Details are also given of a device which compensates for the spurious partial plane polarization that otherwise causes slight differences in reflectance readings when the polarizer on a microscope is rotated, and which simplifies the operations involved in making reflectance measurements on very small grains.