The classical theory of errors can be divided into stochastic and determinate parts, or branches. The birth of the first of therse became inevitable after Bradley's idea of cultivating astronomy and natural science in general by “regular series of observations and experiments” became universally accepted. Such scholars as Lambert, Simpson, Lagrange, Daniel Bernoulli and Euler were responsible for the development of the stochastic theory of errors while Laplace and Gauss completed its construction. About fifty or sixty years ago it was (...) included into mathematical statistics. (shrink)

The use of such concepts as randomness and probability is discussed and stochastic decisions occurring in the sacred texts are studied. In connection with probability, the measurements made in antiquity in order to comply with religious demands are also touched on. The cases where cause rather than randomness was recognized to be at work are indicated, and the statements contained in the Talmud and the thoughts of its later commentator, Maimonides, on hypotheses are linked with the appropriate opinions of Isaac (...) Newton and Jakob Bernoulli. (shrink)

I dwell on the appearance, in the mid-19th century, of a formula describing random sampling without replacement under incomplete knowledge and discuss the statistical aspect of this formula and its applications. I also provide illustrations showing that the fairness of sampling without replacement was being questioned.