Comment on the Measurement of Cosmological Birefringence

Nodland and Ralston (astro-ph/9704196) claim evidence for a wavelength-independent cosmic birefringence. Although it has been shown (astro-ph/9704247, astro-ph/9704263) that the data analysed did not support this conclusion, the possibility of such an effect remains interesting. Here I present a much more accurate test using high-resolution radio images of distant radio galaxies and quasars. A preliminary application sets an upper limit about thirty times lower than the original claim.