Early in the twentieth century an American ship, christened the _Carnegie_, in honor of the philanthropist who supplied funds for the enterprise, was constructed for the express purpose of making accurate charts of the lines of magnetic declination in various parts of the globe. This ship differs from every other vessel of considerable size ever hitherto constructed in that no magnetic material of any kind was used in connection with its structure or equipment. For the most part iron was substituted by copper or other non-magnetic metal. Pins of locust-wood largely took the place of nails; and wherever it was not feasible to do away with iron altogether it was used in the form of non-magnetic manganese steel. The purpose of the _Carnegie_ is to provide accurate charts of magnetic declination for the use of navigators in general. The value of observations made with this non-magnetic ship will be clear when it is reflected that with an ordinary ship the observer can never be absolutely certain as to what precise share of the observed fluctuation of the compass is due at any given moment to the ship's influence. In other words--using technical terminology--he can never apportion with absolute accuracy the influence of declination and of deviation. Yet it is highly important that he should be able to do so, inasmuch as the declination of the compass is an all-important element in reckoning the exact location of the ship, and would be the same for every ship at that place, whereas deviation denotes a purely local disturbance which would never be the same for any two ships of different construction. Not only does the magnetized needle thus tend to vary in the direction of its horizontal action, but it also tends when suspended at the middle to shift its vertical axis. In regions near the equator, indeed, the magnetized needle maintains a horizontal position, but if carried into northern or southern latitudes it progressively "dips," its polar end sinking lower and lower. This dipping of the needle seems to have been first observed by Robert Norman, an English nautical instrument maker, about the year 1590. It was brought to the attention of Gilbert and carefully tested by him in the course of his famous pioneer experiments. Gilbert was led to predicate the existence of magnetic poles, the exact location of which would be indicated by the dipping needle, which, sinking lower and lower as northern latitudes were attained, would ultimately at the magnetic pole itself assume a vertical direction.