Mirages and Illusions

Perceptual errors in mirages

People sometimes ask about the fantastic images supposed to be seen in mirages: ruined castles, marching armies, and so on. A striking example is this passage from William Beauford's 1802 paper in Phil. Mag.:

Of all the phænomena exhibited by nature in her various operations, there are none more curious and extraordinary than those represented by the reflection and refraction of light from fogs and vapours arising from the sea, lakes, and morasses, replete with marine and vegetable salts. For such vapours, by means of the said salts, form various polished surfaces, which reflect and refract the light of the sun, and even the moon, in various directions; thereby not only distorting but multiplying the images of objects represented to them in a most surprising manner; forming not only images of castles, palaces, and other buildings, in various styles of architecture, but the most beautiful landscapes, spacious woods, groves, orchards, meadows, with companies of men and women, with herds of cattle, walking, standing, lying, &c., and all painted with such an admirable mixture of light and shade that it is impossible to form an adequate conception of the picture without seeing: not any scenery represented by the camera obscura can be more beautiful, or more like faithful representations of nature.

He then describes some Irish mirages in agonizing detail:

The first was seen on the 21st of October 1796, about four o'clock in the afternoon, the sun clear; it appeared on a hill, on the county of Waterford side of the river, and seemed a walled town with a round tower, and a church with a spire; the houses perfect, and the windows distinct. Behind the houses appeared the mast of a ship, and in the front a single tree, near which was a cow grazing; whilst the Waterford hills appeared distinctly behind. In the space of about half an hour the spire and round tower became covered with domes, and the octagonal building, or rather round tower, became a broken turret. Soon after this change, all the houses became ruins, and their fragments seemed scattered in the field near the walls: the whole in about an hour disappeared, and the hill on which it stood sunk to the level of the real field. The hill and trees appeared of a bright green, the houses and towers of a clear brown, with their roofs blue.

On the 9th of March 1797, another similar phenomenon was observed … . It had the appearance of a walled town situated on a hill. On one side were houses in ruins, and the ruins of a castle which seemed to fall into the sea. In the middle were two towers broken, on one of which was a flag flying, with houses in ruins between them and the castle. On the fourth were walls and a round tower with windows, which appeared broken in the middle. The hill on which the whole scene was placed was green and brown, and the buildings purple and brown, clear and brilliant, much resembling a transparent painting. The wall which surrounded the town was of a darker brown, with great holes as if made by cannon shot. …

In June 1801, about five o'clock on a fine morning, all the coast opposite Youghal, on the Waterford side, was covered with a dense vapour; that on the right next the sea had the representation of an alpine country; the distant scarpy mountains seemed covered with snow, whilst the foreground, of a brown colour, resembled woods and a cultivated country. Soon the snow was seen to roll down the sides of the mountains into the valleys beneath, and left the gray rocks of the mountains naked and sharp. … On the left, the river and adjacent country were also covered with a vapour, but of quite different appearance from the former. The country seemed laid out in lawns and improvements, in which were situated three gentlemen's seats; the houses well defined, the windows and doors distinct; some of the windows appeared open, and brass knockers were seen on the doors. From the houses were beautiful shrubberies bordered with white Chinese paling; behind the shrubberies were forests of pines; and distant mountains, in fine perspective, closed the scene.

(The brass knockers are a nice touch.) Robert Greenler says he has even been asked to provide photographs showing such things! The short answer is that these fantasies don't really occur in miraged images; they are perceptual errors. As Beauford went on to explain:

For the vapour, being formed into different parts, the light refracted through them causes the confused appearance of ruins, houses, woods, lawns, &c. in the same manner as a board covered in an irregular manner with black and white spots mixed with lines, will at a certain distance resemble a landscape with woods, ruins, houses, trees, castles, &c., and under such imposing forms as to appear real representations.

Castberg (1804) injected a further note of common sense, in his discussion of the classical Fata Morgana observations of Messina seen from Reggio: the cities are so far apart (6500 toises, or over 12 km) that, at this distance, one would be hard pressed to make out individual buildings, let alone individual figures, trees, sheep, etc. This matter of distance is often overlooked in fantastic reports.

The careful observer of mirages H. W. Brandes (1805) was a bit more explicit about the origin of some claims:

The images are usually so distorted, and the object itself appears in such altered form, that one can easily find colonnades, aqueducts, etc. in these apparitions, with the help of an Italian imagination.

And C. W. Baur (1857), after studying mirages on Lake Constance, asked:

But how is it possible that the mirage displays objects that are not present at all, buildings with cupolas, balconies, columns, groves of palms, where the disappointed traveler finds only boulders, sand dunes, and brush? If we deduct from the account what may be credited to imagination and the inflation by fame in passing the description from mouth to mouth, from book to book, perhaps there remains the following natural solution of the mystery. I saw here and there a roundish mass like a tree, heaps of stones or the like, flowing together with their inverted mirror image to form a double shape, which presented on both sides the appearance of perpendicular walls. Here and there on the lakeshore appeared a steep precipice, as if crossed by perpendicular faults, where I could presume nothing of the sort, and also found there on subsequent examination in the afternoon or from a higher standpoint only the narrow edge of a gravel bank or of particularly bare earth. How a rock of this sort can become a building, a short upright spot can become a column, scrub can become a forest, if Wish and Imagination contribute, may be enlightening.

Psychological factors

There is no doubt that the observer's state of mind influences what is perceived. E. A. Mills, who touches upon mirages in his reminiscences of geological field work, says that

The mirage shows many ambiguous images. Desire often insists we are seeing the thing we want. … A vague or confused mirage … often reveals something in the mind of the onlooker.

Vilhjálmur Stefánsson (1913) makes similar remarks about mistaken perceptions in the Arctic. After describing some confusing mirages, and episodes in which David Hanbury took a lemming for a musk-ox, and Gotfred Hansen took an Arctic fox for a polar bear, he says:

In things of this sort there is always a certain amount of suggestion; Hanbury had his mind centered on musk-oxen, and Hansen was expecting to see a polar bear.
He then tells of seeing a grizzly bear that turned out to be a marmot.

Alfred Löw's book on mirages devotes a whole chapter to psychological effects.

Some folks call this tendency to “see” familiar things in random images pareidolia. Besides the examples given on that website, I'll add the astronomical examples of “canals” on Mars (and other misinterpretations of ill-defined, low-contrast markings seen on planets), which are well discussed by William Sheehan in his book Planets and Perception.

The Moon illusion

Another kind of illusion is seen when the Sun or Moon is at the horizon, and it probably affects the way mirages are perceived as well. This is the famous Moon illusion, which makes everything look bigger at the horizon than it really is.

Astronomers are familiar with the spurious apparent enlargement of not only the Sun and Moon near the horizon, but also the constellations. No doubt the same effect also magnifies the apparent size of mirages, which may help account for some of the exaggerated drawings of them.

The Moon illusion is a true optical illusion; it's entirely psychological, with no physical component. In particular, please note that atmospheric refraction cannot magnify (or even appreciably alter) the horizontal angular diameter of the Sun or Moon near the horizon; and refraction usually reduces the actual vertical angular size.

The illusion is well reviewed in

Helen Ross and Cornelis Plug
The Mystery of The Moon Illusion
(Oxford University Press, 2002)

The illusion is greatly enhanced when recognizable objects lie nearby on the horizon; it's often spectacular when the setting Sun is seen between buildings a few blocks away. Though it was described by Aristotle, there is no explanation satisfactory enough to be generally accepted. There's a good discussion of the Moon Illusion on the Web at


Finally, there's a curious instance of this illusion on YouTube, in which an Englishman found that a wind turbine appeared much closer when viewed between city buildings down a long street than when seen in isolation from the shore. Remarkably, he mis-attributed the illusion to atmospheric refraction, thereby mistaking a genuine illusion for a mirage — the reverse of the usual error. (I won't embarrass this observer by providing a link to his misleading video clip.)


Copyright © 2005 – 2009, 2012 Andrew T. Young

Back to the . . .
main mirage page

or the GF home page

or the website overview page