The point where the cannulae penetrated the tissue above the brain was constant, (7.5L, 5P mm) and (12.5L, 5P mm) in stereotaxic coordinates for monkey Y and G, respectively. We lowered the cannulae 9 mm and 10.5 mm on average for monkey Y and G. The depth varied only within ±0.5 mm across sessions. The resulting final position of the cannulae was at approximately 4 ± 0.5 mm from the cortical surface estimated as the depth at which we encountered the first neuronal activity. From the MR imaging of the gadolinium spread, we estimated that the center of inactivation area was at (5L, 1P mm) and (7L, 1A mm) in stereotaxic coordinates for monkey Y and G, respectively. These

values differ from the initial penetration points http://www.selleckchem.com/products/CAL-101.html because the cannulae were not

normal but slightly tilted with respect to the horizontal stereotaxic plane for monkey Y and G. The inactivation was contained within the medial wall of the midposterior portion of the IPS. The medial wall of the IPS includes two anatomically distinct areas, the medial intraparietal area (MIP) and the ventral part of area 5 (5v) (Colby et al., 1988; Lewis and MDV3100 mouse Van Essen, 2000; Saleem and Logothetis, 2012). The distinction between MIP and 5v is based on their myeloarchitecture, and the boundary between the two areas reported in the literature ranges from approximately a quarter to half way along the IPS from the posterior end. In the absence of histology, we cannot determine the precise boundary of these two areas and, thus, do not know whether the inactivated area was MIP, 5v, or both. Methisazone In each inactivation session, a stainless steel beveled-tip cannula (28–30 GA, Plastic One) affixed to a microdrive (NLX18, Neuralynx) was acutely lowered to the aforementioned constant location. Then, typically 5 μl (range: 3.5–10) of muscimol solution (5 mg/ml, pH ∼7.4) was injected at 1 μl/min using a 100 μl gas-tight Hamilton syringe and

a micropump system (Harvard Apparatus). The behavioral experiment began 35–60 min after the injection started and lasted up to 3 hr, well within the accepted time for muscimol action (Arikan et al., 2002). These experimental parameters for individual sessions are listed in Table S1. We alternated between inactivation and control sessions. They were typically spaced 24 hr apart. Exceptions were two inactivation sessions with a 2 day separation from the previous control session, and four control sessions with a 3–9 day separation from the previous inactivation sessions. The recovery of function in control sessions was visually noticeable in terms of the reach endpoint accuracy in the interleaved control sessions (Figures S1B, S1C, and S4D). In a subset of control sessions (four sessions for Y, nine sessions for G), 5 μl of saline solution was injected instead of muscimol.