The Art & Skill of Radio-Telegraphy
-Second Revised Edition-
William G. Pierpont N0HFF
It would be very interesting to know the thinking behind the development of the original Morse code. It had to be tied in intimately with the limitations of the electro-magnetic mechanisms being designed to transmit and receive it. Records show that beginning as early as the B.C.'s reflected sunlight (heliography) by day, and lamps, lights or fires at night, were used for some kind of elementary signalling. By the A.D. 1700's (and well into the 1800's) several semaphore systems had been devised and were in rather extensive use in Europe and elsewhere. These used an alphabetic code formed by the configurations of two or more signal arms or shutters making block patterns (at night some used light configurations) for distance signalling within line of sight.
All these systems (often aided by the use of telescopes) were subject to weather and visibility limitations, and generally required at least two operators at the receiving end -- one to look and the other to write. Where considerable distances (a hundred miles or more) were involved relay stations were established. These signalling systems conveyed a symbolic message or spelled out words for visual reception. A few electric or electrochemical systems were developed using some method of spelling out words by transmitted letter symbols. Morse's system was not the first to use electricity. During the early 1800's several electrical and electro-chemical systems (which overcame the visibility problem, which was complicated by weather conditions) were invented and used. Some of them were quite ingenious, but tended to be cumbersome, rather slow and troublesome to maintain.
Morse's ingenuity was in combining a simple electro-mechanical system with some sort of "linear" coding. Samuel F. B. Morse ingeniously foresaw the newly discovered principle of electromagnetism in combination with some sort of "linear" coding as the key to developing a truly practical telegraphic system. It could provide the relative simplicity and ruggedness needed for the equipment. Like Marconi half a century later, his vision to combine these newly discovered principles and the entrepreneurial drive to bring them into use made telegraphy what it became in the field of communication for many decades. Two features were needed: equipment and a suitable code. As originally conceived it was to be a self-recording system, inscribing the code signals on a strip of paper tape to be read by eye. There was no thought given to "reading" it by ear alone.
His coding system begun in 1832 was a translation system consisting of two essential parts:
Morse seems to have recognized that a receiving operator could readily count up to five dots, but that a higher number of dots would be more subject to errors, as well as taking longer to transmit. With such a system, the duration of the dots and spaces was not critical, but it was a tedious, slow and clumsy system (as well as being rather subject to errors which could only be found on deciphering). Not much ingenuity was required to develop the code symbols for the digits: he simply used from one to five dots to represented the numbers 1 -5, and extended this through 9 and zero by a longer short space following (here indicated by the symbol @).
Here is his code:- 1 2 3 4 5 6 7 8 9 0 . .. ... .... ..... . @ .. @ ... @ .... @ ..... @
With such a system, the duration of the dots was not critical, but the relative spacings were important. What a tedious, slow and clumsy system it would have been (as well as being rather highly subject to errors which could only be found on deciphering). The overall idea was ingenuous, and the actual code signals used for the digits were simplicity itself. But his coding system was the weak link in his whole system, and would hardly find wide acceptance. (Later, this code-translation-book method was applied in China where it made sense to convert the Chinese characters to numbers, using an already available standard Chinese dictionary in which each character had for other reasons been assigned a number.)
Chapter 2 of George P. Oslin's book "The Story of Telecommunications" opens with these words: "Ask almost any American who invented the telegraph, and the answer will be 'Morse', but he did not create the dot-and-dash Morse code, the Morse key, or the stylus recorder." Who was Mr. Oslin, and where did he get this information?
He was a journalist who later became public relations director of Western Union. To prepare this book he exhaustively investigated newspaper articles, magazines, books and more than 100,000 letters and diaries of those involved and condensed it. (He was 93 years old when the book was published.) Pages 13 to 28 are devoted to a summary of the origins of Morse telegraphy, from which the following quotations come. Previous publications had only hinted at what Mr. Oslin has said so clearly. (The numbers in parentheses refer to pages in his book.)
In order to understand the confusion we need to realize first that "Morse's craving for fame was so strong that he postured, pontificated, tried to convince everyone he was great, and was zealous in defending his claims." (28, note 27) To blow up his importance, Morse on several occasions made some quite false statements and exaggerations. It is too bad that he refused to give credit where credit was due, for he would have showed himself a greater man by it. From the very first, Morse made strict contractual relationships whereby he alone was to be credited with all advancements and improvements: all credits for whatever anyone did for him would [publicly] belong to him alone. Yet in a letter Vail, his expert assistant, wrote on March 11, 1853 that "his agreement with Morse provided 'that whatever Mr. Smith, Dr. Gale, or myself should invent or discover, going to simplify or improve Morse's telegraph would belong to all jointly" (24).
However, Morse never shared any of this, and constantly cut Vail off from any public recognition for his work. Because of this, we know almost no details about the development history of the alphabetic versions of the code. We can be sure that if this code had been the work of Morse himself, he would most certainly have carefully elaborated every step of its development. (This is one clue previously published materials provide us.) A second factor was that they were physically separated during most of the first six or seven years: Morse was in New York City while Alfred Vail was working independently in Morristown NJ. This is only a distance of about 30 miles by air, but travel was difficult in those days.
see this in the following: "On October 18, 1837 Morse wrote to Vail: 'I long to see the machine you have been making and the one you have been maturing in the studio of your brain.' Later Vail invited Morse to Morristown, where the artist realized his cumbersome picture-frame equipment [for recording the signals at the receiving end] was to be replaced by the practical and simple Vail instrument. Morse was so upset, Baxter said, that he became ill and was in bed for some weeks at the Vail home." (21) (Morse's feelings were badly hurt.) If Alfred Vail had not joined Morse as assistant in the latter part of 1837, Morse's telegraph system would no doubt have been a failure.
Vail was not only a skilled technician, but had a wider perspective, and must have quickly seen that Morse's complex translation-coding system and its equipment were not really practical: there must be a better way. "It is evident that Henry showed how to telegraph, Morse planned a cumbersome system to do it, Gale made valuable contributions, and Vail developed the code and instruments necessary for successful operation." (25) On October 18, 1888, over 40 years later. Alfred Vail's widow wrote to H.C. Adams, president of Cornell University: 1888: "... Prof. Morse ... sent for me, and on his dying bed [he died 2 April 1872, almost 81 years old], with the forefinger of his left hand raised and moving to give expression to his words, he said: 'The one thing I want to do now is justice to Alfred Vail." (27 note 18) As for his coding system, "Morse's caveat of October 3, 1837, and his letter to Vail on October 24, 1837 announcing the completion of his dictionary of numbers for words did not mention a dot-and-dash alphabet." -- However, he kept working on it until 1843: "Six years after Vail created the Morse code [1937-8], Morse wrote to [F.O.J.] Smith about the numbers-for-words dictionary he was preparing." (23-24)
Vail, in a letter to his father and brother February 21, 1838, regarding a demonstration he had just given to the President and his Cabinet: "... The President proposed the following sentence, 'The enemy nears . . . .... It was then put in numbers and written on the register." (27, note 16) On p. 39 the caption under picture 2.5: "Alfred Vail who created the Morse telegraph key and sounder and telegraph code at Morristown N.J. while Morse was in New York devising a number for each word commonly used. Morse's idea was to transmit numbers instead of words to send messages." "The Engineering News of April 14, 1886, stated that 'credit for the alphabet, ground circuit and other important features of the Morse system belongs not to Morse at all, but to Alfred Vail, a name that should ever be held in remembrance and honor." (24) F. O. J. Smith wrote: "It is evident that Henry showed how to telegraph, Morse planned a cumbersome system to do it, Gale made valuable contributions,and Vail developed the code and instruments necessary for successful operation." (24-25) "Vail watched Morse gradually eliminate him from credit with mounting astonishment and anger, making no public outcry because Morse, involved in a multiplicity of court battles, required all possible support to preserve the patents. When Morse later referred to Vail and his father merely as 'furnishing the means to give the child a decent dress,' Vail supporters boiled, and telegraph journals contained many strong words." (24)
Vail's Thinking ? It would be very interesting to know the thinking behind the development of Vail's " Morse" code. It had to be tied in intimately with the limitations of the electro-magnetic mechanisms being designed to transmit and receive it. Factors which no doubt strongly dominated in Vail's thinking were: brevity, simplicity and accuracy.
Accuracy requires that the receiving operator be able to distinguish immediately between similar characters without confusion or hesitation. (We must remind ourselves that at this point in time Vail was thinking only of reading a record by eye on a strip of running paper tape, not about receiving by ear as was done later.) We must also realize that while "speed" was commercially important, it was by no means so pressingly demanding in the mid 19th century as it is today. Starting with Morse' simple off-on signalling system Vail developed this original idea into a truly practical alphabetic concept, one that does not require further translation. We may suspect that his key idea was to use more than one signal-on duration. (Did musical rhythms also suggest the internal character spaces?)
This was totally different from Morse' code-dictionary concept. Note: Although Morse, in writing out his code dictionary, is said to have written a dash in lieu of five dots, there seems never to have been any hint of his using such a signal element in his code itself. We cannot help wondering how he determined that the use of longer-than-normal internal spaces between elements would not cause the receiving operator confusion in distinguishing between characters.
Vail do some testing to try it out? These interesting aspects seem to have gone completely unreported in contrast to the attempt to associate the briefest code symbols with the most frequently used English letters, which is well reported (however, as if it were Morse's own work). "In November and December 1837, when Vail built the instruments, he visited Louis Vogt, proprietor of a print shop at Morristown, and, over a case of type, learned which letters of the alphabet were used most frequently... He assigned the fewest dots and dashes to those letters." (23)
By January 1838, about three months after Vail had joined Morse, he had produced the first practical "Morse" code, a purely alphabetic code, which included the use of dashes as well as dots and internally spaced characters. [However, at this point not every letter had a separate code character; several (J = G, Y = I, V = L, and S = Z) were combined. This would be ambiguous for receiving by ear, but more easily handled reading by eye from the context of the inscribed tape record.] This alphabetic code would have made coding and decoding almost perfectly straightforward, and let overall transmission speed jump up immediately to about ten wpm. However, he did not tell Morse about it:-- according to information now available, for six years later Morse was still working on his word-number and number-word dictionary. (Morse was so easily upset by some of Vail's excellent inventive developments.)
It is not clear whether any previous inventor had used more than one length of element in a linear code system. (The idea of "linear" is that of a simple signal running along a line in time, in contrast to simultaneous complexity of signals, such as a two-arm semaphore or a printed alphabet.) Vail chose four kinds of linear elements (besides the necessary minimal spacing between the elements of a character) to form characters:
By 1843, Vail had made such major changes to this early 1838 alphabet that the only letters which were not changed were E H K N P Q. These changes included assigning to each letter a single code character.-- It is not at all clear from a comparison of the alphabets and the relative frequencies of the letters why such extensive changes were made, as the same results could have been achieved by changing very few letters. (Were there other factors involved than mere brevity?) Since Morse knew nothing about this new code (he had many other concerns as well) and no one else would yet be using it, no confusion would result by whatever changes were made.
The average character length of the 1838 alphabet, calculated by the same methods used in Chapter 25 was 8.329. Thus the new 1844 code with average character length of 7.978 was actually about 4% shorter than in the 1838 alphabet. (If he had interchanged just two characters, L and T, in the original 1838 alphabet it would have averaged 7.950 units per letter or 4.5% shorter than it originally was, just a bit shorter than the new 1844 code!) Some other variations could have resulted in a still shorter system.
The 1844 code was thus not the "best" possible, but it proved to be very practical. Vail's final code was used successfully by many thousands of commercial operators, and was the standard for wire telegraphy in the United States, Canada and a few other places until nearly the mid-20th century. Relative timing is critically important to prevent confusion and misunderstanding by the receiving operator. The least bit of hesitation in the wrong place within the character, or holding the key down an instant too long would send the wrong character. If these very tiny differences in timing were disregarded, the following letters within a word would be confused: I, O and EE; C, R, S, IE and EI; Y, Z, II, SE, ES, H and the character &; similarly for "on" signals, T, L and Zero could be confused with one another.
Neither the final 1844 code nor its successor, the International Morse code, is perfect. Perhaps no code could be "perfect" for every application, but it proved practical, and together with the promotion of the telegraph instruments it came into wide and successful use. Its efficiency in other languages will vary, depending on the relative frequency of the letters.
The Art &Skill of Radio-Telegraphy
©William G. Pierpont N0HFF
This page last updated August 02, 1998
Modifications and compile by Thom LaCosta - K3HRN - December 2004