LVIII. The effects of finite size and decentring of the source in an angular correlation experiment

Summary In a first approximation any extended source behaves like a point source situated at the ‘centre of gravity’ of the given activity distribution. The first-order errors caused by bad centring of a point source are rather large and exhibit a strong angular dependence. They can be minimized by a randomized block design when it is impossible to avoid them altogether owing to technical reasons.