Abstract

There are two principal features of the geomagnetic field which may be observed in the paleomagnetic study of rocks from a given region. It is quite common to find rocks with a polarity opposite to that anticipated on the basis of the present-day field, and the mean directions themselves may indicate virtual pole positions which progressively deviate from the rotational poles with time. The first is usually interpreted as evidence that the polarity of the geomagnetic field may reverse, the second that progressive movement of blocks of continental extent (i.e. continental drift) has occurred.

Although these features provide information which may be of use in the location of the Permian-Triassic boundary and have an important contribution to make in establishing the broad scale paleogeographic features of the time, their use is still limited by the small number of results available. It is only in recent years that collections of sufficient size have been routinely subjected to magnetic cleaning. In the absence of stability tests many older results which appear reliable must be excluded.

Reversals of the magnetic field are fairly common in the Lower Triassic. In stark contrast to this, the geomagnetic field appears to have been remarkably stable during Permian times, fully justifying Irving’s introduction of the “Kiaman Magnetic Interval.” Detailed investigation of the Permian however does reveal a number of periods of reversed polarity, although little as yet may be said of their duration. These reversals are more common in the Upper Permian than in the Lower Permian. The change in behavior of the geomagnetic field appears to coincide with, or approach very closely to, the Permian-Triassic boundary. If use can be made of this, then from the magnetic standpoint the top part of the Ochoan might be included in the Triassic.

The information carried in the mean magnetic declination and inclination provides a means of establishing ancient geography. The data now available from most continents imply that considerable relative displacements of continental blocks must have occurred. The information may be used not only to construct paleogeographic maps but as a means of testing configurations arrived at in other ways. It is possible to represent the results from many sampled Permian horizons in a given region by a single virtual pole, which implies no significant relative motion during Permian times. However, this cannot be assumed to be applicable generally (Embleton, 1970). In the present paper attention has been concentrated upon results from rocks of Permian and Lower Triassic age, but in some regions more general “Triassic” results must be used. Models of Dietz and Holden, Du Toit and others have been examined in the light of the paleomagnetic data and compared with results based solely on paleomagnetic data. It was found that the Dietz and Holden model compared closely with that derived paleomagnetically.