The geological time scale and how to measure it:

The thing about geological time is that there is lots of it. So much, it's difficult to grasp how much. There are lots of metaphors around - you'll probably have come across the one about imagining the history of the earth compressed into twenty-four hours, and humans not appearing until two minutes to midnight, or two seconds, or however long it is. Personally, I don't think this properly conveys the sheer vastness of it all. Then again, perhaps nothing can, but I think that the analogy below comes close.

Take a piece of imaginary string, and think of your age, to the nearest ten years. Make a mark on the string to represent your own life, one millimetre for every decade. So, if you are thirty years old, make a mark three millimetres from the end of the string.

Now measure forty centimetres of string, and make another mark. This represents the earliest alphabet, 4,000 years ago. Now you'll need to imaginarily get up from your chair. The next mark goes 3.4 metres along the string. This represents the earliest cave paintings, 34,000 years ago.

Now, imaginarily leave the house and walk until you've measured a hundred metres of string. This represents the appearance of modern humans at about 100,000 years ago.

Keep on walking until you've done six and a half kilometres, reeling out the string as you go. This should take you a little over an hour, depending on how fast you walk. This represents the extinction of the dinosaurs, sixty-five million years ago.

Keep walking for another thirty six kilometres. This is when the first land plants are known, 425 million years ago in the Silurian period. This is about a tenth of the age of the Earth, so for most of Earth's history there were no land plants, and hence no land animals.

Only another seven and a half kilometres. This is when the rocks discussed if in the rest of the book were formed, in the Ordovician period, about 460 million years ago.

Keep walking. You need to go another 400 kilometres before you get back to the origin of the Earth. At about the 360 kilometre mark, you can stop for a breather - this point represents the oldest preserved life forms on the surface of the Earth. If you need to, stop again at 380 kilometres, for the oldest rocks currently existing on the surface of the Earth.

Finally, 1,500 kilometres (about 160 kilometres more than the distance from John O'Groats to Land's End) from where you started, you can stop. You've reached the beginning of the universe, 15 billion years ago. Look back along the string, and think of the first mark you made, representing your life. If you're very lucky, your life will be one centimetre worth of string. Small, aren't you?

The previous section was designed, not to give you a sense of your own insignificance, but to show how immense geological time is. Not only is there lots of it, but it is also on a totally different scale. Lengths of time that would be highly significant on a human timescale are barely visible in geological terms. Take the Aberystwyth Grits, for example. These are sedimentary rocks of Silurian age in Wales. Technically they are known as turbidites, which are beds of sandstone deposited in a sudden event. Between the sandstones are beds of mudstone, which represent normal marine deposition between the sandstone events. Although each sandstone bed was deposited virtually instantaneously, probably within a few hours, the mudstone bands in between represent, on average, about four hundred years. There are hundreds of these mudstone beds in the Aberystwyth Grits.

Figure 1: Geological time scale for the Phanerozoic. Stages (Ashgill, etc.) shown for the Ordovician only. The Builth Inlier contains possible Arenig, Llanvirn and basal Caradoc rocks. Stages can be divided into substages, and substages into biozones. Graptolite biozones represent, on average, about a million years.

Figure 1 shows the generally accepted division of geological time. The beginnings of these time units are defined very exactly, for example the base of the Devonian is defined as being the first appearance of the graptolite species Monograptus uniformis at Klonk in the Czech Republic. The endings are not defined, except by the beginning of the next unit of time. This is to avoid embarrassing gaps or overlaps. For example, if it turns out that fossil X, which you used to define the end of unit A, actually occurs later than fossil Y, which you used to define the beginning of unit B, there is a sequence of time that is both in unit A (because of fossil X) and unit B (because of fossil Y). This can lead to confusion. Much simpler just to define the beginning of everything, and say that the earlier one stops when the later one starts. This using of fossils to date rocks isn't circular reasoning, because it is the order of appearance of fossils that is important. This order can be worked out independently for several different places. If the same fossils appear in the same order at different places, it is reasonable to infer that this apply in other places too. As a result of some thousands of person-years of geological mapping, the sequence of fossils through time is well known.

Within an area, rocks are often correlated by their individual characteristics. This is called lithostratigraphy. For example, the merest glimpse of a bright red sandstone in northern England is enough to make geologists say "Aha! Permo-Triassic New Red Sandstone!" and feel very pleased with themselves. This method tends to be useful only over relatively small areas, but is much more accurate than biostratigraphy. For example, an ash bed represents an almost instantaneous event, so allows recognition of a single time plane. However, because very few events leave global markers in the sedimentary record, fossils are generally used for correlation between areas.

Unfortunately for palaeontologists, it's not quite as simple as finding a fossil, checking the right book and finding the age of the rock; not all fossils are useful for dating. Some fossils are so rare that they are effectively useless, as you'll never find one when you need them. Archaeopteryx, for example, is only known from eight specimens, all of which are from a single locality anyway. Some fossils are only found in a small area, so can't be used to correlate the rocks in other places. Exceptionally preserved fossils, such as soft-bodied animals from the Burgess Shale, fall into this category. Some remained the same for such a long time that they can't be used with any precision. The classic example is the inarticulate brachiopod Lingula, which has remained the same for the past 500 million years. The best sorts of fossils to use are those that are abundant, widespread and evolved quickly. Graptolites fit the bill nicely, dinosaurs do not. Graptolites evolved very rapidly - it is difficult to say exactly how long any individual species lasted, but on average it was probably about 2 million years, with some living significantly less and some very much more. They are abundant - it has been calculated that there are about three billion of them per cubic kilometre of black shale - so you can usually find one, as long as you're looking in the correct sort of rock. (Graptolites aren't usually found in very shallow-water deposits, probably because they were destroyed by turbulence.) Some species are found globally, which means that it is possible to correlate graptolitic rocks across the entire planet. In addition, many species of graptolites existed at the same time, so it is possible to correlate using many species rather than just one. Builth Inlier rocks are mostly dated using graptolites (see the section on stratigraphy in the research page for more details).

Graptolites aren't the only fossils that are used for dating. Microfossils are commonly used, especially planktonic foraminifera (single-celled organisms with calcite skeletons), palynomorphs (spores and pollen). Acritarchs (single-celled algae), conodonts (teeth of a very early vertebrate, depending on how you define vertebrate) and chitinozoans (no idea, but probably some sort of animal egg) are commonly used in the lower Palaeozoic. Microfossils are defined as being any fossil that can only be seen under a microscope. (This definition is not entirely accurate, or satisfactory. Some foraminifera can be several centimetres across, but are still counted as being part of micropalaeontology, whereas something like a very small snail would be considered the job of a mollusc worker, not a micropalaeontologist. Micropalaeontologists are palaeontologists who work on microfossils, not abnormally small palaeontologists. There is no correlation between the size of palaeontologists and the size of the fossils they work on.) Because they are microfossils, they tend to be abundant (often thousands per kilogram of rock). Many of them floated in the upper parts of the water column (foraminifera, acritarchs, chitinozoans), or swam (conodonts), or were blown by the wind (spores, pollen), which means that they tend to be very widely distributed.

Use of fossils allows us to say that two bits of rock are the same age, or which one is older, but fossils can't tell us how many years old a rock is. For that, absolute dating techniques, such as radiometric dating or amino acid racemisation are needed, but that's going outside palaeontology. (Absolute dating gives a numerical age. Relative dating, which is what fossils give us, involves putting things in age order, but doesn't say how many years old they are.) In general, rocks are absolutely dating using radiometric dating - there are other methods, but they tend not to be useful as far back as the lower Palaeozoic. Radiometric dating utilises the fact that some elements are radioactive, and decay into other elements (called daughter elements) at a known rate. The technique involves measuring the amounts of these elements in a rock. Knowing the ratio of the original element to daughter element, and knowing the rate at which one changes into the other, it is possible to work out when the rock was formed. (The actual process is a bit more complicated than this. If you want to know any more details, there are many good sources on the web.) Not all rocks are suitable for this. Trying to get a radiometric date for a sedimentary rock, for example, is a little bit pointless. All you will get is when the constituent minerals were formed, not when the rock itself was laid down. Igneous rocks are what's needed.

How can we put absolute dates on the biostratigraphical scale? Radiometric dating can only be done on igneous rocks, but fossils are only found in sedimentary rocks. Fortunately, volcanic ash bands are often found in sediments - there are many thousands in the Welsh Basin. This allows radiometric and biostratigraphic dating of practically the same moment in time, at least within the limits of accuracy of radiometric dates. However, there are many places throughout the world where, at a particular moment in time, there was no volcanic activity. If you live in Britain and are observant, you will have noticed this lack. In these cases tying in the absolute and biostratigraphic scales is somewhat more difficult, but it can still be done. The obvious way to do it is to correlate the rocks with those in an area that did have volcanism. The correlation can be either by fossils, as explained above, or by the sediments themselves, for example, sea level changes can be global. There can also be geochemical markers useful for correlation, the best-known example being the iridium anomaly at the Cretaceous - Tertiary boundary. Similarly, isotopic records for carbon, oxygen or strontium can be used. But all of these analyses are horrendously expensive, require specialist equipment, and are easily ruined by contamination. In most cases, the easiest, cheapest and most accurate way to date a rock is to look at the fossil content

Back to top