The apparent pole wander (APW) path for Gondwana is still not clearly established, in particular, for Silurian-Devonian times. A controversial debate places authors who argue for an X path, running directly through Africa on a reconstruction of Gondwana against those who advocate a large loop passing by southern South America, the Y path. Most of the paleomagnetic data used to draw this loop come from the Lachlan Orogen (Australia). A paleomagnetic study was carried out in the well-dated Ambone and Ural volcanics in the central subprovince of Lachlan Orogen, New South Wales. Anisotropy of magnetic susceptibility measurements confirms detailed mapping of the region and shows that these massive dacitic sills and/or lava flows are flat lying. Among the different localities studied, only one yields interpretable paleomagnetic results. Two components of magnetization can be identified: a midtemperature direction yielding a corresponding pole in Australian coordinates λ = 67.9¿S/$phi$ = 084.4¿E (B = 5; n = 21; dp = 17.5¿/dm = 23.1¿) and a high-temperature direction with a corresponding VGP λ = 24.4¿S/$phi$ = 060.6¿E (B = 5; n = 25; dp = 1.4¿/dm = 2.5¿). The first is interpreted as corresponding to an Early Carboniferous pole position and can be regarded as an overprint probably related to the Early Carboniferous Kanimblan orogenic event. The second does not correspond to any expected Silurian-Devonian or younger pole position. This magnetization is thought to be primary in origin; however, secular variation has apparently not been averaged out in the single lava flow sampled. Therefore the earliest Devonian paleopole position probably lies in a 30¿ cone around the obtained VGP, and this position can only match the X-type APW path for Gondwana. It is in particular very different from coeval poles obtained in the eastern subprovince of the Lachlan Orogen, and it is mostly used as key poles supporting the Silurian-Devonian loop for the APW path of Gondwana. Therefore some poles from the Lachlan Orogen must be affected by rotation to explain these results. The Lachlan Orogen was hence not stable up to the mid-Paleozoic, and data from this region should not be used as representative for Gondwana.