The theoretical scientist makes his discoveries within the pursuit of questions concerning explanation. Given the results of observation and experiment, he looks for laws combining the observational data into a consistent causal net. The more general and comprehensive the laws, the greater their explanatory power. However, in order to be accepted the law must not merely combine known data by way of a single rule; it is required to predict new data accessible to observational test. The criterion of scientific truth is the verification of predictions; and explanatory power is thus reducible to predictive power.