The procedure, however, is difficult, and many tests have shown that it can be inaccurate, and it is at times not even considered reliable by mainstream scientists.
Suppose we have a tank partly filled with water, and a hole in the bottom through which the water is leaking out of the tank.
We wonder how long the hole has been there, that is, how old the hole is.
Sometimes a hypothesis must be made that may be plausible but has not been proven.
At other times an additional measurement can eliminate the need for one assumption, although no science can be done without assumptions at some level.
This method has particular significance because it is the only method that purports to give a value for the age of the Earth (that is, the time at which the Earth and asteroids condensed out of the planetary nebula) rather than only a lower limit on the age of the Earth (that is, the age of the oldest surviving rocks on the Earth).
There are a number of assumptions involved, but if they all hold, when the ratio of Pb for the various samples, the points should lie on a straight line.
We could measure (a) how much water the tank holds, (b) how much is still in the tank, and (c) the rate at which it is leaking out.
We can calculate the age of the hole by subtracting (b) from (a) to find out how much water has left the tank, and then dividing this by (c), the rate at which it is leaking out.
If they disagree, it may be because lead has been lost at some point in the history of the sample, for example, if there was an episode of heating above 1000 degrees C.