Initial experiments by Mishkin revealed that damage to either the hippocampus or the amygdala had very little effect on delayed nonmatching to sample (DNMS) performance.