This chapter explains what the demand to deal with errors of measurement amounts to and at the same time make some remarks about the origin and nature of errors of measurement. It shows how well Philip Kitcher’s theory of explanation is able to address the problem of errors. Measurements, especially in physical science, are normally recorded together with an estimate of the likely error involved when the reading was taken. Explanations in physical science normally aim to provide people with understanding of why quantities are exhibited with certain values, even when these are encased within imprecision intervals. The kind of understanding which is increased when it is shown that derivations involving laws and observation sentences from the two domains have the same form evidently has nothing to do with understanding the phenomena. The example bears a resemblance to a formal analogy between two theoretical principles.