On sources of error in the biochemical study of perilymph (guinea-pig)