iced. For younger samples—ty t contamination does not alter somuc for older samples it can be a serious problem because so feoms arebeing counted. In t instance, to borro is like miscounting by a dollaro a t is more like miscounting by a dollar o count.

    Libby’s metion t t of carbon-14 in tmospe at t ory. In fact it  been. e no tmosp Eartism is deflecting cosmic rays,and t t can vary significantly over time. t some carbon-14 dates are more1If you oms determine  session, t t a statistical convenience-a kind ofactuarial table for elemental terial  isntt every atom in t for exactly 30 seconds or 60 seconds or 90 seconds or some otidilyordained period. Eacom  survive for an entirely random lengtime t o do iples of 30; it mig until t mige auries to come. No one can say. But  for te ofdisappearance s an average rate, in ot to any large sampling. Someone once , for instance, t dimes  30 years.

    dubious ticularly so es just around time t people firstcame to tter is so perennially in dispute.

    Finally, and pertle unexpectedly, readings can be t by seeminglyunrelated external factors—sucs of tested. Onerecent case involved te over s in  monks in amonastery graveyard  tial conclusion t t into doubt by tion t ten alot of fiso be older t t  got to tantalizinglyunresolved.

    Because of ted scomings of carbon-14, scientists devised oting ancient materials, among trons trappedin clays, and electron spin resonance, ic ions of trons. But even t oft date anyt 200,000 year