The apparatus, squatting on a sturdy table in the small room beyond, had the slightly haywire look of an experimental model. But it was little short of a miracle to those who knew how it had been built--on the basis of radioed descriptions of the ruined device the excavators had dug up on Mars. Even more intriguing, however, was the row of neatly labeled boxes on a shelf. There in cushioned nests reposed little cylinders of age-tarnished metal, on which a close observer could still trace the faint engraved lines and whorls of Martian script. These were the best-preserved specimens yet found of Martian record films. Sound and pictures were on them, impressed there by a triumphant science so long ago that the code of Hammurabi or the hieroglyphs of Khufu seemed by comparison like yesterday's newspaper. Men of Earth were ready now to evoke these ancient voices--but to reproduce the stereoscopic images was still beyond human technology. Dalton scrutinized one label intently. "Odd," he said. "I realize how much the Martian archives may have to offer us when we master their spoken language--but I still want most to hear _this_ record, the one the Martians made right here on Earth." Thwaite nodded comprehendingly. "The human race is a good deal like an amnesia patient that wakes up at the age of forty and finds himself with a fairly prosperous business, a wife and children and a mortgage, but no recollection of his youth or infancy--and nobody around to tell him how he got where he is. "We invented writing so doggone late in the game. Now we get to Mars and find the people there knew us before we knew ourselves--but they died or maybe picked up and went, leaving just this behind." He used both hands to lift the precious gray cylinder from its box. "And of course you linguists in particular get a big charge out of this discovery." "_If_ it's a record of human speech it'll be the oldest ever found. It may do for comparative-historical linguistics what the Rosetta Stone did for Egyptology." Dalton grinned boyishly. "Some of us even nurse the hope it may do something for our old headache--the problem of the origin of language. That was one of the most important, maybe _the_ most important step in human progress--and we don't know how or when or why!" "I've heard of the bowwow theory and the dingdong theory," said Thwaite, his hands busy with the machine. "Pure speculations. The plain fact is we haven't even been able to make an informed guess because the evidence, the written records, only run back about six thousand years. That racial amnesia you spoke of. "Personally, I have a weakness for the magical theory--that man invented language in the search for magic formulae, words of power. Unlike the other theories, that one assumes as the motive force not merely passive imitativeness but an outgoing will.