The idea that the Vikings navigated across the Atlantic using birefringent Iceland spar (calcite) to locate the Sun's position on cloudy days is surely one of the most ingenious and captivating recent hypotheses about ancient materials use. The suggestion itself is an old one1, but has been given strong support in new experiments by Ropars and colleagues2.

The claim, which has understandably enjoyed much media interest, has a lot going for it. References in Viking sagas to a 'sunstone' used in seafaring sound akin to magic, but there is now a good physical basis for thinking that Iceland spar — abundant in the Viking homelands, as the name implies — might be used in this way with sufficient accuracy.

A narrow beam of polarized light passing into the mineral is split into two by the optical anisotropy that causes birefringence. The 'ordinary' beam behaves as it would in glass; the 'extraordinary' beam is parallel but displaced from it, defying Snell's law. Light passing through a hole in a screen over a calcite crystal therefore forms two images on the far side. When the crystal is oriented to equalize their brightness, it completely depolarizes the light.

Sunlight acquires a slight polarization as it is scattered by the atmosphere. The researchers demonstrate how, with calibration on a clear day, the depolarization point at which the split images are equally bright could be used to locate the Sun when it is obscured by cloud, from the light arriving from a patch of blue sky. They have constructed a wooden device containing a calcite crystal that could have been used by Viking sailors to pinpoint the Sun even at twilight.

They say that this proposal is made all the more plausible by the recovery of a piece of Iceland spar from an Elizabethan ship wrecked off the coast of the Channel Island of Alderney in 15923. It may have been carried as an alternative to the compass, they say, which was vulnerable to disturbance from the iron cannons.

The idea that birefringence of Iceland spar might have been used by navigators from the Dark Ages until at least Elizabethan times is, however, problematic. It is hard to see how this would have been common practice without it ever having been recorded or coming to the attention of natural philosophers. It is just the kind of trick Giambattista della Porta, a specialist in optics who conceived of the telescope before its invention, would have revelled in discussing in his 1558 book Natural Magic, had it been known. When Rasmus Bartholin wrote the first experimental account of birefringence in Iceland spar in 1669, he made no mention of such uses; neither did the eminently practical Christiaan Huygens when he offered the first explanation of birefringence 21 years later.

This is not to say that the hypothesis of Ropars et al. is wrong. Rather, it underscores the need for studies of historical materials usage to integrate the scientific arguments with careful consideration of historical sources and context.