In contrast to relative time, absolute time systems attempt to establish the age of an object or event in actual numbers of years. Theoretically, the principle behind absolute age determinations is simple. All one has to do, is to find some non-cyclic system which changes at a constant known rate, determine how much change has taken place and, if one knows the rate of change, one can calculate how much time has elapsed. 

Over time, many people have taxed their ingenuity to find ways of determining the age of the earth. From the time when it became obvious that Bishop Ussher's 6000 year chronology of the earth simply ran counter to evidence, many calculations were attempted. They included attempts to calculate the age of the earth on the basis of the salinity of the oceans, as well as the rate at which the earth lost heat. None of these was entirely successful, not because they were conceptually flawed, but because they lacked sufficient information. Even so, they all clearly pointed to the fact that the earth is extremely old. 

Several techniques have been used with some success within the recent past. They include dendrochronology (aging based on tree rings), lichenology (based on the growth pattern of lichens), and the study of warves, annual layers deposited by the meltwaters of glaciers. All these methods leave something to be desired, being limited to relatively recent materials, but they can nevertheless give reasonable ages. 

Every radioactive isotope decays exponentially, but in a predictable fashion. The rate of decay can be expressed in half-lives, which is the amount of time that it takes for one half the parent atoms to decay into daughter atoms. For any given parent/daughter pair, this half life is constant and is not affected by physical or chemical change. Some commonly used pairs and their half-lives are listed below: 
 
 

Parent/ Daughter	Half-Life
Thorium 232/ Lead 208	13.9 B.Y. billion years)
Rubidium 87/ Strontium 87	4.7 B.Y.
Uranium 238/ Lead 206	4.5 B.Y.
Uranium 235/ Lead 207	0.7 B.Y..
Potassium 40/ Argon 40	1.3 B.Y.
Carbon 14/ Nitrogen 14	5710 yrs.

     .  .    . 

If we use, for example, U235/Pb206 we note that the half-life for this decay series is 0.7 BY. What this means is that after 0.7 BY. half of the atoms that started off as U have now changed into Pb. If we wait another half-life (another .7 BY.) half of the remaining U atoms will again have changed, leaving one quarter of the original U atoms. After one more half life, there will be 1/8 U atoms and 7/8 Pb atoms, and so on indefinitely. Thus, for any moment in time, there will be a unique ratio of parent/daughter material. 

In practice, calculation of the age of a sample is done in the following fashion. A minute portion of the sample is introduced into a mass spectrometer to get an accurate analysis of the composition of the sample. From this we obtain a ratio of parent/daughter material. Using this ratio we can calculate the number of half lives which must have elapsed since the formation of this material. Having calculated the number of half-lives we then multiply this number by the length of a half-life and obtain a date. Once the sample is introduced into the mass spectrometer, all subsequent steps are carried out automatically by machines and computers. 

Although absolute dates seem correct to the last decimal place, the entire system has some degree of error built into it. Analytical techniques are never sufficiently precise to count every last atom, and the smaller the sample, the greater the uncertainty. Computers which do the calculations also calculate the standard deviation of the sample. Suppose that a lab analysis indicated that the age of the sample was 50 million years. (Such a date is meaningless without the degree of variation around the mean, the standard deviation expressed as a + or - value). Suppose further that the standard deviation was given as +or-3 M.Y.. What this means is that there is approximately a 66% chance that the sample that was between 47 and 53 million years old. If you wanted to be more secure about the date you may assume that there is a 95% chance that the date lies between 44 and 56 million years ago and you can be 99% confident that the event took place between 41 and 59 million years ago. Thus, even the best of absolute dates incorporates some degree of uncertainty. 

Aside from laboratory constraints, there may also be geochemical problems such as leakage or addition of parent or daughter material which may have occurred in the past. In fact in some ways, absolute dates are less useful than relative dates. At least the latter will give you the correct sequence of events even if it does not tell you how long ago the events took place. 

The above comments should be in no way construed as a dismissal of the value of absolute dating. The elements used for radiometric dates are found more commonly in igneous and metamorphic rocks. These rocks rarely contain fossils and could otherwise be dated only with the greatest difficulty, if at all. Because of the general lack of fossils in rocks of Pre-Paleozoic age, absolute dates are of great use there. Absolute dates have helped us unravel the history of the early earth, and even though world wide correlations of that earlier time still involve much uncertainty, radiometric dates have given us a fairly coherent picture of what happened then.