ABSTRACT

Carbon-14 dates more than 10,000 years old are subject to at least one serious error, which is (at the moment) not amenable to correction.

Wood pebbles from a single thin clay layer, in west Florida, give C-14 dates which range from 22,000 to 29,000 years, with no assurance that either of these dates is correct. C-14 dates from various wood fragments and other material in coastal zones appear to provide a date, for a high stand of MSL during the Wisconsin, at almost any time desired from 20,000 years ago to 40,000 or more years ago. Not all of these dates can be correct, or there would have been no Wisconsin. For reasons having to do with the growth and decay mechanisms affecting continental glaciers, it is unlikely that any of these dates are correct. In any case, there is no standard by which "good" dates can be distinguished from "bad" dates.

A suite of samples which has been dated by both C-14 and K-Ar methods yields dates which differ by an order of magnitude or more; either the "young" C-14 dates represent much older materials, or the "older" K-Ar dates represent much younger material. With no additional method of dating, one cannot be certain which type of date--if either--is correct. The likelihood of contamination is higher for the C-14 results, however, and therefore, this suite of dates may include mid-Wisconsin numbers for mid-Pleistocene events.

Morner has reported that a small contamination of late Wisconsin or Holocene carbon may provide mid-Wisconsin dates for pre-Wisconsin materials. The bristle-cone pine curve, widely used to correct late Holocene C-14 dates, cannot be extended back far enough to reduce the uncertainties cited here. It now appears that pre-Holocene C-14 dates are of little value.