A detailed shipborne magnetic survey in the western North Atlantic consisted of forty east‐west lines, spaced 20 miles apart, extending between 23°N and 34°N and from 75°W to 60°W. Several magnetic provinces are evident. From west to east, there is the magnetically smooth zone previously described, a 100‐km‐wide rough‐smooth transition, then a 300‐km‐wide band of correctable sea‐floor spreading anomalies (the USNS Keathley sequence), whose eastern boundary (called the Bermuda discontinuity) is characterized by anomalies of rather high amplitude, and finally a region in which no certain anomaly correlations could be found. A model is derived for the Keathley anomalies, which, based on three Joides deep drill holes, are supposed to have been generated between approximately 155 and 135 m.y. B.P. at a rate of at least 2.4 cm/yr. This value, based on Joides drilling results, is the average spreading rate between 155 and 120 m.y. B.P. as measured parallel to transform faults discovered by an unpublished airborne survey. Closely spaced fracture zones between 120 and 135 m.y. B.P. suggest slow spreading during that time. This implies a rate exceeding 2.4 cm/yr for the Keathley anomalies between about 34°N, 63°W and 36°N, 67°W. An assumed spreading rate of 3 cm/yr for the Keathley sequence near 30°N leads to a reversal frequency of 3 per million years, approximately the upper Tertiary value. Possible left‐handed transform fault offsets occur in the Keathley sequence at about 23°N and 28°N. Certain resolution of smaller offsets must await more detailed surveys. As none of the observed offsets approach the 180‐km offset of the present ridge at 24°N, it is concluded that the present ridge trace is not an exact image of the early break. The paleospreading pole for the Keathley sequence is found from the rate of northward narrowing of the anomalies and from possible transform faults. The rate of northward increase of anomaly amplitude is compared with models based on various magnetic poles along the North American polar wandering curve. The agreement seems best for a pole between the Triassic and the Jurassic poles, the latter being taken near the present geographic pole, but scatter in the data is too great to be certain. If the magnetized layer is 0.5 km thick, its magnetization intensity is about ±0.005 km. Using linear cross correlation, predominate northeast, northwest, and north trends are found in the magnetic fabric of the uncorrelatable region. Comparison with bathymetric studies suggests that the first two are ‘ridge’ and ‘fracture zone’ trends, respectively. We speculate that the rift between South America and Africa developed about 135 m.y. B.P., causing spreading rates to fall and directions to change in the North Atlantic, thus producing the Bermuda discontinuity, a band of high, rough basement and high‐amplitude magnetic anomalies. The high basement ridge formed 135 m.y. B.P. apparently affected the sedimentation pattern in the western North Atlantic throughout the Cretaceous and most of the Tertiary. The uncorr...
