THE fact that an animal is able to discriminate between two stimuli is the foundation of a method used extensively in animal psychology. It is, however, open to various sources of error. While looking through a number of papers dealing with this method I found that some of these possible pitfalls are often not sufficiently realized.