SPEAKING-MACHINE. krious attempts have been made to imitate the human voice by means of a machine. In Ina, the Imperial Academy of St. Petersburg proposed, as one of their prize questions, an inquiry into the nature of the vowel sounds, and the construction of an instrument for imitating them. The prize was awarded to M. Kratzeustein (` Journal de Physique; xxi.), and his prin ciple consisted in the adaptation of a free reed, in all respects similar to that represented under IIARnosaau, to a set of pipes of peculiar forms, determined by repeated trials. About the same time, M. Kenipe len, of Vienna, best known for his ingenious fraud, the so-called automaton chess-player, succeeded in producing the vowel sounds by adapting a reed to the bottom of a funnel-shaped cavity, and placing his hand in various positions within the funnel. He also contrived a hollow oval box, divided into two portions, attached by a hinge so as to resemble jaws, and by opening and closing them the sounds which issued from a tube connected with the reed produced A, o, ou, and an imperfect E, but not i. He afterwards succeeded in obtaining, from different jaws, the sound of the consonants P, M, L ; and by means of these vowels and consonants he could produce such words as mamma, papa, aula, lama, mulo. He afterwards succeeded in imitating the human organs of speech by having only one mouth and one glottis. The mouth consisted of a funnel- or bell-shaped piece of elastic gum, to which was added a nose made of two tin tubes. When both tubes were open, and the mouthpiece closed, a perfect x was produced; and when one was closed and the other open, an 11 was sounded. He also obtained a tolerable resemblance of the four sounds 13,13, E, T, or rather a modified r, and at length succeeded in producing entire words and sentences, such as Ccmstantinopolis ; opera ; astronomy; Vous ctes mon ami ; Je roes aims de tout mon case. We learn from a letter from
Mr. Thos. Collinson to Dr. Hutton, who had seen the machine in London, and afterwards at Kempelen's house in Vienna, that the sounds issued, not from a speaking figure, but from a rectangular box about three feet long, placed upon a table and covered with a cloth. Kempelen set it in action by introducing his hands beneath the cloth.
Professor Willis, in the third volume of the ' Transactions of the Cambridge Philosophical Society,' shows that by using a shallower cavity than that employed by Kempelen, be could dispense with the introduction of the hand, and by sliding a flat board over the mouth of the cavity could obtain the whole series of vowels, in the order u, o, A, E, I. The apparatus was, of course, connected with a wind-chest and a double pair of bellows. Mr. Willie also used cylindrical tubes containing a reed, with a contrivance for varying the length of the tubes by means of sliding joints. When the tube was much less than the length of a stop-pipe in unison with the reed it sounded 1 : and by increasing the length of the tube it gave E, A, 0, and u ; but what was remarkable, when the tube was lengthened so as to be 14 times the length of a stop-pipe in unison with the reed, the vowels began to be sounded in an inverted order,—namely, u, o, A, r,— and then again in a direct order, 1, E, A, o, tt, when the length of the tube was equal with twice that of a stop-pipe in unison with the reed. When the pitch of the reed was very high, it was impossible to sound some of the vowels, which Mr. Willis remarks is exactly the case with the human voice, female singers being unable to pronounce u and o on the highest notes of their voices.