William Godwin's Diary


Lecture by Sir Charles Wheatstone (1802-75), developer of telegraphy, describing the speaking machine he had constructed, an improved verion of the model designed by Wolfgang von Kempelen in 1791. The young Alexander Graham Bell, inventor of the telephone, was later inspired by seeing Wheatstone’s machine and in 1863 constructed his own version of it. The Morning Chronicle report of 9 June 1835 does not give the date of the lecture; however, it seems a very likely match as no other references to Wheatstone lecturing at the Royal Insitution around this time have been found.

‘SPEAKING MACHINES. Some short time since Professor Wheatstone of King’s College, delivered a lecture at the Royal Institution on Speaking Machines. The learned Lecturer commenced with alluding to the speaking heads said to have been made by Roger Bacon and Albertus Magnus, and repeated the story of Thomas Aquinas having broken the machine of the latter, when the sage exclaimed, “There goes the work of thirty years.” Of these machines little was now known. He then proceeded to state that he intended to confine himself almost entirely to a description of the speaking machine made by Kempelin, the celebrated contriver of the well-known automaton chess player. This distinguished mechanic was occupied for a considerable time in endeavouring to find some musical instrument that he could use for the purpose of giving the different vowel sounds; at first he used the bell of the clarionet, the greater or smaller portion of the mouth of which he stopped to produce the various sounds. Mr. Wheatstone then proceeded to produce these different sounds in the way proposed by Kempelin, and afterwards showed that the same sounds would be more distinctly given by passing the air from bellows through metallic tubes of different lengths. He then explained the nature of the different vowel sounds, as laid down by Bishop Wilkins, in his Essay on a Real Character, and afterwards proceeded to show the nature of the machine made by Kempelin, a model of which he exhibited. It consisted of five parts; the first, the reed, representing the human glottis; second, an air-chest, with internal valves; third, bellows, to serve as lungs; fourth, a mouth with extermal and internal lips, andc.; and fifth, nostrils, as in the living subject. The reed was formed in imitation of the reed of a bagpipe-drone, the hollow portion being square, and the tongue of the reed being of the thickness of a playing-card, and made of ivory and resting in it horizontally. The hollow portion was inserted into the air-chest, and the discharge of air occasioned a vibration of the ivory tongue, which produced the requisite sound. The vibrations being regulated by a moveable spring, of course modulated the sound. At the opposite end of the air-chest to that in which the ivory slip was placed, the nosle of a pair of double bellows was inserted. The machine required for each sound six times the quantity of air used by a man in speaking. The mouth was a bell-shaped piece of [illeg] (Indian rubber) applied to the air chest, and so adapted that the sound of the reed issued through it. Independent of its communication with the reed producing the requisite sound, a tin tube connected it with the air-chest, by means of which it might be kept constantly full of air. The [illeg] was composed of two tin tubes communicating with the mouth; when the latter was closed and both tubes remained open, the sound of M was distinctly pronounced; when one of the tubes was closed, the sound N was heard. The sound was regulated by the modification and compression of the mouth. We believe that the above will give a tolerably accurate idea of the nature of the machine. Mr. Wheatstone then produced a number of distinct words from the model of the machine he had made, such as Mamma, Papa, Summer, Plum, andc. The letter R was very indistinctly pronounced, but the lecturer observed that Kempelin had stated that he was satisfied with the sound, as a great many people could not pronounce it better. Mr W. then quoted a passage from Kempelin’s work entitled “Le Mecahnisme de la Parole, suivi de la description d’une machine parlante.” to the effect that it was possible to acquire in the space of three weeks a suprising facility in playing on the speaking machine: above all if it was applied to the French or Italian languages, but the German was much more difficult in consequence of the constant recurrence of consonants in that language. He succeeded in pronouncing any French or Italian word required, but a long German word was mich more difficult. He could pronounce a few entire phrases, but these were necessarily short, as the bellows of his machine was not sufficiently large to furnish the wind: for instance, he could make the machine distinctly say “vous etes mon ami,”—“je vous aime de tout mon coeur;” or in Latin, “Leopoldus secundus Romanorum Imperator semper Augustus.” He was convinced that without much difficulty a machine with keys might be arranged so as to produce any sentence distinctly. Mr. W. then proceeded to allude to two speaking machines presented to the French Academy in 1783, and which were spoken of in high terms by Diderot, but his observations were very indistinctly heard; indeed, this was the case, we regret to have to state, with a considerable portion of the lecture. He concluded with remarking that he conceived that the chief use of such a machine as he had that night exhibited, would be to fix the pronounciation of languages. Kempelin, however, in the preface of the work we have just alluded to, states that he believes it would be found to render easier the mode of teaching the use of language to the deaf and dumb, and also of use in correcting in some the vices of pronounciation, which in general arise from an improper use of the organs of speech. Although it was not stated by the lecturer, we believe that Kemperlin placed his machine within the model of a young girl, which of course appeared to give utterance to the different sounds.’