AUTOMATO N. 



153 



many other operations which require the human hand 

 to regulate them, have often been contrived. But 

 there is one kind of imitation of nature, which, 

 perhaps, infinitely surpasses" either the representa- 

 tions of animals or the Androides, we mean the 

 imitation of speech. This may assuredly be consi- 

 dered as the utmost stretch of invention ; for of all 

 the faculties conferred on living beings, that which 

 most eminently 'istinguishes man is speech. 



A brazen hero, we are told, was once construct- 

 d, which is said to have uttered words ; but this pro- 

 bably consisted in deception only : nor is it wonder- 

 ful that it should not have been easily detected, 

 when we reflect on the propensity of the public lately 

 to view one of little ingenuity with admiration. M. 

 Kempelen, however, a Hungarian gentleman, who 

 had previously distinguished himself by mechanical 

 productions, has undoubtedly effected the imitation 

 of speech where no deception is practised. We shall 

 briefly advert to the principles he followed, as we be- 

 lieve no account of them whatever has hitherto been 

 detailed in Britain. 



M. Kempelen having directed his attention to- 

 wardsthe practicability offorminga speaking machine, 

 limited his expectations to the production of vowels 

 only. At first he entertained no hopes of obtaining con- 

 sonants, far less did he deem it possible to unite them 

 with vowels, and thus express words or syllables. In the 

 course of his investigations, he tried all musical instru- 

 ments, even horns and trumpets, with the view of finding 

 which of them emitted sounds approaching nearest to 

 the human voice : but although he was aware that the 

 reeds of hautboys, clarionets, and bassoons, came near- 

 er the voice of mankind, because there is a faint resem- 

 blance between their operations and the functions of 

 the human glottis, and also knew that a reed stop, 

 called voce humcina, had been adapted to organs, his re- 

 searches were ineffectual. The sound of those reeds 

 was found, on comparison, to be a very imperfect imi- 

 tation of what they were intended to represent. At 

 length having accidentally heard the reed of a bagpipe, 

 he conceived that it exceeded all others in this re- 

 spect, and thence made it the subject of his future 

 experiments. 



M. Kempelen then proceeded to a minute and as- 

 siduous study of the mode in which the human 

 speech is produced, which has led to an interesting 

 dissertation On the Mechanism of Speech. There 

 the anatomical position of all the different organs is 

 shewn and described, and also the different relations 

 of each sound to another. After considering these 

 things, he supposed that the fundamental part of voice 

 consist:, in A. But this was far from aiding his pur- 

 pose ; and he could obtain no other vowel, whether 

 grjTe or acute, from a reed connected with a tube. 

 However, after long study, he contrived a hollow 

 oval box, divided into halves, which were attached 

 by a hinge, thus resembling jaws. These were adapt- 

 ed so as to receive the sound issuing from the tube ; 

 and by means of opening and closing them, he heard 

 the sounds A, O, OU, and an imperfect E ; but no 

 indications of I, or the German ii. His attention 

 was next directed to consonants ; and after the labour 

 of two years, he obtained from different jaws P, M, L. 

 With these vowels and consonants, he could compose 



T0L. Hi. PART I. 



syllables, and even words, as mama, papa, aula, lama, 

 mulo, because each touch of his instrument produced 

 a different sound. Still he had to conquer a great 

 difficulty in the first letter not having ceased when 

 the second commenced ; and on attempting to pro- 

 cure the sounds in immediate succession, the letters 

 were confounded together. Papa, instead of being 

 one word, evidently consisted of so many different 

 letters ; and also the too sudden discharge of air in- 

 to the tube produced a faint K. Thus aula nearly 

 resembled ka-ku-kl-ka. Another imperfection like- 

 wise arose in an aspiration following the consonant, 

 and papa then resembled ph-a-ph-a. As M. Kem- 

 pelen proceeded in ascertaining the possibility o 

 producing the sound of letters, he surmounted these 

 difficulties, though it cost him a great deal of trou- 

 ble. But the proper combination of them he saw 

 must result from imitating nature in having only one 

 glottis, and one mouth, from which all the sounds 

 should issue, and where their union should be form- 

 ed. His invention therefore terminated in construct- 

 ing a machine which, in some measure, imitated the 

 human speech. 



The speaking machine is of simple structure, and 

 consists of only five principal parts. 1. The reed, 

 representing the human glottis ; 2. An air chest, 

 with internal valves ; 3. Bellows, or lungs ; 4. A 

 mouth, with its appurtenances ; 5. Nostrils, as in 

 the living subject. We shall not attempt to expa- 

 tiate on each of these parts, which would lead to a 

 long discussion ; and in order to avoid this detail, we 

 shall briefly explain, so far as we arcable without figures, 

 the general composition of each. The reed, though 

 not cylindrical, is formed in imitation of the reed of 

 a bagpipe drone, which probably many of those who 

 peruse this article may have seen. The hollow portion, 

 however, is square, and the tongue of the reed, which 

 vibrates, consists of a thin ivory slip, resting upon it 

 horizontally. This hollow portion, or tube, is in- 

 serted into the air chest ; and the discharge of air 

 occasioning a vibration of the ivory, the requisite 

 sound is produced. To soften the vibration, the 

 part supporting the slip is covered with leather, and 

 a moveable spring, shifting along the upper side of the 

 slip, brings the sound of the reed to the proper pitch. 

 The sound is more acute as the spring is moved for- 

 ward to the outer extremity, because the vibrations 

 then become quicker; and if shifted farther from the 

 anterior extremity, the sound becomes more grave, 

 as the vibrations are then slower. The extremity of 

 the ivory slip should not be applied close to the tube 

 where it rests, but should remain a little open, that the 

 air may penetrate and occasion the vibration. Thus 

 we observe that a common bagpipe reed may be 

 closed and produce no sound. A slight curvature 

 of the ivory slip arises from the pressure of the 

 spring, which is enough for the object desired. 



One end of the air chest, which is of an oblong 

 figure, receives this voice-pipe, as we shall call it, 

 containing the reed ; and into the opposite end is in- 

 serted the mouth of the bellows. Both the apertures 

 are guarded by leather, to prevent the unnecessary 

 waste of air. Two smaller air chests are then put 

 into it, each having a valve above closed by the pres- 

 sure of a spring, and each having a round aperture 

 u 



