No. 71: Sep-Oct 1990
The advent of radiometric dating seemed to solve once and for all the problem of assigning dates to the key events in the earth's history. Indeed, all of the reference books confidently label charts with firm dates for the appearance of fishes, the demise of the dinosaurs, and so on. Alas, things are not quite as certain as they appear. Radiometric dating is not all that precise; errors may be large indeed.
Take the Pennsylvanian period for example. It is part of the Carboniferous period, when many of the great coal deposits were laid down. The classical duration of the Pennsylvanian -- used in many texts -- is 34 million years. A meticulous new study of central European stratigraphy now pegs the Pennsylvanian as spanning only 19 million years. Now that's a 44% change!
This new figure for the duration of the Pennsylvanian has already cast doubt on the origin of the famous Pennsylvanian cyclothems (repetitive strata) in North America. It had been thought that these seemingly cyclic deposits were correlated with sea level changes forced by variations in the earth's orbit (the Milankovitch periods). With this substantial compression of Pennsylvan-ian time, this correlation falls apart. The cyclothems, which are of impressive area and thickness, now seem to have been created by some other, still unrecognized phenomenon. (Klein, George deV.; "Pennsylvanian Time Scales and Cycle Periods," Geology, 18:455, 1990.)
Comment. Even worse, perhaps, is the fuzziness conferred on the entire geological time scale by this compression of the Pennsylvanian and the possibility of similar revisions for other periods.