maybe we can modify dhruv's solution a little bit. First, we do as
dhruv said, measure the volume of the whole pipe and compare it with
the theoretical value. If it is wrong, we know that it is not even all
the way. If it is correct, we can now fill up the pipe to half its
length and then compare that value to half of the theoretical volume of
the pipe. Again we apply the test: if it is wrong, we know it is not
even all the way. If it is correct, then we test 1/4 of the length of
the pipe. We can do so until halving the measured length negligible.
This process can also be modfied a little to find out where the
inconsistencies occur along the length of the pipe.