We have obtained middle Carboniferous to Permian paleomagnetic results from the Junggar block in order to test its tectonic coherence and to reconstruct its late Paleozoic position. From Hoboksar, on the northwestern margin of Junggar, a paleomagnetic pole from basic/silicic flows yields ϕp=299 ¿E, λp=70 ¿N (N=15 sites, K/A95=36/6.4¿). In the same general area, a pole from lava flows, plutons and as well as from remagnetized limestones/sandstones, is at ϕp=350 ¿E, λp=69 ¿N (N=26 site, K/A95=15/7.5¿). Farther to the southwest in Keramayi, a pole from late Paleozoic dike swarms lies at ϕp=230 ¿E, λp=80 ¿N (N=23 sites, K/A95=16/7.8¿). In the sothern margin near Urumqi, from sedimentary rocks and volcaniclastics of the Middle to Upper Carboniferous Qijiagou and Aoertu formations, a pole is produced at ϕp=7¿E, λp=72¿N (N=15 sites, K/A95=28/7.4¿). From the Middle Carboniferous Liushugou Formation and dike swarms intruding into the formation, two poles are located at ϕp=231 ¿E, λp=69 ¿N (n=4 sites, K/A95=46/13.7¿) and ϕp=172 ¿E, λp=55 ¿N (N=7 sites, K/A95=17/15.1¿), respectively, though reliabilities of these two poles are in question.

Magnetic polarities of all the poles in our study are exclusively reversed. Based on positive fold tests and a baked contact test, we believe that these magnetizations were acquired during the Kiaman reversed interval (250--320 Ma). More accurate ages of the magnetizations cannot be resolved at present. Despite widely separate locations, the pole positions are roughly located on a small circle segment around the Junggar block. Similar inclinations suggest a coherent Junggar block since Permian time. Based on the results of this study and other published paleomagnetic data, a mean middle Carboniferous to Permian paleomagnetic pole is calculated at ϕp=329 ¿E, λp=82 ¿N (N=6 poles, K/A95=26/13.4¿). Recognition of late Paleozoic apparent polar wander for Junggar is obscured by a lack of detailed geochronologic data and possible relative horizontal rotations around vertical axis.

The paleolatitude for the southern Junggar predicted by this pole is only 4¿ south of the present latitude. Difference exists between the mean pole from Junggar and Permian poles from other tectonic units in Asia (Tarim, Siberia, Kazakhstan, and North and South China), indicating post-Permian relative motions. At the boundary between Tarim and Junggar, this motion is likely to have a left-lateral sense. ¿ American Geophysical Union 1991