Page:Popular Science Monthly Volume 51.djvu/806

788, where one pursuing special lines may make up his own instruments, construct his own machines, or devise models to be used as patterns in the machine shop proper. The power for the building is furnished by three dynamos and an engine located in the basement. An air compressor and a vacuum pump form part of the outfit. The laboratories are large, intended for several students working together, or small, for individual workers. At the present time, students are pursuing investigations upon the mechanical equivalent of heat and the coefficient of viscosity of a liquid. There are eight or ten of these individual laboratories for single students, each supplied with all needed facilities. Every laboratory room, large or small, is supplied with gas for light and fuel, electricity for light and power, water, compressed air, and vacuum pipes. In one of the two constant temperature rooms is a highly interesting piece of apparatus conceived by Prof. Michelson and perfected by Prof. Michelson and Prof. Stratton. Its purpose is the direct production of standards of length. The principle involved is the determination of the length of the metre in waves of light. This determination was first made by Prof. Michelson, whose original investigation was published by the Bureau Internationale des Poids et Mesures, at Paris. The instrument produces light waves, from a given substance, of known length, and then mechanically lays off standards of length, up to two decimetres, in waves of light. Another notably interesting piece of apparatus constructed by Prof. Michelson and Prof. Stratton is a harmonic analyzer, which has cost two years of work. A first pattern was quite fully developed, only to be abandoned for the design shown in the illustration on next page. The purpose of the apparatus is the analysis and synthesis of any harmonic curve (Fourier's series), provided not more than eighty elements enter into it. Every one knows the vast use made to-day of traced curves in all branches of science. These curves are frequently resultants and combinations of two or more simple curves. It is a matter of difficulty to recognize and separate the elements thus combined. The machine in question enables the