35 Palaeont. afr., 23, 3~0 (1980) AN OVERVIEW OF PALAEOMAGNETIC CHRONOLOGY WITH SPECIAL REFERENCE TO THE SOUTH AFRICAN HOMINID SITES by P.L. McFadden Department C!f Physics , Universiry of Zimbabwe. Salisbury, Zimbabwe Present address: Research School of Earth Sciences, Australian National Universiry, P.O. Box 4, Canberra ACT 2600, Australia. ABSTRACT The phenomena of secula r varia t\on, polarity reversals and apparent polar wander are dis­ cussed . The calibration of eacli of these phenomena for use in palaeomagnetic chronology is outlined and the use of each of these calibrated scales for dating is briefly explained. A suc­ cessful application of the pola rity reversal d ating technique is presented as an example of the potential for palaeomagnetic chronology in South Africa. In this example it is shown that the age of the important Member 3 in M aka pan is about 3 M y. It is concluded that palaeomag­ netic chronology has a vast p otential in South Africa; a palaeomagnetic laboratory specifi­ cally oriented to chronological problems would be extremely valuable. CONTENTS Page INTRODUCTION .. .. ......................... .............. .... ............ ... ......... ............................. .. .......... .. .... .................. 35 Disadvantages of correlative techniques ........... ....................................... .......... .... ....................................... 35 Advantages of correlative techniques ... ... .......... .... .. .. .... ......... ... .. ................................................ .................. 36 PALAEOMAGNETIC CHRONOLOGY .......... .. .. ... ... .... .... .... ................... .... .. .. .. ... .............. ....................... 36 Secula r variation ................................ ... ... ... ........... ..... ... .. .. .. ............. ... .. ......................... ... .. .. .. ... ................ 36 Polarity reversals ....................... ... ............................................... ... .. ...... .................... ... ............................ .. 36 Apparent polar wander ....................... ... .. ...... .. .... .. ...... ... ............. .. ...... .. ................................ ... ......... .......... 37 USE OF THE CALIBRATED SCALES .......... .. .. ........... .......................................... ........ ... .......................... 37 Secula r variation da ting ................... ... ... ... .. .. .... ... .. .......... .. .. ............. .. ... ..... ... ........................ ... .................. 37 Polarity reversal da ting .. ..................................................................................... .... .... ... ............................ 38 Apparent polar wander dating ................... .. ... .. .... .. .. ... .... ............ .......... .. ............ .......................... ... .......... 38 THE SOUTH AFRICAN HOMINID SITES ...................................... ...... .............................. ... ....... ........... 39 The Makapan polarity pattern ................. ... .. .... .. .. .. ......... .. ............. ....... ............... ................. .... ..... ............ 39 Interpretation ...... ........ .. ................. ... ......................................... ............. .. ............. ............................... 39 Age conclusions ........ .... ...... ................................................................... ............ .. ...... ........ ..................... 40 CONCLUSIONS ....... .... ...... ............................................................................................ ... ....................... ... 40 REFERENCES ...... .... ...................... .. ... ... ........... ........... ............................................................ .. .. ... ............. 40 INTRODUCTION bration with respect to an absolute dating tech­ In the study of fossils it is extremely important nique. Examples of this are faunal correlation and to know the age of specimens so that temporal re­ palaeomagnetic chronology. For faunal correlation lations between specimens, particularly from dif­ the external temporally correlated phenomenon is ferent localities, can be determined. To this end morphological change occurring in animals as it it is necessary to have available a wide range of appears in the fossil record. For palaeomagnetic dating techniques. For any system to be used as a chronology the phenomenon is change in the dating tool there must exist some phenomenon earth's observable magnetic field which is recorded which alters with time, causing a measurable magnetically by the rock. change within the system. If the temporally corre­ lated phenomenon is inherent to the system, pro­ Disadvantages of correlative techniques ceeds in a uniform direction with time and occurs Frequently the correlative phenomenon used at a known rate, then the system may be used as does not alter smoothly or continuously but an absolute dating tool. The most widely used ex­ radomly and often discontinuously. For example, ample is radioactive decay, resulting in radio­ with faunal correlation the morphology of a given metric geochronology. If, however, the temporally animal may remain constant for an extended correlated phenomenon is external to the system period of time and then alter rapidly and signifi­ (the system now acting solely as a recording de­ cantly to another stable form. Thus, on a given vice) and proceeds at a randomly varying rate, correlative scale it may be impossible to distin­ then the system cannot be used as an absolute guish two points in time which may in fact be dating tool but must be considered as a correlative quite widely separated, owirig to the lack of any dating technique. Evidently optimum value from a chang~ in the measured parameters of the system correlative technique is obtained only after cali- during the time interval. 36 Calibration of a correlative scale can only be earth's surface may be very different from the performed at a finite and frequently small number change at another locality even if the localities are of points, thereby producing the problem of inter­ quite close together. Over a period of some tens of polation. For a discontinuous alteration it is fre­ thousands of years the various components of the quently impossible to interpolate, and, even when magnetic field average out to give a geocentric a continuous alteration exists, because the alter­ axial dipole. As a result of palaeomagnetic investi­ ation is external to the system, the actual point of gations it has been discovered that at irregular in­ calibration may not have been recorded in the sys­ tervals the geocentric axial dipole component actu­ tem. Further, the recording of a calibrated event ally reverses its direction, giving rise to the does not ensure its observation in sampling. phenomenon of polarity reversals. Finally, the state of the event being calibrated These two distinct types of change in the direc­ may not be unique to that particular point in time. tion of the earth's magnetic field plus the existence For instance, the direction of the earth's magnetic of slow tectonic processes within the earth give rise field may return to a specific direction on several to three phenomena - secular variation, polarity different occasions. Great care must therefore be reversals and apparent polar wander- each of which taken to avoid errors resulting from such ambigui­ may be calibrated for chronological purposes. ties. Secular variation Advantages of correll_!tive techniques The magnetic direction is defined by two angles, Although the advantages of correlative tech­ the declination and the inclination. Declination is the niques are few, they are of vital importance. The angle measured in a horizontal plane between true obvious advantage is that they often allow the ex­ north and magnetic north. Inclination is the angle tension of chronological investigation to systems between the direction of the magnetic field and the which are not amenable to absolute methods and horizontal, measured in the vertical plane. Both of which would otherwise remain undated. Occasion­ these angles must be calibrated, and in a few ally vast amounts of data can be used to determine places direct instrumental observations of the the calibration of a point on a correlative scale earth's field have been made over a period of sev­ with a far greater precision than is possible with a eral hundred years, thus providing the initial part single absolute reading. Further, since several of the curve. In the Cape direct observation re­ types of absolute techniques may sometimes be cords exist from as far back as 1595, and a com­ brought to bear in calibrating a point on a corre­ prehensive list was published by Beattie ( 1909). lative scale, the accuracy of the age of the cali­ For areas in which historical readings are not brated point may be known with greater reliability available, and in any case for the period before di­ than is possible with a single absolute method. rect observation, carbon-dated material has to be used to construct the calibration curve. Evidently, PALAEOMAGNETIC CHRONOLOGY therefore, the secular variation method can only be In palaeomagnetic chronology use is made of used back to about 2 000 years B.P. The main field changes in the earth's observable magnetic field. of application is consequently archaeomagnetism, The most useful changing parameter is the direc­ a good review of which is given by Aitken ( 1970). tion although for the very recent past