It is very easy to photograph sounds and to reproduce them simultaneously with the projection of a picture, but to reproduce all the sounds of a cowboy scene accurately is, at present, almost impossible. If a hero says “Good-bye” to a heroine with a kiss like a creaking board, in the middle of a twenty-reel drama, instead of improving upon the effect of your imagination, which tells you that it is real, and which acts the scene for you better than it can be shown in life or sound, it would be like putting up a blackboard across the screen with the words written upon it, “this is not real--it is only a fake.” That is what is wrecking the talking cinema. As a scientific proposition it is easy, but the results are not good enough at present, and, if we can improve, let us first consider the loud speaker. The talking cinema will come as a matter of course. It is so easy to record sounds upon the film by photography with reproduction by the selenium cell or the neon tube; it is easy to photograph the wave; it is easy actually to impress the sound wave upon the film with the picture, or to use a gramophone; but reproduction is not like a human voice. Neither as yet is any reproduction ever like the voice itself. Let that painful fact be remembered. Unfortunately, from the business point of view, the long distance reproduction effect is usually satisfactory even when re-broadcasted upon telephones, and for business purposes it is immaterial whether the voice that records the profit or loss is harsh or pleasant. So the greatest effects we shall soon see from wireless and sound are these: we shall be able to speak to people all over the world by relaying and a combination of land-line and radio; we shall easily be able to connect our office with a wireless station on the coast, radio across the Continent, and then connect by land-line to another office on the other side of the Atlantic. All this is so easy that no one can doubt that we shall soon listen in to native jamborees; no one can doubt that we shall hear the strange cries of partisans at a baseball match taking place a few miles from New York; no one can question these things, and when reproduction becomes so accurate that the very nature of the people is revealed to us through their speech, surely we might be a little more neighbourly even with those whom we now pretend to love? Relations are notoriously quarrelsome. If you are in a concert hall and the number of people is varied, it will alter the effect of the sound. You have only to look at a sound-wave photograph produced from a violin to realise from its dainty intricacies that the least variation of any of its harmonics or the very exact shape of its wave beats will reveal all the difference between a beginner and the finest musician in the world. But these things are seldom noticed in wireless.