For a detailed study of the earthquakes of any district, an interval of twenty-one years is too brief. Long-period variations of frequency cannot be established. We can form no satisfactory conception of the distribution of seismic energy in space, for some foci may lie inactive for a much longer time, while others may continue in operation without apparent change. But, to ascertain the characteristic features of the earthquakes, to investigate their relations with those which precede and follow them, or to trace their connexion with the structure of the central districts, such an interval is possibly of sufficient length. In any case, the defects resulting from its brevity may be partly compensated by uniformity in treatment and in the methods of investigation.