The posttectonic St. Lawrence Granit (360¿5 Ma) and associated subvertical dikes have two magnetizations, N and H, N points north with an intermediate upward inclination (declination, 005¿; inclination, --50¿; alpha (95), 10¿; paleopole 12¿N,120¿E). We consider N to have been acquired close to the time of initial cooling. The H magnetization points south with low inclination. H is also the dominant magnetization in the nearby Permo-Carboniferous Spanish Room Formation. H is considered to be an overprint acquired chemically during the Kiaman Reversed Superchron (Late Carboniferous and Permian). The N magnetization corresponds to a paleolatiude of 31¿7¿S, very different from the near-equatorial values expected from the commonly accepted Late Devonian and earliest Carboniferous (Tournaisian) reference paleolatitude for cratonic North America. This is generally consistent with the hypothesis of a southerly displaced Acadia proposed by Kent and Opdyke, although the displacement is much greater (30¿ rather than 15¿). However, we show that all inclinations observed from Middle to Upper Devonian and Tournaisian rocks from the craton are not significantly different either from the Kiaman paleofield, or from the H overprints observed from Newfoundland. Declinations in some other cratoic studies show small departures of only a few degrees (with one exception detailed, but it is noteworthy that the declination observed in the most comprehensive of them (Catskill red beds) also agrees with the Kiaman paleofield and with the H overprints from Newfoundland. Declinatiions in some other cratonic studies show small departures of only a few degrees (with one exception based on limited sampling and two from the Colorado Plateau which may have been rotated) which could be caused by experimental inaccuracies. Moreover the directions obtained from cratonic studies are 97% reversed, as they are in rocks with Kiaman magnetizations. We suggest, therefore, that the commonly accepted Late Devonian to earliest Carboniferous reference paleofield for North America is either entirely Kiaman in age or has been seriously affected by Kiaman overprinting. Consequently, the 30¿ of southerly displacement which the N magnetization of the St. Lawrence Granite appears to indicate, and the 15¿ invoked by Kent and Opdyke, we consider to be illusory. Three Late Precambrian rock units (623 to 606 Ma) have scattered declinations but similar inclinations yielding paleolatitudes of about 35¿. The scattered declinations presumable were caused by local pre-Carboniferous relative rotations. There is little or no evidence of Kiaman overprinting in Precambrian rocks, indicating that overprinting is comparatively superficial and is probably related to Late Paleozoic groundwater circulation.