Lowitz arcs were the most controversial of the halos. They were first recorded in 1790, but even until the 1990s their very existence was doubted because they had not been clearly observed or photographed. Now, see the gallery, digital cameras, easy Internet communication and much increased interest in atmospheric optics combine to produce images each year. Nonetheless, Lowitz arcs appear very rarely. When they do they are usually fleeting, faint and indistinct. They remain enigmatic and we still need more observations and photographs.

Looking for a Lowitz arc? The traditional search is for arcs extending from a sundog to the nearby 22° halo. Do this, but look out for other features. The 22° halo at 10 and 2 o’clock is often brighter where a Lowitz arc touches it. The 22° halo is conversely rather weaker next to the parhelion. The upper Lowitz arc can be most evident above the 22° halo.

Lowitz arcs are closely related to Parry arcs and the upper Lowitz arc is always tangential to a suncave Parry. Therefore, look closely at the region between a Parry arc and upper tangent arc. With luck, a faint Lowitz arc will be seen curving downwards. When the sun is low and the twin ‘V’s of an upper tangent and suncave Parry are visible, look also for a middle Lowitz arc touching the Parry.

The classical explanation for Lowitz arc formation is via Lowitz oriented plate crystals. These ‘rotate’ or take all rotational positions about a near horizontal ‘Lowitz axis’. Lowitz rays pass between two prism side faces inclined at 60°. As with Parry arcs, there are three possible routes and for Lowitz oriented crystals they produce upper, middle and lower arcs. The Lowitz orientation is unusual and perhaps not aerodynamically very favourable. This might account for the arcs’ rarity but there may be other factors.

Recently Marko Riikonen et.al.* analysed several Lowitz displays with clear and well defined arcs and concluded that, for some, the classical explanation was not adequate. The arc relative intensities were better simulated using semi-regular hexagonal plates or prisms rather than the regular ones previously assumed. Furthermore, the observed extent of arcs suggest that crystals take limited orientations about the Lowitz axis rather than all rotational positions. The Lowitz story is not over yet.