefore, that we have followed a theory based upon a single, and not very important principle, and in so doing we have neglected many practical considerations, and violated the true and vital principles that should govern the selection of a code for the use of the navy. Lieutenant Niblack has shown, beyond question, that the conditions and requirements afloat and ashore are so different that the navy should develop or select a system of signals that are suited to its own peculiar uses. In other words, the navy cannot rely upon the army to think for it in this matter, nor can it successfully bully Nature's laws and ignore service conditions by pursuing a course that rests upon one lone idea—that of having the same code as the army.
The fatal defects of the American Morse code, and the fact that it cannot well be adapted to naval conditions, are made glaringly apparent in this paper. That the space or time element is a serious objection will be admitted by those who are called upon to read signals by this code. It is far easier to read signals by the Myer code, and surely we should have the code that renders the reading of signals as easy as possible. The difficulty of using the American Morse code with the steam whistle, the fact that it cannot be adapted to a system of night signals like the Ardois, and that a hoist of six lanterns would be necessary "to transmit the numeral six," are reasons enough to show that this code is totally unfit for the navy.
It is almost amusing to note the predicament in which the navy finds itself as a result of the adoption of the American Morse code. Lieut. Niblack states that "the American Morse code cannot be used in the Ardois, and in our service we have a special code called the 'Ardois ' to use with the apparatus." In order, therefore, to communicate effectively with the army at night, the latter must learn the Ardois code! In other words, by adopting the American Morse, we have not even facilitated nor simplified the means of communication between the Army and the navy. Both the army and the navy must now learn two codes! The attempt to make one code answer is a failure—this one lone, shadowy principle, with which we have flirted, has jilted us.
On the other band, had the navy adopted a simple and practical system —one that could be adapted to all purposes of naval signaling, by day or night—the army would have been put to no more inconvenience than at present, and the navy would have gained immensely in having but one code to learn in making its own signals. In fact, the army might also have gained, because the system that would be adapted to naval purposes would probably be simpler and more easily learned than the Ardois code.
It is hardly proper for one who has had no practical experience with the Ardois system to attempt to discuss its merits or defects. But it would appear, from Lieut. Niblack's paper, that the Ardois alphabet is by no means a good one, and that there are many serious objections to a system that requires five lights in a hoist when it would be possible to get along with four. When we consider the restrictions as to height of masts that are imposed on board small ships, monitors and torpedo-boats, it is clear that the fewer the lights the better.
Lieut. Niblack's treatment of the subject of speed and other signals used at night in squadron cruising, is very much to the point. If squadrons must manoeuvre with few or no lights in time of war, officers should be prepared for such an emergency in time of peace. An officer who has always been accustomed to cruising in a squadron loaded with lights, like a torch-light procession, will experience a violent change passing into the conditions that obtain in time of war. An officer-of-the-deck who must direct the helm, control the engine, watch the other ships, and at the same time see that three or four men hoist or lower as many lights or signals, may avoid disaster in an emergency, but it will be more by good luck than good management.
"The simpler the whole question of signaling can be made, and the fewer the codes which men have to learn, the better we will get along, as we have no corps of trained signalmen, and we require practically all our officers and men of the active combatant force to be up on signaling." This sentence sums up the whole matter, and Lieut. Niblack's plan of a modified Myer code, that can be "adapted to all purposes of signaling" is the true and only practical solution that has yet been proposed.
With "three service codes" and " three experimental codes in use in the North Atlantic squadron," with no official method of using the American Morse code on the fog-whistle, with no method of distant day signaling, and with a general signal book that is twenty years behind the times, it would indeed be difficult to imagine a condition of greater confusion and inefficiency than now exists in the United States Navy in the matter of signals. Under such circumstances, with important questions unsolved, with a multiplicity of codes, and poor codes at that, it is not surprising that many officers and men find themselves in in a state of complete bewilderment as regards signals. It is probable that there are now fewer officers and men who are competent to make and read signals than ever before in the history of the navy. Certain it is that the subject has never presented to them so many difficulties as at present. It may well be asked whether the navy can be considered in a state of readiness for war, with the signal question in its present condition? If this question can be answered in the affirmative, it follows that signals are of very little importance in the navy and have very little to do with the proper conduct of naval movements or operations in time of war.
Lieutenant J. M. Bowver, U. S. Navy.—I have read with much interest the able and carefully prepared article on Naval Signaling, written by Lieut. A. P. Niblack, U. S. N., and I congratulate him upon his masterly treatment of a subject that has not heretofore received the attention that it deserves.
The changes made in recent years have not all been beneficial. I refer particularly to the changes made in the wig-wag code.
Among the officers of the navy we have lost, by a process of elimination, due to changes of code, a large number of qualified signal men. When I made my first cruise, the officers of and above the grade of master, as a rule, could not read or make wig-wag signals, the reason being that the code had been changed since they learned it. My impression is that the change referred to was made in 1870, and, although only a slight one, it was sufficient to practically disqualify a large number of officers.
It is important that all line officers should be able to use the wig-wag; but since 1870 about fifteen classes, representing more than 500 officers, have been graduated from the Naval Academy qualified in the Myer code only to be disqualified by the adoption of the Continental Morse, and those that learned the latter were, in turn, disqualified by the adoption of the American Morse.
The numerous changes of code would not matter so much were it not that it requires much practice to become proficient in any code, particularly when one has to forget a code or codes already learned.
I believe that the Modified Myer code, as recommended by Lieut. Niblack, would prove excellent for all purposes of signaling. The American Morse, so long as the army clings to it, would have to be taught to the signalmen.
Regarding shapes, I consider them excellent on board large ships when not engaging an enemy; but in battle, presenting as they do a large target, they would become worse than useless, because the enemy's fire would inevitably derange them so that either they could not be expanded, or, being expanded, they could not be collapsed.
I therefore favor a preventer set of flags, made to take the place of the shapes, for use when the latter become impracticable, i. e., when they become jammed in action, and in small craft where they cannot be conveniently used.
Captain J. W. Philip, U. S. Navy.—Lieut. Niblack's article . . . has interested me very much, and I agree with him in his conclusions. The Myer code, with the proposed modifications, seems to me the best for adoption for all of the methods of communication. The adoption of shapes instead of flags, the masthead light, and the elimination of the fifth light in the Ardois system, are also in the line of improvement. I have asked the Department that there be placed on the fore and main trucks of the cruiser New York a double lantern, similar to that used in the Ardois apparatus, the lamps to be connected by wires to a keyboard. These lanterns are useful in case the Ardois or other apparatus is masked or out of order; they can never be masked, and would be in position for use as a white or red truck-light for squadron purposes, such as guard ship, etc.
Lieutenant T. B. M. Mason, U. S. Navy.—The lecturer certainly deserves our thanks for the intelligent way in which he has handled his important subject, and for the fearless way in which he has pointed out the defects in our present methods of communication.
He is undoubtedly correct in advocating the adoption of a four-element Myer code. The old code is the best that we have ever' had, and the proposed elimination of the fifth element is a vast improvement.
The Myer code is equally adapted to the best methods of sending day and night signals now known, either visual or phonetic. That it does not correspond with the codes used in ordinary telegraphy seems an advantage, as it could be used by signalmen without being intelligible to civilian operators. A written message in the code characters could be sent or taken down in writing, on receipt, by any ordinary operator.
It would be a great improvement to use the same code with all our instruments for communicating, thus doing away with the necessity of teaching our signalmen half a dozen different codes. When we had the old Myer code we had plenty of good signalmen ; now we have but few, and even those are not proficient in all of the codes and often become confused.
That a distant all-around signal should be made from the highest point of the station is undoubtedly true, and an arrangement of masthead lights and shape signals absolutely necessary.
I quite agree with the lecturer in his conclusions as to what we need, and sincerely hope to see the whole subject placed in the hands of some officer of modern experience who can evolve system out of chaos and pull down our present tower of Babel.
The findings of the Philip Board will certainly be approved by every officer who has had charge of a modern .ship sailing in squadron.
The subject of carrier pigeons is one of vast importance to us, and the service owes its thanks to those who, almost entirely unaided by the government, have succeeded in making a beginning in establishing such means of long distance communication. A single bird, reaching land from an outlying picket vessel might give timely warning of an approaching enemy, allow of a concentration and save a city.
The New York is to be fitted for masthead signaling and could easily be fitted to display the shape signals. The above sketch shows her present signal equipment.
Lieutenant W. Irving Chambers, U. S. Navy.—I have given this subject considerable thought and agree with the writer on all the essential points. I have also conversed with many officers who have had recent experience with the confusing system of signals in present use and, without exception, the opinions expressed have favored a return to the Myer code. Give us that code, and the details of using it under all circumstances will readily follow.
I venture to disagree, however, on a minor point, where (p. 448), in a spirit of fairness, the writer wishes to point out "one rather serious defect in the Myer." Having had an exceptional amount of experience in signaling with the Myer code by fog-whistle (on board the steamship Loch Garry, with the Greely Relief Fleet, where I had personally to attend to all the signaling, very much of which was in fog or thickly falling snow), I can stoutly affirm that no such defect exists, if we regard 1 as a blast (of any duration) and 2 or 3 as two and three toots respectively. In this case Y is simply 3 blasts, and ''front '' (or interval) 3 toots ; N is 2 blasts, and T is 2 toots. It is very easy to make this distinction between a group of toots and a blast, and my experience teaches me that if the sounds, toots or blasts, are of equal duration, the result, especially with a "smart signalman" who is trying to show off, is a confusing succession of sounds improperly spaced.
In regard to the proposed arrangement of shapes for distant day visual signaling, I think practice would demonstrate that the varying effects of atmosphere and light would frequently so distort them as to render it difficult to distinguish the difference between the two, and that the mechanical difficulties of the arrangement of two shapes in one would prove too great for smooth working. I would, therefore, suggest that the shapes be all alike and double. Figures (a) or {b).
For all collapsible shapes I would prefer some such preparation as oiled silk to rubber, on the score of durability, and should wish to arrange the covering so as to prevent access of water to the springs and joints. The difficulty of mast construction would enter as a factor. We would need light masts with about sixteen grooves, and this difficulty could perhaps be met by making the small masts as shown in figure (c). The flanges should be flush riveted.
This, to my mind, is the best method of construction for all military masts; but that is somewhat outside of this subject.
I fancy, however, that the simplest and best (though perhaps the slowest working) arrangement of signal shapes would be that whereby the shapes were wholly hoisted at each display, and such device might be tried on any mast of sufficient height. Figures (a), (b) and (d).
The double shapes would require a little greater length of mast than the single shapes, and the hoisting shapes a little greater length than fixed shapes, but a mast 50 feet in clear would amply suffice for double shapes of 4 feet diameter.
I do not think that torpedo-boats or torpedo hunters should be provided with Ardois lanterns, but should be required to rely upon the flashing masthead lantern which all ships should have. The most important war or lookout warning signals should be arranged to be transmitted by one lantern in the shortest time.
I think the thanks of the service are due to Lieut. Niblack for the zeal and patience he has displayed in following up this subject, which is of rapidly growing importance.
Commander C. M. Chester, U. S. Navy.—It is only necessary for one to have attempted to study the subject of signals in the navy to comprehend the vast amount of labor that has produced the excellent paper of Lieut. Niblack.
Assuming that we are to retain our woefully complex system, which has resulted in the common practice of making a signal to "send a boat," and then returning the message by a bearer of dispatches, the propositions of the lecturer become valuable. Indeed the paper is full of valuable suggestions for any system of signals, but what I would impress upon the Association is that our present code of signals is entirely inadequate to the growing demands of the service.
The U. S. Navy Code of Signals consists of 10 elements, represented by 10 digits, which, combined in groups of 4, allow 9999 signals to be indicated. As our signal book contains many more than this, we resort to repetition of numbers, qualified by a pennant or signal to go before. Should the pennant be obscured, as it frequently is, a signal sent to "Ram the enemy at full speed" may be interpreted "How," or the reverse. In order to make all the possible combinations of the signal code, we use at the present time 31 flags. This cumbersome system has resulted, as I have said, in sending messages by boats, rather than as signals, and thus preventing the practice in signaling which is an absolute necessity for preparation for battle.
In 1889 Commander C. H. Davis proposed an entire new departure in signals, based upon the system of the International Code. His communication was referred by the Department to the Board on Organization, and, after a careful investigation, its adoption was recommended.
The system contemplated the use of 20 elements, represented by 20 consonants of the alphabet. It is only necessary to work out the permutations possible, to show that over 116,000 signals can be made with four flags, or other symbols; ample to accommodate any code, without repetitions.
The great advantage claimed for the system is that no combination of flags can have a double meaning.
Having been detailed by the Bureau to prepare a new code of signals, on the basis of the recommendation of the Board of Organization, I worked continuously on the subject for over a year, and submitted to the Department what was to be known as the "Introduction to the Code" and the "Fleet Evolutions," and left the "General Signal Book," partially completed, in the hands of Lieut. John F. Parker, in April, 1891.
As the most difficult part of signaling takes place during the night, and as at this time communication is generally more necessary, the code was prepared to cover these requirements, knowing that day signals would easily be assimilated to the night code.
In the proposed code the object has been to combine the fewest number of elements that will contain all the symbols used, without repetition, for in repetition there is danger.
It is apparent that two elements, which make only nine combinations in permutation, cannot be used; but that three elements, permitting twenty seven combinations, will suffice for all the requirements of the service.
It is made mandatory that all signaling in the navy should be done by the Navy Code, thus insuring the facility in its use necessary in war by constant practice with it in time of peace.
The following sketch will illustrate the method of using the various symbols in combination, and it is readily seen that any three dissimilar objects may be utilized.
It is hardly necessary to go into a description of the code, but by permutation the three elements, called, say, red, white and green, represented in any way desirable, enable us to make the twenty seven letters of the alphabet. Taking twenty of these for the code proper, we have six letters left for the qualifying signals, "Interrogatory" and "Preparatory" (the only two used), and the "Answering," "Annulling," etc. The "&," or 27th letter, has a special signification, described in the "Introduction to the Code."
The "Signal Book" has three parts, besides the introduction.
The "Fleet Evolutions" contains:
1st. 20 "Battle Signals," indicated by a single flag.
2d. 380 "Evolutionary Signals," indicated by a hoist of two flags each.
The General Signals are all made with three-flag hoists, and the book has a possibility of 6840 signals; a larger number than in the present signal book.
The Telegraphic Signals are made with four-flag hoists, and there can be made 116,280 signals. Thus additions to the present Telegraphic Dictionary may be made and its scope greatly increased.
The General Signal Book contains a numeral table, similar to that in the International Code, so that any number may be expressed by the letters of the code.
There are, however, ten auxiliary symbols, used to indicate the nine digits, shown by pennants in the day signals, and by two lights at night. The "Zero" and "&" are identical.
Thus messages in numerals may be sent, without the numeral table, although, where confusion from long distance or other causes are likely, they would not be used.
The night signal method, described in the "Introduction to the Code," consists of three electric lanterns, each with a red, white and green shade, arranged vertically, with nine wires, leading to a keyboard, containing the twenty seven letters and ten numerals. The key indicating a letter closes the circuit for one color in each of the three lanterns, while that for a numeral shows two colors, in the same way. The search light is used by showing one, two or three flashes, corresponding to red, white or green of the code.
The sound signals are, dot, short dash and long dash.
The wig-wag, as right, left, vertical, corresponding respectively to the red, white and green of the code.
A very imperfect description from memory of the proposed code has been given, but I will have accomplished my object in taking part in this discussion if the necessity of discarding the old system, with the old ships of war, be shown and substituting something else for a different class of vessels. This, I think, is better than to rehabilitate the old Myer code, as suggested by Lieut. Niblack.
A prime feature of the proposed navy code is the use of shapes for distant signaling, as well as for use on board the low, mastless vessels, which are likely to be common in the future.
I worked for some time over an idea by which the three shapes, square, cone, and inverted cone, could be displayed mechanically on a hollow iron mast, from a turret, but other duties prevented my completing the project.
The semaphore symbols of the code can be easily worked this way, and even this is better than trying to hoist the 4, 5 or 6 flags necessarily used at present. It may be said that the three colors taken as the basis of the new system are objectionable, particularly in the Very code, and yet Mr. Very himself has recommended their use, and the Italian navy has adopted his recommendations.
The changes made from the Myer code to the English Morse, and later to the American Morse code, were properly made to accord with the army, that we might communicate with that branch of the service, when necessary. In doing so we have almost lost sight of the fact that we have a navy code, upon which we must rely in war time, and which, by disuse, will have become of little value. I would keep the American or English Morse or Myer code to talk to our army friends, as we would send a telegraph message (through an operator), but it should be held as almost a misdemeanor to communicate with our sister ships in anything but the language we shall use during battle.
The drills in signaling should be as rigid as the requirements of the old spar and sail exercises.
Lieutenant H. S. Knapp.—The paper given us by the writer is an excellent presentation of the inefficiency and insufficiency of the present means of signaling in use in the navy, together with a remedy therefor. Every officer who has been cruising recently can testify to the confusion resulting from the use of our many different codes, service and experimental. Every officer, from admiral to sadet, ought to be able to understand any signal that he can see, or hear; and if we had a systematic code, carefully and consistently worked out, it would not be too much to expect of any officer that he should be familiar with all sorts of signals. As it now stands, one needs the memory of a Loisette to remember all the codes with which we are burdened.
From personal experience, I can heartily endorse all that Lieut. Niblack says regarding the shortcomings of the American Morse code. Without criticising the reasons that led to its adoption, I unhesitatingly express the opinion that it has been shown painfully cumbersome and inadequate to naval needs. It is harder to learn and more difficult to retain, and it is distinctly slower than was the Continental Morse or the Myer. In doing away with it, however, I am disposed to favor a return to the Continental Morse rather than to the Myer, and for two reasons: first, an army signalman learning the navy code, or a navy signalman learning the army code, would have many less letters to learn over, owing to the duplication of sixteen letters in the two Morse codes; secondly, it would be possible with it to communicate with foreign ships using the same code. It is worth while getting as close as possible to the army in the matter of signals, though it is quite unnecessary to adopt a code for the simple reason that it is the one in use by the army; and the desirability of a common signal code in combined operations with the naval force of other nations, a not infrequent occurrence, was proved during the time that our service used the Continental Morse. Lieut. Niblack himself shows that the Continental Morse can be used in the same way as he proposes to use the Myer. Either would have to be modified to make it a four-element code, and either gives thirty characters when so modified. With the universally accepted convention that the dot is indicated by a motion to the signalman's right, and a dash by the motion to his left, the rest of the general scheme works itself out as well with the Continental Morse as with the Myer.
The scheme is a valuable one as showing that a consistent code can be evolved for all purposes. I have never made any special study of flag or shape codes, and do not feel competent to criticise the changes in these regards proposed in the paper. But I venture to express the hope that, when a code shall be finally adopted, as much attention be given to the instruction of officers and men in general and night signals as is now devoted to the wig-wag. My experience goes to show that most of the instruction given afloat in signals is in the use of the wig-wag code.
That the Ardois, or some similar, apparatus has come to stay there is little doubt. Lieut. Niblack claims for it the advantages of "rapidity and accuracy in signaling." The latter point is freely granted, and the value of that sort of apparatus for night work is unquestioned for that reason if for no other. Moreover, its range is very considerable. I doubt, however, the practicability of its use on low-masted vessels, and consider the use of either the modified Continental Morse or Myer code with the two-element truck-lamp, fitted with Fresnel lenses, the best substitute for it. As a "preventer" on any ship these truck-lamps should always be fitted. I believe that, with a proper keyboard and an arrangement whereby a reduced current could be made to flow constantly through both lamps, sufficient to keep them at a low red heat while signaling, nearly or quite as great speed could be attained as with the Ardois, though not with the same certainty of the signal's correct reception, of course.
The report of the Philip Board concerning night speed-signals and Lieut. Niblack's remarks thereon appeal to every watch officer. To say that the present regulation speed-lights are a nuisance is to state the case mildly. To be of any value, they should be electric lights (during the time I was attached to the Squadron of Evolution one ship used oil lamps habitually, and as habitually the officers-of-the-deck of the other ships placed no dependence whatever upon the display); also, to be under the immediate observation and control of the officer-of-the-deck, they should hoist from or near the bridge. But in that situation, the glare resulting is a serious source of danger. A squadron is not on dress parade at night; every purpose is gained when the vessels keep in good touch—near enough for mutual support and to be well in hand by the Commander-in-chief, and far enough apart not to endanger one another. These objects can be attained without speed-lights and the lights should go. We ought to discard them in peace and so gain the experience and confidence in night squadron sailing under war conditions that would be of utmost importance in case of actual hostilities.