Explorers, navigators, and surveyors who relied on sextants, octants, or transits to solve latitude and longitude locations normally determined the height (angle) of the sun or other astronomical body above the horizon (altitude) at a precise time to solve such problems. If for whatever reason (e.g. mountainous terrain, ice, fog, etc) the natural horizon can not be seen calculation of place become difficult. In that case the navigator can resort to an 'artificial horizon' to estimate the angle of, say, the sun above the unseen horizon. This is done by determining the angle of the reflected image of the sun with the sextant. The image is reflected from a level and steady mirror, and its its apparent angle is exactly twice that of the the angle above the actual horizon. The problems with this include the facts that the mirror must be perfectly level and steady for good results. In the first case the level criterion can be met by flooding the bottom of a box with liquid mercury which when undisturbed with provide a level reflective surface. At sea, however, the movement of the mercury can only be assured by either going ashore for the observation, or invoking an elaborate gimbal or other motion damping technique. (interestingly some early Arctic explorers, when faced with not having liquid mercury, sometimes used molasses as a substitute—it wasn't as 'shimmery.” and even as a frozen thin plate was reflective enough to do the job.
While artificial horizons were manufactured for use by surveyors working in mountainous terrain, these were generally not used by navigators or ocean explorers. Rather they used “make do” solutions to navigation in (usually) Arctic environments where ice and fog were common impediments to horizon visibility. Moreover, floe ice cover damped wave action and provided a stable base for “going ashore” to make observations. Therefore, most artificial horizons used nautically were ship board made, and are rarely seen today—even in museum collections (the New Bedford Whaling Museum, for example, has but a single example}. This example is 6 3/4” long x 5 1/4” wide x 3 1/4” high (roughly the same as the New Bedford Museum specimen. It is made from three panes of glass (bottom and two sides) and solid ends of wood, with a cork stopper where mercury would have been added. The glass is thickly glazed together with wire supports which are thickly coated with fabric and paint. It is a rare example for a nautical or surveyor collection. Good+