Corrigendum

Analytical steps towards a numerical calculation of the ruin probability for a finite period when the risk process is of the Poisson type or of the more general type studied by Sparre Andersen, Astin Bulletin vol. 6, 54-65, by OLOF THORIN, Stockholm. On pp. 56-57 of the above-mentioned paper there is a digression about K(t), the distribution function for the time between successive claims. Some examples are given and something is said about their generality. Here a wrong statement has slipped in, which ought to be corrected. In fact, it is said that the class of probability laws given in the example e. is dense in the class of all probability laws concentrated on the positive half-axis. However, it is not difficult to see that the example e. is not sufficiently general for this conclusion. Therefore, please read the statement in view (the midst of p. 57 or more precisely the tenth through the twelfth rows from above) as referring not to the example e. but to the following example e'.