CIRCUMSTANTIAL EVIDENCE is a kind of indirect evidence. Suppose something happens in the presence of witnesses who observe it, then their evidence is direct evidence of the oc currence. But now suppose there is no such direct evidence of the event, then we have to rely on indirect evidence, if any. Such in direct evidence usually consists of certain circumstances surround ing the event or in some way connected with it. By piecing these circumstances together we may get an intelligible, coherent ac count of the whole course of the event or events in question. This piecing together of the data is a kind of hypothesis, and there may be several rival hypotheses to account for the facts. But as in all cases of the use of hypotheses, the hypothesis which gives the most adequate and consistent explanation is usually accepted. In ference from circumstantial evidence resembles ordinary induc tion to the extent that it involves the use of hypotheses and their verification, but it is unlike ordinary induction inasmuch as it does not aim at a generalization, being concerned only with some par ticular event. The systematic character of inference is perhaps most obvious in the case of inference from circumstantial evi dence, for it manifests most clearly the process of the imaginative construction of a coherent system out of fragmentary evidence. The most familiar occasions for the exercise of inference from cir cumstantial evidence are criminal cases. Criminals naturally take precautions against the possibility of direct evidence, and are us ually betrayed by circumstantial evidence. See EVIDENCE. See A. Wolf, Essentials of Logic (1926).