When the horizon of the visible landscape is obscured by darkness or inclement weather such as fog, or by other causes,
making it difficult to measure the altitude of a celestial body, an instrument known as an artificial horizon is used to serve
as an alternative to obtain accurate altitude readings on shore. When properly leveled so that a reflection of the sun is
obtained, the angle between the sun and its reflection is measured and then halved to obtain the altitude. The instrument is
particularly useful when high precision is required, such as for setting chronometers.
With the development of reflecting instruments in navigation, the horizon of the natural landscape became one of the bases
for measurement of the altitude of the sun or a star. When the natural horizon was not sufficiently visible, a replacement was
necessary. A mirror could provide a simple solution if it were possible to set it upon a horizontal plane with sufficient
precision, which, however, was rarely the case. The invention of an artificial horizon was inspired when it was realized
that a liquid automatically sought its own level, an instrument that subsequently took various forms.
Land Horizons versus Sea Horizons
When taking a sight with a celestial object, the sea horizon that is vertically below the object the altitude of which is
required must be distinct and clear. When the altitude of a celestial object is to be taken at sea, the observer uses the natural
or sea horizon as the line of departure when the horizon is obscured, or when a sextant cannot be conveniently used. On
shore, however, he has recourse only to an artificial one to which his observations must be referred. This consists of a
reflecting plane that is parallel to the natural horizon, upon which the rays of the sun or other celestial object fall, and are
reflected back to the eye placed in a proper position to receive them. The angle between the real object, and its reflected
image being then measured by means of a sextant, is double the altitude of the object above the horizontal plane. The
observer using a sextant could dispense with the visible horizon when taking sights of the sun or of a star even when
visibility was acceptable by substituting the artificial horizon.
Probably the earliest reference to an artificial horizon appeared in the manual on navigation entitled Arte de Navegar by
Pedro de Medina, an able navigator of wide experience. First published in 1545 at Valladolid, it was translated from the
original Spanish into Dutch, French, Italian and English. Medina described a device to be used by an observer using a
cross-staff consisting of a horizontal bar to be attached on the observers ship to serve as a substitute for the horizon when
Among the first to devise a replacement for an indistinct or invisible natural horizon, was John Hadley (1682-1744), one of
the inventors of the quadrant or octant. In about 1730 he devised a simple spirit level that he attached to his instrument. Two
years later the London instrument maker John Elton (fl. 1730-1732), who had designed a new form of quadrant for taking
altitudes without a horizon, published a description of an improved artificial horizon in Philosophical Transactions. This
consisted of two fitted spirit levels positioned at right angles to each other to the quadrants frame. The difficulty of holding
the instrument steady on shipboard, however, and keeping it perfectly vertical while making an observation made each of
these early forms impracticable.
More Sophisticated Form
A more sophisticated form of the device consisted of a trough containing mercury, the surface of which served as a
reflector. Credited with having been the first to invent the mercury trough was the elder George Adams (c. 1703-1773), the
noted London maker of mathematical instruments. In about 1738 he devised an instrument which, in addition to the mercury
and trough, featured a glass roof inserted gable-wise in such a manner to prevent the wind from affecting the mercurys
tremulous motion. The principle was based on the first law of optics, that the angle of reflection from a mirror is equal to
the angle of incidence. When employing the artificial horizon, the sextant is used to measure the angle between the sun and
its image on the mercury surface, this angle being equal to twice the suns apparent altitude. The apparent altitude is the arc
of a vertical circle between the apparent direction of the observed object and the plane of the sensible horizon or horizontal
plane upon which the observers eye stands.
Useless On Board a Ship
An early description of the mercury horizon appeared in the eighteenth century classic Elements of Navigation published in
1754 by John Robertson (1712-1776). The mercury horizon proved to be useless on shipboard, however, unless the vessel
remained perfectly steady, for the slightest movement caused the mercury surface to tremor and render it useless for
In a later edition of his earlier book, which appeared in 1748, George Adams added accounts of two newly invented
artificial horizons, both of which had been designed by naval men. Captain Henry Ellis (fl. 1748-9) devised an artificial
horizon in which a plate of glass floated on the mercury to check trembling. This problem was that this Floating Mirror of
Captain Ellis all too easily slipped beneath the mercury surface on which it rested. William Wales (1734-1798),
astronomer and master of mathematics at Christs Hospital, did not recommend it in his revision of Robertsons textbook on
navigation, saying that the mercury soon submerged the glass, and the normal roof cover over the mercury was preferable.
Another form of the artificial horizon for use at sea was the whirling speculum or horizontal top, invented by Captain John
Serson (fl. 1735-1750), who was subsequently lost at sea on the H.M.S. Victory. The instrument had been tested at sea in
1743 and tested again by the engineer John Smeaton in 1752, who substantially improved it and described his spinning
artificial horizon with instructions for its use in the Philosophical Transactions in that year. Smeatons top had a polished
speculum surface of a diameter of about 3-1/2 inches, and the top was fitted with a brass ring placed at right angles to the
axis of the top. The sharp spinning point of the top rested upon a cup having a hard surface such as flint or agate, which kept
friction to a minimum. It did not appear to have met with much success in practical use, and in fact in the 1772 edition of
Robertsons work it was reported that mariners found it imperfect.
Many would-be inventors on both sides of the Atlantic attempted to produce improved forms of the artificial horizon during
the nineteenth century. In 1830 Phineas Spear of Portland, Maine, re-invented an artificial horizon in the form of a bubble
level attached to a sextant, for which he was awarded a patent. He described it as A fog glass, a horizontal bar, and a
level. A more portable form of the instrument, described in the catalogue of the firm of Keuffel & Esser as a Reflecting
Horizon instead of a Mercurial Horizon, consisted of a circular plate of black glass about 2 inches in diameter, mounted
upon a brass stand, half an inch deep, and having three foot screws to set the plane horizontal, the horizontality of which
was determined by a short spirit level.
In 1834 an artificial horizon was designed by Lieutenant A. B. Becher of the Royal Navy that consisted of a small pendulum
fitted to a sextant and the motion of which was damped in oil. In the mid nineteenth century the firm of Benjamin Pike, Jr. of
New York advertised the standard type of artificial horizon for $20 to $25 while the circular type was available for $6 to
Tremulous Motion Solved
A major innovation was a compact mercury horizon patented in 1868 by Captain Christopher George of the British
merchant service. It consisted of a circular iron trough containing mercury. A piece of thin paper was placed upon the
mercury surface and then a glass disk having parallel faces was floated over it at the same time that the paper was
simultaneously being removed. The advantage of Georges device was that the entire surface was available for
observation. In 1880 an artificial horizon was developed for the Dutch navy by P. J. Kaiser which featured a reservoir
made of leather and filled with mercury. It had a hollow tray on top covered with a cover made of mica. Squeezing the
reservoir caused the mercury to flow onto the tray. An innovation introduced in 1890 by Admiral G. E. Fleurian of the
French Navy, consisted of a gyroscope made to spin by means of a special mechanism that would continue spinning for
several minutes after being attached to a sextant.
The need for an artificial horizon could be as great for surveyors on land as they were for navigators at sea. A surveyor on
land can experience the same problem while using a sextant to take a sight for finding longitude ashore when no visible
horizon is available, or for checking chronometers. The Lewis and Clark expedition, for example, was equipped with three
artificial horizons. Before departure the explorers were advised in great detail by Major Andrew Ellicott and Professor
Robert Patterson of the University of Pennsylvania about the types of scientific instruments that would be needed and how
to use them.
Ellicott stated that the instrumentation required was relatively simple, such as he had used in the survey of the southern
boundary with the Spanish territory. For determining both latitude and longitude he specified that they would require a good
sextant, a well-made watch with a second hand, and an artificial horizon. He noted that in his experience one that used
water as the reflecting surface in preference to other liquids commonly used proved to be useful when the object observed
was sufficiently bright to reflect a distant image. He cautioned the need to keep the container for liquid separate from the
cover, to avoid potential motion of the liquid that could be caused by the wind. Ellicott also proceeded to construct another
artificial horizon for the expedition of a different form. The block of talc for the reflecting surface appears to have been an
innovation not used by others. Ellicott obtained a block of the substance from the mineral collection of the American
Yet Another Solution to the Problem
Professor Robert Patterson of the University of Pennsylvania provided another artificial horizon that he had made,
consisting of a glass pane cemented to the side of a wooden ball and adjusted by a spirit level and platform. He assured the
explorers that it was more adaptable for taking altitudes of the moon and stars and of the sun under dull conditions. The
third horizon, using the index mirror of a sextant attached to a flat board and also adjusted by a spirit level and platform,
was particularly useful with bright objects such as the stars.
By the beginning of the twentieth century the bubble octant was developed primarily for use in aircraft, and following the
demise of astronomical navigation by mid-century, a sextant with a built-in gyroscopic horizon activated by an air pump
came into use.
Silvio Bedini is a historian emeritus with the Smithsonian Institution in Washington, D.C., and a Contributing Editor
for the magazine.