The determination of deep‐seated geologic structure from the study of earthquakes has much in common with the determinations made of more shallow structure by the seismographic refraction‐method of prospecting. But the earthquake solution presents more difficulties. (1) We do not know just where earthquakes will originate. Therefore, we must build seismographic stations scattered throughout the domain to be investigated. These stations are expensive to establish and maintain, and, since the results of the investigations are not of any obvious economic value, the funds to support the stations are difficult to obtain. We cannot, therefore, have stations distributed every few miles from our sources. (2) We do not know when an earthquake will originate. Therefore, the stations must be kept in continuous operation with the resultant cost of maintenance and supplies. We cannot operate our recording drums at very high speeds (with resultant high accuracy in times measured), because of the expense of photographic paper for continuous operation.