The Canals of Mars

Summary

From 1877 astronomers drew fine straight lines across Mars, and Percival Lowell argued they were canals built by Martians to irrigate a drying world. Spacecraft mapped the whole planet in 1971 and 1972, and the network is not there.

Most of the lines correspond to nothing on the Martian surface. They were drawn by experienced observers, working at good telescopes, over several decades. Why they drew them is the finding.

What is documented

The word. The Italian astronomer Angelo Secchi had already used the word canale for dark markings on Mars in the late 1850s, for broad features rather than lines. In Italian the word can mean a natural channel or an artificial waterway.

Schiaparelli, 1877. Giovanni Virginio Schiaparelli, director of the Brera Observatory in Milan, observed Mars at the favourable opposition of 1877 with a Merz refractor of 22 centimetres aperture. He produced a new map with a new system of names and reported about 40 canali: narrow dark lines crossing the brighter regions of the planet. The same close approach produced a real discovery. In August 1877 Asaph Hall found the two moons of Mars.

Doubling. In 1879 Schiaparelli reported that some canals seemed to split into two parallel lines, which he called gemination, and at the opposition of 1881 to 1882 he saw the geminations more clearly. From 1886 he observed with a new refractor of about 50 centimetres.

Schiaparelli's own view. Schiaparelli's early reports were cautious. In 1895, in an essay for the Italian magazine Natura ed Arte, he went further. Of the geminations and their regularity, he wrote that the idea of intelligent beings having at least a secondary part in them "non può esser considerata come intieramente assurda" (cannot be considered entirely absurd). Of the doublings that came and went, he wrote: "la supposizione di un intervento intelligente è più che mai indicata" (the supposition of intelligent intervention is more than ever indicated). He presented this as what followed if his observations were true.

Lowell. Percival Lowell, a wealthy Bostonian, founded an observatory at Flagstaff, Arizona, in 1894. From 1896 he observed with a 24-inch refractor built by Alvan Clark & Sons. In three books, Mars (1895), Mars and Its Canals (1906) and Mars as the Abode of Life (1908), he argued that the canals were irrigation works built by a civilisation on a planet running out of water. As evidence he cited their straightness, their uniform width, and the way they radiated from particular points, which he called oases. By 1906 he had recorded more than 400 canals, 51 of which he saw double at least some of the time. He made a Mars globe at each opposition from 1894 to 1916, twelve in all.

The observers who could not see them. A number of professional astronomers reported canals, but many more could not see them. In 1903 Edward Walter Maunder of the Royal Observatory, Greenwich, and J. E. Evans published experiments in which people copied drawings of discs bearing small spots and irregular markings but no lines. Viewed from far enough away, the markings were often copied as continuous lines. They concluded that the canals were an optical effect.

Wallace's critique. In 1907 the naturalist Alfred Russel Wallace published Is Mars Habitable?, a book-length reply to Lowell's Mars and Its Canals. He argued that Mars was too cold, and its atmosphere too thin, for liquid water, and that the markings Lowell took for engineering seemed to him to "bear the unmistakeable impress of being due to natural forces".

Meudon, 1909. In 1909 Eugène Antoniadi, who had long accepted the canals, observed Mars with the 83-centimetre refractor at Meudon, near Paris, then the largest telescope in Europe. He saw no canal network.

Mariner 4, 1965. NASA's Mariner 4 flew past Mars on 14 and 15 July 1965, passing 9,846 kilometres from the surface at 01:00:57 UT on 15 July. It returned 21 pictures and part of a 22nd, covering about 1 per cent of the planet. They showed a cratered, Moon-like surface and no canals.

Mariner 9, 1971 to 1972. Mariner 9 entered orbit on 14 November 1971, the first spacecraft to orbit another planet. A global dust storm, which had begun on 22 September, delayed imaging of the surface until it cleared. By the time the spacecraft was switched off on 27 October 1972 it had returned 7,329 images covering the entire planet. They showed the giant volcanoes and the canyon system now called Valles Marineris. They did not show a canal network.

The comparison. In 1975 Carl Sagan and Paul Fox compared the classical canal maps with the Mariner 9 images. They found that a small number of canals may correspond to rift valleys, ridge systems, crater chains and linear markings of light and dark, but that "the vast bulk of classical canals correspond neither to topographic nor to albedo features". Those canals, they concluded, appear to bear no relation to the real surface.

Leading explanations

Perception at the limit of vision. Established. Near the limit of what the eye can resolve, scattered spots and the ragged edges of dark areas are seen, and drawn, as continuous lines. Evans and Maunder produced the effect in 1903 with drawings and no telescope. A larger telescope in 1909 showed no network, and the Mariner 9 comparison found that the great majority of canals match nothing on the ground. The canals were real experiences of real observers, and the observers were reporting in good faith. The features were not on Mars.

A few canals may trace real features. Proposed by Sagan and Fox for a small number. They found that some canals may coincide with rift valleys, ridges, crater chains or linear light and dark markings. Where a canal lies along a real feature, the eye had something to work from. That does not rescue the network.

Irrigation works built by Martians. Rejected. Lowell's hypothesis. It depended on the lines being real, straight and connected, and the lines are not real. Wallace rejected it on physical grounds in 1907, Antoniadi saw no network in 1909, and Mariner 9 found none.

Current status

Debunked. There is no canal network on Mars. The great majority of the classical canals correspond to no feature of the surface, and the few that may coincide with real rifts, ridges, crater chains or markings do not make a network. Mars has since been mapped by later orbiters in far greater detail. Nothing short of finding the network on a surface that has now been photographed in full would reopen the case.

The registry keeps the record because of who was wrong. The canals were not drawn by cranks. They were drawn by the director of a major observatory, then by a well-equipped observer with an observatory of his own, and by others, among them Antoniadi, who accepted them for years before a larger telescope showed him no network. Near the limit of vision, the eye supplies lines. The case shows how skilled, honest observers can agree on something that is not there.

Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.