The monkeys’ health and weight were monitored daily. Each animal underwent two surgeries as described previously (Elsley et al., 2007). Briefly, in the first surgery we implanted a head post for head restraint, a scleral search coil for monitoring of two-dimensional gaze position and a recording chamber permitting bilateral access to the SEF (stereotactic coordinates Monkey S: AP = 25, ML = 3; Monkey Z: AP = 24, ML = 2). In the second surgery, we implanted chronically indwelling bipolar hook electrodes bilaterally in five neck muscles involved in orienting the head both horizontally PD0332991 purchase and vertically. We focus here

on the activity of three of these muscles (OCI, obliquus capitis inferior; RCP maj, rectus capitis posterior major; SPL cap, splenius capitis; see Fig. 4), as these muscles form the core of the horizontal head-turning synergy (Corneil et al., 2001) and are robustly recruited by extracellular stimulation of the oculomotor system (Corneil et al., 2002; Elsley

LY2835219 in vivo et al., 2007; Farshadmanesh et al., 2008; Chapman et al., 2012). Similar profiles of recruitment were observed across all muscles, so for the sake of simplicity we have pooled normalized recruitment levels across all muscles (see below for a description of our normalization procedure). ICMS-SEF was delivered through tungsten microelectrodes (impedance 0.5–1.2 MΩ at 1 kHz) lowered through 23-gauge tubes secured within a Delrin grid. The stimulation MRIP sites from which we derived the data reported here are a subset of the sites visited previously (Chapman et al., 2012), screened to be those from which a saccade with a predominantly horizontal component could be evoked (neural activity in the vicinity of the stimulation

electrode was not systematically recorded). Briefly, SEF sites were classified as those sites from which a prolonged train of biphasic stimulation pulses (100 μA, 300 Hz, 200 ms) reliably elicited a saccade while a monkey looked around the room; as reported in our previous work stimulation at these parameters also evokes a robust neck muscle response (Chapman et al., 2012). Once an eligible SEF site was encountered, stimulation duration was shortened to 30 ms. Thus, during the behavioral paradigm described below, ICMS-SEF consisted of ten individual stimulation pulses, delivered at 300 Hz (i.e. 100 μA, 300 Hz, 30 ms). While stimulation duration was designed to be very short to preclude evoked saccades, the fixed stimulation current of 100 μA is considerably higher than that used in some studies of the SEF with longer stimulation durations (Schlag & Schlag-Rey, 1987; Martinez-Trujillo et al., 2003; Stuphorn & Schall, 2006), but in the same range as that used in others (Chen & Walton, 2005; Yang et al., 2008; Kunimatsu & Tanaka, 2012).