Do you need to know how old a bone is perceived to be before you can date it properly. If so how do you know how old it is before you date. If your answer is index fossils how do you know how old they are before you date them?

No, you don't need to know how old a sample is before it can be dated properly.

There are a number of different radiometric dating methods. Each of them has its own optimal range, because of the different half-lives of the isotopes used. Two extreme examples: there will be no detectable C-14 in a carbon sample that's more than 75,000 years old -- so we don't date Carboniferous Period coal deposits using C-14! On the other hand, it would be impossible to date a 75,000-year-old sample using uranium-lead dating because you would be unable to detect a significant difference in lead vs. uranium content in such a sample.

I expect that if you really had no idea how old a sample is, you would use several different dating methods to establish a rough age. Then you would date again using at least two methods that work well for samples of that age.

A careful worker will always use more than one method of getting a number, if more than one method is available. If the two numbers agree, you have more confidence in your measurement.

Index fossils are index fossils because they are always found in rocks that date a certain age (or rather range of ages). It's not something we pull out of our hats.

Fossils are not often directly dated. First, they're too valuable to destroy. But more importantly, the mineralization process is similar to the formation of sedimentary rocks--it precludes accurate dating.

Instead, we normally date rocks around the fossil. The major exception is carbon-14 dating. First, it operates on a timeline too short for rocks to form. Second, fossil mineralization won't happen (under most conditions) in 75,000 years, the maximum age for C-14.

Incidentally, geologists understood that the earth had to be at least several hundred million to a billion years old, a century or more before we even knew about radioactivity. They knew this by noticing that rock formations are like layer cakes -- and that rocks are not formed quickly enough (except by volcanoes) for the process to be noticeable even in several human lifetimes. Radiometric dating simply allowed geologists to put hard numbers to their reasonable estimates.

A last note and reminder: radiometric dates can be pretty accurate -- as close as a part per thousand or less, though most of them are only accurate to a few percent. But the importance is not so much the range of numbers that a radiometric date gives, but the much larger range of numbers that it excludes. A sample may be "roughly" dated to 300 million years, give or take 100,000 -- but that means that we know it's not 5,000 years old!

Some references:

• The USGS article on Radiometric Dating, from their website on Geologic Time.
• A very thorough article on Radiometric Dating, from the journal of the American Scientific Affiliation. This article also includes a discussion of non-radiometric dating methods.
• The Wikipedia article is thorough, but may be a bit confusing to a novice.

Dan Berger