the button more or less, thereby varying the electric current and deflecting the needle of the galvanometer to one side or the other. The instrument was said to indicate a change of temperature equivalent to one-millionth of a degree Fahrenheit. It was tested by Edison on the sun's corona during the eclipse observations of July 29, 1875, at Rawlings, in the territory of Wyoming. The trial was not satisfactory, however, for the apparatus was mounted on a hen-house, which trembled to the gale, and before he could get it properly adjusted the eclipse was over.
It is reported that on another trial the light from the star Arcturus, when focussed on the vulcanite, was capable of deflecting the needle of the galvanometer. When gelatine is substituted for vulcanite, the humidity of the atmosphere can also be measured in the same way.
Edison's crowning discovery at Menlo Park was the celebrated 'phonograph,' or talking machine. It was first announced by one of his assistants in the pages of the Scientific American for 1878. The startling news created a general feeling of astonishment, mingled with incredulity or faith. People had indeed heard of the talking heads of antiquity, and seen the articulating machines of De Kempelen and Faber, with their artificial vocal organs and complicated levers, manipulated by an operator. But the phonograph was automatic, and returned the words which had been spoken into it by a purely mechanical mimicry. It captured and imprisoned the sounds as the photograph retained the images of light. The colours of Nature were lost in the photograph, but the phonograph was said to preserve the qualities even of the human voice. Yet this wonderful appliance had neither tongue nor teeth, larynx nor pharynx. It appeared as simple as a coffee-mill. A vibrating
