While the great strength of steel makes it of pre-eminent value in the arts, steel in the huge dimensions of modern roofs and bridges has the demerit of expanding with heat and contracting with cold in a troublesome degree. A notable case is that of the steel rails on the elevated railroad of New York. If this fault, common to all metals, can be materially reduced or abolished, then steel enters upon a new field of golden harvests. Here, by dint of acumen and skill the goal has been reached by M. Charles Edouard Guillaume, of the International Bureau of Weights and Measures in Paris. A few years ago he began investigating the singular magnetic qualities of nickel-steels. Then in studying expansibility by heat he discovered that when the nickel was increased to 36.2 per cent. the alloy was almost indifferent to changes of temperature, expanding but one part in one million when warmed from zero to 1° Centigrade. Because of this insensibility, the alloy at the suggestion of Professor Thury is named _invar_. In observations of invar which extended through six years, an elongation of one part in 100,000 was detected; subsequently its changes of length each year seemed less than one-millionth. This slight inconstancy may be overcome by further experiment; in the meantime while invar is not available for standards of length of the first order, such as those of the Bureau of Standards at Washington, there is a vast and useful field for the alloy. It offers itself for secondary standards, to be compared at intervals with primary standards at Washington or other capitals of the world. A leading application will be in surveying. Already wires of invar have been employed by the Survey of France with the utmost success, dispensing with the burdensome apparatus formerly needed in compensating variations due to temperature. With invar wires ten men have advanced at the rate of five kilometers a day; ten years before, with ordinary steel measures, fifty men advanced one half a kilometer, that is, with but one fiftieth as much efficiency. In time-keeping invar is likely to be as valuable as in surveying. At the Bureau of Standards and the Naval Observatory at Washington, pendulums of invar have been adopted with gratifying results. In ordinary watches and clocks the alloy will banish the compensating devices now requisite, of brass and steel which expand with heat and shrink with cold. For chronometers of the highest grade it is desirable that invar be improved with respect to its stability, an improvement which appears to be highly probable.