334 MAGNETOSTRATIGRAPHY mathematician Karl Gauss) In the laboratory, a magnetometer is used to measure the intensity and direction of the magnetic vector of the rock sample At this point, all the magnetizations acquired by the rock since it was formed are present The goal is to identify the original magnetization at the time of rock formation (i.e the natural remnant magnetization) Subsequent (younger) magnetizations are referred to as overprints These younger magnetizations need to be removed; in a sense we can think of the process as cleaning up the magnetization history of the rock so that only the natural remnant magnetization remains (or at least can be identified) This is done in one of two ways, depending on which kind of magnetometer is used Alternating-field demagnetization subjects the rock samples to a strong alternating magnetic field that destroys the weaker magnetizations in the rock, so that only the stronger natural remnant magnetization remains Thermal demagnetization heats the rock; during this process weaker magnetizations tend to disappear first, at lower temperatures than the stronger natural remnant magnetization Neither method is usually strong enough to realign the primary magnetic mineral grains in the rock, and thus remagnetize it Magnetostratigraphical Correlation Once the magnetic polarity of a succession of rocks has been determined, a magnetostratigraphy can be established (Figure 3) The basic unit of such a magnetostratigraphy is the magnetostratigraphical polarity zone (magnetozone for short), a body of rock with normal or reversed polarity Now, the problem is to establish whether the observed succession of reversed and normal intervals has a pattern that can be correlated with that of another succession and/or with one or more segments of the global polarity time-scale In other words, this piece (preserved by a local rock succession) needs to be matched with another piece or with a piece of the geomagnetic-polarity time-scale This matching is termed ‘correlation’ In correlation, a signature – a distinctive pattern of magnetic-polarity reversals – is looked for in order to establish a match (Figure 3) The global pattern of magnetic reversals is irregular and nonperiodic, so that distinctive long intervals of magnetic reversals can be recognized Reversals took place wordwide, independently of rock types and environments, and were geologically instantaneous (in rocks millions of years old, the 5000 or so years it takes for the reversal to happen is insignificant) However, polarity events are not unique, so that only a long succession of polarity intervals of distinctive lengths can be correlated Furthermore, sedimentary hiatuses (unconformities) and changes in sedimentation rate can confuse the picture Because of these problems, an independent method of correlation is often needed to provide a tie point (datum) against which to correlate the magnetostratigraphy In other words, some other idea of the general age of the local rock succession – either derived from an index fossil or a numerical age – is usually needed to help narrow the possible correlation of magnetic-reversal histories This means that magnetic-polarity-based correlations are typically not an independent means of correlating strata, although, once an index fossil or numerical age places the local slice of magnetic-polarity history ‘in the ballpark’, the matching of magnetic signatures often provides a more exact correlation than can be obtained from fossils or numerical ages alone Secular Variation The Earth’s magnetic north pole is close to, but not the same as, the geographical north pole This means that, in most places, there is a small east-west difference between true north and magnetic north The angle of this east-west deviation, measured from anywhere on Earth, is called the declination For example, in California the declination is about 20 to the east, whereas in New York it is about 10 to the west of true north The angle that the magnetic field makes with the Earth’s surface is called the inclination At the equator, the inclination is nearly horizontal, whereas at the magnetic pole it is vertical The magnetic field varies globally on geologically short time-scales of a few hundred years These variations in declination, inclination, and field intensity are called secular variation Secular variations are not magnetic reversals, but they are well documented over at least the last 10 000 years, and such palaeosecular variation can provide a succession of magnetic events that may be useful in correlation, particularly in archaeological research See Also Analytical Methods: Geochronological Techniques Lava Palaeomagnetism Sedimentary Rocks: Ironstones Stratigraphical Principles Tectonics: MidOcean Ridges Further Reading Butler RF (1992) Paleomagnetism: Magnetic Domains to Geologic Terranes Boston: Blackwell Kennett JP (ed.) (1966) Magnetic Stratigraphy of Sedi ments Stroudsburg, PA: Dowden, Hutchinson and Ross Khramov AN (1958) Paleomagnetism and Stratigraphic Correlation Canberra: Australian National University [English translation, published 1960]