Parallel allocation of attention prior to the execution of saccade sequences.