Vegard's rule can be used in certain circumstances. For example, in 8YSZ (8% yttria stabilized zirconia) then the 8% yttria isn't likely to make a huge difference. RI ZrO2 ~ 2.15 @ 632.8 nm RI Y2O3 ~ 1.85 @ 632.8 nm. Thus modified RI ~ [0.08 X 1.85] + [0.92 X 2.15] ~ 2.12 - 2.13...
John Francis Miller Yes, that's true. Many materials are birefringent and trirefringent (although in the latter case 2 of the RI's are usually pretty close). For low aspect ratio particles (AR < 3:1 in non-ISO 9276-6 parlance) then all axes of the particle are presented to the impinging radiation (see Alex Lips) and thus an averaged RI can be used (see Esper Larsen). In the case of AR's > 5:1 (analogous to liquid crystal formation) then the z-axis is presented preferentially to the impinging radiation because the particles have a tendency to flow align.
The imaginary component of the RI provides even more entertainment, and we have to look to Bouguer (or Beer-Lambert, if you prefer) to find a route through.
Actually, there is one thing to consider with the refractive indices of anisotropic crystals. While the dielectric function is a second rank tensor (function), the refractive index is not. The reason is that the latter is a wave property, so you actually calculate two values for the ordinary and the extraordinary ray (or both extraordinary rays) in dependence of the direction of the wave...