location. To obtain a water reward, the monkeys were required to perform a saccade within 500 ms after the offset of the central fixation point toward the remembered location (window of 5° × 5°) and to maintain fixation at the remembered target location for a random period of 300–600 ms. If the animal correctly performed all the required steps, the target stimulus reappeared and the animal received a water reward (Fig. 1a). Eye movements were recorded at 500 and 1,000 Hz using video eye trackers (EyeLink II and EyeLink 1,000, Kanata, ON, Canada). Presentation of the behavioural stimuli, monitoring of the responses, and reward delivery were controlled using CORTEX as the experimental control software (NIMH, Bethesda, MA, USA). Task events, vertical and horizontal eye positions, and digitized neural signals were stored together using either a Plexon MAP or Omniplex System (Plexon Inc., Dallas, TX). We visually inspected the eye traces and the trials in which the eye movement trajectory deviated from the normal pathway were excluded. Monkey 1 correctly performed 94.68% and monkey 2 correctly performed 89.97% of memory-guided saccade trials.
