From photography let us pass to a glance at the atmospheric conditions which greatly affect its work. The weather from day to day depends upon factors so variable and unstable that prediction beyond twenty-four hours is unsafe. “Suppose a stratum of air,” says Professor Balfour Stewart, “to be very nearly saturated with aqueous vapor; that is to say, to be just a little above the dew-point; while at the same time it is losing heat but slowly, so that if left to itself it would be a long time before moisture were deposited. Now such a stratum is in a very delicate state of molecular equilibrium, and the dropping into it of a small crystal of snow would at once cause a remarkable change. The snow would cool the air around it, and thus moisture would be deposited around the snowflake in the form of fine mist or dew. Now, this deposited mist or dew, being a liquid, and giving out all the rays of heat possible to its temperature, would send its heat into empty space much more rapidly than the saturated air; therefore it would become colder than the air around it. Thus more air would be cooled, and more mist or dew deposited; and so on until a complete change of condition should be brought about. In this imaginary case the tiniest possible flake of snow has pulled the trigger, as it were, and made the gun go off,--has altered completely the whole arrangement that might have gone on for some time longer as it was, had it not been for the advent of the snowflake. We thus see how in our atmosphere the presence of a condensable liquid adds an element of violence, and also of abruptness, amounting to incalculability, to the motions which take place. This means that our knowledge of meteorological phenomena can never be mathematically complete, like our knowledge of planetary motions, inasmuch as there exists an element of instability, and therefore of incalculability, in virtue of which a very considerable change may result from a very small cause.” In view of the inherent difficulties it is certainly creditable that the predictions of the United States Weather Bureau should prove true six times in seven, greatly inuring to the safety of mariners, of passengers by lake and sea, and to the saving of crops under threat of destruction by storms. CHAPTER XXIII SIMPLIFICATION Simplicity always desirable, except when it costs too much . . . Taking direct instead of roundabout paths. . . . Omissions may be gainful . . . Classification and signaling simpler than ever before. For a simple task the inventor’s means should be as simple as possible. Mr. J. J. Thomas in his “Farm Implements” says:-- Simplicity of Build Desirable.