This paper summarizes results from paleomagnetic studies sponsored by Lithoprobe on the Kapuskasing structural zone (KSZ). Data from Archean rocks outside the KSZ indicate that the Wawa Subprovince has not been significantly rotated or translated (< 5°) relative to the Abitibi Subprovince. Results from the granulites and amphibolites indicate that the KSZ underwent several kilometres of uplift at ca. 2.51 Ga and then 10 ± 5° west-northwest tilt with several kilometres of further uplift between 2.04 and 1.88 Ga from thrust faulting on the Ivanhoe Lake fault zone. Localized chemical remagnetization occurred at 1.1 Ga along the west side of the Shawmere anorthosite. Paleomagnetic data from the 2.45 Ga Matachewan diabase dike swarm indicate that it was emplaced within one reversed to normal polarity interval of less than 5 Ma. Their polarity pattern indicates major north-trending faults with several kilometres of dip-slip displacement. Their remanence confirms that the Superior Province was deformed around the KSZ into an oroclinal flexure with 40° changes in trend between 2.04 and 1.88 Ga. Results from eight 1.1 Ga alkali syenite–carbonatite complexes show that the KSZ and adjacent subprovinces have undergone only minor uplift (< 6 ± 2 km) since emplacement. Also, these data refine the radiometric ages of some complexes, demonstrate that the use of superchrons to correlate Keweenawan units in the Midcontinental Rift is unsound, and show that Keweenawan magnetic field was symmetrical. Many specific conclusions that relate to a given unit or limited area were drawn in the KSZ paleomagnetic studies.

SUMMARY The Early Jurassic Telkwa Formation, comprising the base of the Hazelton Group, was sampled at three areas in the central part of the Stikine Terrane, British Columbia. Detailed alternating field and thermal step demagnetization on samples from nine sites from the Red Canyon area and 10 sites from the Zymoetz River area, both of which are in the Bulkley Ranges, and eight sites from the Telkwa Range area isolate stable remanence magnetization components. The components are interpreted to be primary magnetizations because of the presence of two polarities and the much improved agreement of their inclinations after full tilt correction of the sites from the Bulkley Ranges. Within‐locality declinations agree well; between‐locality declinations are discordant. The expected cratonic declination is ˜340° whereas declinations from this study are 265°, 227° and 328° for the Red Canyon, Zymoetz River and Telkwa Range areas respectively. These results are similar to those of Monger & Irving (1980) and we concur that the discordant declinations are the result of differential rotation about vertical axes between sample localities. Our observed mean inclination of 52° is in good agreement with Monger & Irving's (1980) result of 54°. They interpreted this to indicate ˜1300 km of northward displacement since the Early Jurassic and their interpretation was substantiated by several later studies. However, with revisions to the geologic time‐scale and the Early Jurassic reference pole for North America, the Early Jurassic Canadian Cordillera palaeomagnetic data base appears latitudinally concordant. This indicates that Terrane I and southern Wrangellia were in much the same latitudinal position relative to the North American craton as they are now. Tectonic displacement models developed in view of discordant Cretaceous palaeomagnetic data from the Canadian Cordillera must now consider latitudinally concordant Permian to Early Jurassic data and Middle Eocene to Recent data. This tectonic senario appears more reconcilable with the regional tilt hypothesis. However, this hypothesis cannot account for discordant Cretaceous data from bedded volcanic rocks in Terrane I. It is therefore probable that Cretaceous Cordilleran tectonics were characterized by both regional tilting and moderate latitudinal displacement.

Well-grouped stable magnetizations have been isolated at 14 of 20 sites sampled from the Spetch Creek pluton. The single-polarity primary magnetization directed at D = 026.2°, I = 73.5° (α95 = 4.4°, paleolatitude 59 ± 7°N, paleopole 73°N, 078°W, A95 = 7°) was acquired around 88 Ma during the Cretaceous normal polarity superchron (118–84 Ma). This direction is discordant from the expected mid-Cretaceous direction (D = 332.5°, I = 75.1°) for North America. The difference could be caused by one of two end member models: the all-tilt model requires a 15° east-side-up tilt about a horizontal axis striking 353°, and the displacement–rotation model requires 330 ± 770 km of northward displacement combined with 54 ± 14° of clockwise rotation. Regardless, this result provides a negative test of the Baja British Columbia model, which requires ~ 2400 km of northward displacement.A review of previously observed mid-Cretaceous magnetizations from the Coast Belt, which are also discordant, indicates that they exhibit common characteristics, although their discordance is not uniform. Assuming that present horizontal approximates paleohorizontal, post-mid-Cretaceous latitudinal displacements inferred from individual results vary between 330 and 3500 km northwards. Relative rotations about vertical axes vary between 17 and 57° clockwise. Such variation cannot be accounted for by the displacement and rotation of a superterrane as a whole. Recent studies emphasize that many of the intrusions have probably been locally tilted. Consistent with known geology, these discordant poles are best explained by the "tilt and moderate displacement" model. This model invokes moderate (500–1000 km) post-mid-Cretaceous northward displacement of the amalgamated Insular, Coast, and Intermontane terranes west of the major dextral Canadian Cordilleran fault systems, combined with variable local block tilting east- and northeast-side-up. Northward displacement was driven by Kula and (or) Farallon – North American plate interactions from 90 to 56 Ma. Tilting is most likely due to a combination of northeast–southwest compression, differential uplift, and extension.

Paleomagnetic measurements have been completed on 400 specimens from dated Archean granites and Matachewan dikes in the Michipicoten and Gamitagama greenstone belts in the western Wawa Subprovince of the Superior Province, Ontario. Detailed alternating-field and thermal step demagnetization analyses were used to isolate stable remanence directions. A single-component remanence was isolated within three adjacent dated granitic plutons on the eastern margin of the Michipicoten belt, including the Hawk Lake trondhjemite, the Southern external granite, and the Eastern external granite (HSE). The maximum possible age for this remanence is constrained by the intrusion of the last pluton at 2694 Ma. The corresponding HSE paleopole is located at 10°W, 41°S (dp = 8°, dm = 13°). A second paleopole, NB, is derived from the Northern external granite and the Baldhead River quartz monzonite, which give U–Pb zircon ages of 2662 and 2668 Ma, respectively. Their single-component remanence defines a paleopole on the Archean apparent polar wander path (APWP) at 15°E, 27°S (dp = 8°, dm = 13°), with a maximum possible age of 2.66 Ga. A third paleopole, GD, is derived from the north-northwest-trending Gamitagama diabase dikes and yields a position of 57°E, 41°N (dp = 7°, dm = 14°), which agrees with poles determined by other workers from the 2454 Ma Matachewan dike swarm. The GD pole, along with previously determined Matachewan dike poles, demonstrates that a tectonically stable craton has existed since intrusion of this extensive dike swarm, and it improves the precision of the 2454 Ma Matachewan pole on the APWP. These poles, when compared with coeval poles from the eastern side of the Kapuskasing Structural Zone in the Superior Province, imply no tectonic rotation or translation between the Wawa and Abitibi subprovinces along this Early Proterozoic structure.

The Middle Eocene Ootsa Lake Group is exposed in the central portion of the Stikine Terrane, where it was sampled along the shoreline of Tahtsa Reach and Whitesail Reach. The group consists of dominantly subaerial flows, which range in composition from basalt to rhyolite, that unconformably overly the Jurassic Hazelton Group. Detailed alternating-field and thermal stepwise demagnetizations were done on all specimens from the 21 sites collected. The presence of a normal- and reversed-polarity remanence, a positive fold test, and high coercivities and unblocking temperatures indicate that a prefolding primary remanence has been isolated. The mean tilt-corrected direction of D = 002.2°, I = 69.2 °(α95 = 7.4°) from 13 sites for which paleohorizontal is well known yields a pole position at 354.6°E, 88.0°N (A95 = 11.5°), which is statistically indistinguishable from published 50 Ma reference poles for cratonic North America. This evidence demonstrates that the proposed large-scale northward displacement of Stikinia since mid-Cretaceous was completed by at least Middle Eocene time. This result is consistent with other paleomagnetic results from Stikinia, Quesnellia, and the Coast Plutonic Complex indicating that much of the allochthonous Cordillera had assembled and docked with cratonic North America by the Middle Eocene.