Measurements were made of the birefringence of bleached rods and cones of six non-mammalian vertebrates. The variables included fixatives and temperature, and the immersing medium namely ethyl alcohol, urea, type of Ringer solution, and sucrose. The temperature variations of the birefringence of rods and cones were similar in frogs and in goldfish, respectively, but showed a significant interspecies difference. Rods fixed faster than cones, and the birefringence of cones is higher than that of rods. The birefringence of frog rods rose with the osmolarity of sucrose, but was unaffected by urea. Over 30% of the rod volume is occupied by solid: fixation is believed to destroy the two types of intrinsic birefringence which are postulated to exist as a result of the experiment on osmolarity, and to leave only form birefringence intact. A theory of receptor birefringence is also tested.