40 On Telegraphic Communication. 



ON TELEGRAPHIC COMMUNICATION BY MEANS OF 

 A NUMERICAL CODE. 



BY LIEUT. J. HERSCHEL, E.E. 



The object of telegraphy is the transmission of a series of 

 words, signs, or symbols, representing ideas ; and this object 

 will be best attained when those words, signs, and symbols are 

 transmitted with the least possible expenditure of labour and 

 time, consistently with a minimum risk of error. To arrive 

 at a just conclusion as to the means to this end, it is necessary 

 to consider the nature of these words, signs, and symbols, 

 and, possibly, of the ideas which they represent. 



The signs and symbols which it is required to transmit are 

 few in number as compared with the words ; the letters of the 

 alphabet, the ten numerals (not necessary since they have names, 

 or may be represented by words), and a very limited number 

 of symbols, such as punctuation signs, algebraical signs, etc. 

 (which may for the most part, if not entirely, be known, like 

 the numerals, by their names, or the words representing them) ; 

 these include all that, in practice, it is desirable to transmit of 

 this class. 



The principal medium of communication of ideas is words. 

 All words are susceptible of scription in letters, and existing 

 telegraphy does transmit them by means of the letters which 

 compose them, or at any rate by the transmission of the more 

 important letters which form the words. By a " word/' I 

 understand any combination of letters which has an accepted 

 signification, whatever place that combination or word may 

 have in syntax or grammar, or whatever its signification may 

 be. Existing telegraphy therefore may be understood to com- 

 municate ideas by the transmission of a series of combinations, 

 simple or complex, of a limited number of symbols and letters. 

 Assuming that- telegraphy can, in practice, only transmit 

 simple signals, out of which can be formed a variety of com- 

 binations, it is clear that the simplicity of communication 

 must depend on that of the combinations necessary to repre- 

 sent the words, etc., communicated. Now the primitive signals 

 are but two in number, by the repetition and arrangement of 

 which letters, symbols, and numerals are represented, 'lhtfse 

 latter may be called the telegraphic alphabet ; and it is evident 

 that the more extended this alphabet, the greater the variety 

 of the arrangement of the primitive signals must be. Accord- 

 ing to the existing system this alphabet contains at least 

 forty individuals. It is the object of these remarks to show 

 that ten will suffice ; and even should it appear that the actual 



