Figure 1

Polarization dependence of the viscosity for two experiments regulated at 66 and 76 mK; the heating dc current is switched between 0 and $250\mu \mathrm{A}$ every 9 and 7 s, respectively. Time increases from right to left. The inset shows how we determine the ratio of the viscosity in the cold (i.e., large viscosity) and hot states. For each sequence of three states (black symbols), we eliminate the transients by keeping the last 4 s of each step (large black symbols). $\eta (m,T+\delta T)/\eta (m,T)$ is determined such that, after correction of the first and last steps by this ratio, the data lie on a single local parabola (thin line, red online). Associating the obtained $\delta T$ with the polarization of the intermediate step gives $R_{\mathrm{th}}(m)$.Reuse & Permissions

Figure 2

Polarization dependence of the thermal conductivity of ${}^3\mathrm{He}$ $\kappa (m)/\kappa (m_0)$, for a series of experiments at 27 bar, for different base temperatures and heating powers. The points are obtained assuming a temperature independent $f$ (see text). The dark and open squares correspond to the experiments of Fig. 1. The continuous (gray) noisy line corresponds to the interpolation scheme described in the text, when the above assumption is not made. The experimental polarization dependence is smaller than the predictions for $s$-wave scattering or the $s\mathrm{\text{−}}p$ approximation, shown by the upper and lower continuous lines, respectively.Reuse & Permissions