"How long does it take for a stalagmite to grow?" Is there a single cave tour when somebody doesn't ask that question? And what can we answer? There used to be a standard reply "One cubic inch every thousand years" which suddenly reduced to only a sixteenth as much when we went metric, to "one cubic centimetre per thousand years"! And both answers are about as good as each other. We all know speleothems which haven't been actively growing for thousands of years, and there is the golden syrup tin in Orient Cave, Jenolan, which must have accumulated at least two cubic centimetres per year since it was placed there, and the protective netting in numerous caves which often has quite large amounts of calcite deposited on it. So rates and ages must vary tremendously. And unfortunately stalagmites and stalactites, no matter how much some of them may look like birthday cakes and candles, don't have a label saying I'm x years old. So how can we answer these questions? It's not all guess work, though nothing is as simple as counting the candles.

There are several methods that can be used and they all have the same basic concepts:

1. There is something we can measure that changes with time - let's call it x, in the time-honoured scientific jargon. X will be something different for each different method, and we'll discuss what x is later for each method.
2. We know or can make a good guess at x's initial value
3. We can measure x now
4. We know the rate at which x changes with time

These concepts may seem very simple but we need to be aware of them because sometimes one or more of them mightn't be true and that means our estimates of age might be wildly wrong. Now let's look at the various methods and keep an eye out for possible pitfalls at the same time.

A) Carbon dating

1. X, which we measure, is the ratio of two different carbon isotopes. An isotope is a different version of the same element: the two isotopes of carbon have the same chemical properties and behave in just the same way but one of the isotopes, C14 [this is chemical shorthand again for an isotope of the element carbon with an atomic weight of 14. Look up isotope for yourself in your dictionary!], is slightly heavier than the other, C12 (approximately 14 times as heavy as a hydrogen atom instead of 12 times). However, as time goes on the heavier isotope changes: it decays and turns into C12, so the ratio of C14 to C12 also changes with time.
2. We assume that we know what the initial ratio is: we can get a good idea from measuring the ration in very recently formed speleothems, such as deposits on wires and other debris left in caves which we know can't be very old.
3. We measure the ratio of C14 to C12 by mass spectroscopy, which needs about 20 gm, or nuclear accelerator methods which need only a few milligrams but are more expensive.
4. The rate at which C14 changes to C12 is known from experiments: half of whatever C14 is there will have changed to C12 in 5730 years. (This is called the "half life" of C14. This means that after 11460 years there'll be only a quarter as much, after 17190 there'll be an eighth and so on. After about 35,000 years there is so little of the original C14 left that the changes become too small to measure accurately, so 35,000 years is about as old as we can estimate using carbon dating. (If you see carbon dates quoted as 35,000 it almost always means "at least 35,000 and could be much older".)

B) Uranium series dating

1. Here we measure the ratio of uranium-234 (U234 - which is about 234 times as heavy as a hydrogen atom) to thorium-230 (Th230 - 230 times as heavy as a hydrogen atom). U234 changes to Th230 as time goes on, half of it every 245,000 years. This means that the upper limit for uranium dating is much higher than it is for carbon dating, around 350,000 years.
2. There are small quantities of uranium and thorium in almost all rocks and soils, although it may be only 1 part per million (less than 0.0001 %). As the water seeps down it picks up both uranium and thorium but, because the thorium is much "stickier" it usually gets left behind before it gets as far as the cave. We assume that only uranium is deposited in the speleothems and no thorium, so any thorium we measure must have came from uranium decaying to thorium.
3. We measure uranium and thorium by separating them out chemically and then monitoring the amount of radioactive decay.
4. The rate of decay (half-life) of uranium is known from experiments, as it is for carbon (remember that any uranium series age of 350,000 years means at least 350,000 years and could be lots older).

C) Electron spin resonance (ESR) dating.

This technique is "hot off the press" and still being refined. But it's really exciting because it can potentially date back more than 1 million years, well beyond the limits of radiocarbon and uranium dating. ESR dating is very similar to thermoluminescent (TL) dating which is used for dating sediments, volcanic rocks and some archaeological artifacts. However, TL dating hasn't proved very successful for dating speleothems. Both methods actually measure similar things, although they use different techniques.

1. Both ESR and TL measure the number of imperfect or broken chemical bonds in a sample. You will remember from the last ANDYSEZ that calcium carbonate (calcite or aragonite) is made up from different atoms "holding hands" or bonded together to form compounds and lattices. As time goes on some of the atoms "let go hands" - more and more of the bonds get damaged. This damage is caused by the very low level background radiation that is everywhere, even in caves, although in caves it's usually less than it is above ground.
2. We assume that initially all the bonds are perfect, This is a reasonable assumption because whenever we measure a recent sample we don't find any broken or damaged bonds.
3. Here's the essential difference between ESR and TL dating:
   1. In ESR dating the sample is ground up and placed in an ESR spectrometer. This instrument produces a magnetic field which is absorbed by the damaged bonds. The more damage, the more power is absorbed, and this is what we measure.
   2. In TL dating the prepared sample is heated in a lightproof measuring chamber. As the sample heats, the sample acquires sufficient energy to repair themselves and, as each one does so, it gives out a flash of light which is recorded - the more damaged bonds, the more flashes of light.
4. The rate at which damage increases after the stalagmite is formed is trickier to measure. We need to know the amount of background radiation because this is what causes the damage. We can measure this in several ways, by chemical analysis, by gamma spectrometry and by special dose meters. We also need to know how much damage is caused by how much radiation, because not all samples are equally sensitive. We do this by carrying out experiments in which we blast small parts of ground-up sample with high doses of artificial radiation to see how it reacts.

So far, these three methods, radiocarbon, uranium and ESR dating, are the best we have available. They all have their place. Carbon dating is best for young samples, uranium series is useful for up to 350,000 years and ESR is the only technique for samples older than 350,000 years, which is the upper limit for uranium dating. The techniques not only have different age ranges for which they are best but they also overlap in the age ranges to which they can be applied. This is very useful at it gives us a way of independently checking the age estimates - and they don't always agree perfectly! Sometimes there may be some undetected difficulty in the analysis, sometimes the assumptions we make may not be valid for a particular sample, or maybe there's just more to learn. Perhaps we should remember what Poul Anderson, scientist and science fiction writer said: I have yet to see any problem, no matter how complicated, which, when you looked at it the right way, did not become still more complicated.