We presented optical and near-infrared multi-band linear polarimetry of the highly reddened Type Ia SN~2014J appeared in M82. SN~2014J exhibits large polarization at shorter wavelengths, e.g., $4.8$% in $B$ band, and the polarization decreases rapidl
y at longer wavelengths, with the position angle of the polarization remaining at approximately $40^{circ}$ over the observed wavelength range. These polarimetric properties suggest that the observed polarization is likely to be caused predominantly by the interstellar dust within M82. Further analysis shows that the polarization peaks at a wavelengths much shorter than those obtained for the Galactic dust. The wavelength dependence of the polarization can be better described by an inverse power law rather than by Serkowski law for Galactic interstellar polarization. These suggests that the nature of the dust in M82 may be different from that in our Galaxy, with polarizing dust grains having a mean radius of $<0.1 mu$m.
We present an extended optical spectropolarimetry of R CrB from 1998 January to 2003 September. The polarization was almost constant in the phase of maximum brightness, being consistent with past observations. We detected, however, temporal changes o
f polarization ($sim 0.5$ %) in 2001 March and August, which were the first detection of large polarization variability in R CrB near maximum brightness. The amplitude and the position angle of the `transient polarization were almost constant with wavelength in both two events. There was a difference by about 20 degrees in the position angle between the two events. Each event could be explained by light scattering due to short-lived dust puff occasionally ejected off the line of sight. The flatness of the polarization against the wavelength suggests that the scatterer is a mixture of dust grains having various sizes. The rapid growth and fading of the transient polarization favors the phenomenological model of dust formation near the stellar photosphere (e.g., within two stellar radii) proposed for the time evolution of brightness and chromospheric emission lines during deeply declining periods, although the fading timescale can hardly be explained by a simple dispersal of expanding dust puff with a velocity of $sim 200-350$ km s $^{-1}$. Higher expansion velocity or some mechanism to destroy the dust grains should be needed.
