In many of the sciences there is no means of ensuring absolute accuracy in the measurement of a particular quantity. The difficulty can to some extent be overcome by making several independent sets of observations and by combining these to give the most probable result. This done, the degree of precision of the calculated value can in its turn be assessed. It is seldom a straightforward matter of taking the average reading, and in many cases a knowledge of the theory behind the methods of combining observations becomes essential. It is this knowledge that Professor Smart provides.