.. is no longer used to three drawbacks associated with it. First is self- absorption. When its beta particle disintegrates, the energy level produced by the breakdown is very low, and difficult to detect. Thus it required a large sample to detect something more than a fraction of the disintegration. Secondly, carbon absorbs rather well.

Many samples were contaminated due to the fallout in the air from atomic tests in the 50s. The samples could be cleaned, but not with the certainty needed for accurate readings. Last, the reaction rate of carbon dioxide varied when exposed to magnesium during the procedure, perhaps giving way to a few dates that were clearly too old for the control sample. Gas-Counting was the method that replaced the solid-carbon method. This method was popular for a time in the late 50s, after which time many laboratories switched to using proportional counting or scintillation counting.

Gas-counting basically involves the rapid combustion of a sample (acidification in the case of inorganics), and then many stages of chemical removal of electronegative and radioactive impurities. Then, because of the presence of radon (half-life of 3.82 days), the sample is stored for two weeks until this impurity has decayed below the limit of detection. This can make the entire process stretch out to as much as a month for a reading. Liquid Scintillation Techniques are also used in some laboratories. Though the chemistry involved is considerably more complicated, it usually achieves the same quality of results as obtained with the gas counting method. However, there are two advantages to the scintillation process.

One is that the equipment needed for counting is more readily available than is for the gas counting method. This is an important cost factor because, while the gas counting method requires much human intervention and processing, the equipment used for scintillation is usually semiautomatic, and therefore uses less manpower. Secondly, there is a much greater flexibility with the scintillation process. With the flip of a switch, the laboratory can change from calibration to a reading for an unknown sample. Accelerator Mass Spectrometry is the latest technique being used to date samples. Whereas the other three are based on observing the decay products of C14, AMS counts the number of C14 atoms present relative to the number of C12 and C13 atoms in a particular sample.

To do this, the AMS system was built on the technology of the mass spectrometer. With the MS, a magnetic field is applied to a moving charged particle, and the particle is deflected from the straight path along which it was traveling. If charged particles of different mass, but the same velocity, are subject to the same magnetic field, the heavier particles are deflected the least. The AMS works similarly, however, the charged particles are driven to very high speeds by large voltages of electricity. This allows the minute levels of C14 to be detected among the higher levels of elements such as N14 and CH13.

The AMS system has two major disadvantages. The first is the high construction and maintenance costs. Building the facility alone can run as much as $2 million. The second is the need for rigorous pretreatments (such as the removal of calcium hydroxyapatite from bone, and cellulose from wood) to keep even minute levels of contamination from causing major errors in dating results. However, AMS is more efficient than the conventional methods for several reasons.

1) It can accurately date a sample 1000 times smaller than can conventional techniques. This means smaller artifacts or specimens can be, for the most part, preserved by the archaeologist instead of being destroyed in the dating process. 2) Other chemicals can be detected more easily in a sample. This allows the laboratory to detect the effects of contamination on the results. 3) With AMS a laboratory can run as much as 1000 tests per year because of the 2-3 hour testing time. With conventional methods, it can take 1-3 days.

Test Results Most of the controversy and misunderstanding surrounding radiocarbon dating revolves around the test results pointing to a date or dates that someone does not agree with. As with all scientific methods, there are limits to what radiocarbon dating can and cannot do. Also, there are questions concerning errors, accountability, and a consistency of standards to be considered. The Maximum age limit for conventional methods is about 40 000 years, whereas the limit for AMS is controlled by the individual machines stability and contamination introduced or missed during the processing of small samples. As the sample gets older, levels of C14 can drop to the point where it is difficult to measure against the background level naturally present.

When this is the case, results are quoted as either “infinite” or “background.” The Minimum age limit for all methods is not less than 200 years. Results pointing to an age of less than 200 years are reported as “modern.” Contaminations here can occur due to the worlds fossil fuel and atomic bomb activity, and results showing more C14 than would normally be seen in contemporary samples have been quoted. Error is present in every scientific process. Normally, experimental error is calibrated by repeating the measuring process. However, this is not very practical in radiocarbon dating due to the sample size, time, and cost factors.

The error term (s), then, is estimated for each sample and applied as a known measurement. Going with the normal distribution curve, the true result is expected to have a 68.3% chance of being within 1s of the experimental result, a 95.4% chance of being within 2s of the experimental result, and a 99.7% chance of being within 3s of the experimental result. To date there is no accepted convention as to just how a laboratory should apply the error estimation. All laboratories include an estimation based on Poisson statistics (industry wide standard based on known counts). But how those statistics are applied will differ from lab to lab.

This is because, over a short time period, the decay rate for C14 is not exactly predictable. However, the percentage of error can be reduced significantly by an increase in count time in the conventional methods. Liquid Scintillation gives eight counts per minute per gram for a modern sample. To insure a maximum error of 1% (an error term of approximately 80 years), a five gram sample would need to be counted for about 250 minutes. The older the sample, the longer it takes to produce the same error term.

Increasing the minutes of counting, or the size of the sample will also affect the outcome in the way of a more accurate result. In addition, several samples must be sent out to different laboratories so that a “blind” and impartial result will be produced. Finally, accuracy and precision are often used interchangeably when speaking in terms of error. However, they are quite different. Accuracy refers to systematic errors in relationship to a true result. There may or may not be precision associated with these results.

Precision, on the other hand, refers to how tightly associated the results are with each other. It basically measures the randomness of errors. These results may or may not be accurate in terms of the true results in this case. Thus, an accurate result is more desirable than a precise result because of the relationship between it and the true result. If possible, though, both are pursued.

Lastly, a statistical model will be applied to eliminate any extreme results before a test result date is settled upon. Calibration Radiocarbon dating works very well when trying to establish a broad estimate of age. The problems start when trying to establish calendar dates. This is where calibration, specifically that of dendrochronology, is most helpful. Dendrochronology, the establishing of calendar dates through the study of tree rings, can calibrate with great accuracy radiocarbon dating so that it can more efficiently be used to establish specific ages and dates for the archaeologist. Long chronologies can be plotted by starting with living trees and overlapping them with older, felled trees.

This process is continued as far back as possible. Through use of trees like the Bristle Cone Pine, we now have chronologies dating back 8 000 years. Dating these findings by radiocarbon methods then gives us a statistical curve of calibration based on the comparison of absolute dates (tree rings) and approximate ages (radiocarbon). Laboratories can then use this statistical model to account for any natural error inherent in the radiocarbon process. Bibliography Baillie, M.G.L. Tree-Ring Dating and Archaeology.

Chicago: The University of Chicago Press, 1982 Bowman, Sheridan. Radiocarbon Dating. Los Angeles: University of California Press, 1990 Michael, Henry N. and Ralph, Elizabeth K. Dating Techniques for the Archaeologist. Cambridge: The MIT Press, 1971 Orna, Mary Virginia. Archaeological Chemistry: Organic, Inorganic and Biochemical Analysis.

Washington: The American Chemical Society, 1996 Purdy, Barbara A. How To Do Archaeology The Right Way. Gainesville: University Press of Florida, 19966 Robbins, Maurice and Irving, Mary B. The Amateur Archaeologists Handbook. New York: Thomas Y.

Crowell Company, 1965 Stokes, Marvin A. and Smiley, Terah L. Tree-Ring Dating. Chicago: The University of Chicago Press, 1968 Graduate Paper written in Spring of 1998. Grade received: A 9.5 pages plus bib.