ot i u o

as a proscription.3 temporal resolution of these maps can improve sky-In fact, the extensive literature on clear-sky polar-ization left by Rayleigh’s contemporaries has been greatly expanded in 20th century studies by ~among others! Dorno, Jensen,4 Neuberger,5 Sekera,6 Coul-son,7 and Volz.8 Many of these studies ’ measure-ments have been made with naked-eye polarimeters either held by hand or mounted on a clinometer.8,9 Yet even electronic polarimeters usually yield only spot observations of polarization averaged over their narrow fields of view ~FOV!. This restriction has tended to limit the angular extent of polarization studies, which often are confined to the Sun-zenith ~or principal! plane or to the zenith itself.7,10 light polarization models as well as the physical in-sights that underlie them. 2. Measuring Skylight’s Degree of Linear Polarization Compared with such seemingly straightforward quantities as skylight’s radiance and chromaticity, measures of its polarization may appear obscure to newcomers. Graphs of polarization ratio versus scattering angle16 and hemispheric maps of Stokes parameters and of ~partial! polarization planes17 can at first seem far removed from simple observations of skylight through a polarizer. Of course, historically, skylight polarization was first quantified by such simple means. To identify skylight’s neutral points, observers such as Brewster and Arago needed only to interpret visually the images that they saw through a polarimeter ~e.g., the description of the Savart polari-scope’s use in Ref. 5!.18 To quantify polarization further, we now use Stokes parameters to define various measurable de-grees of polarization. For Stokes parameters I, Q