For each of the eight periods of the sun-spot cycle, the percentage of years of seismic and volcanic activity has been calculated from the data of Fig. 1, as shown in the table at the end of this article, and has been plotted in the frequency curves of Figs. 10 and 11. For Sayles, 13 complete sunspot cycles are available, and for Jensen, 10. In the same way the average intensity of the phenomena of the years of each period has been calculated and plotted in Figs. 12 and 13. In order to obtain the mean curve it is necessary to combine the six shown in Figs. 10 to 13 into one. The first process is the combination of Jensen's four curves into two which shall be comparable to the two of Sayles. This is done in Figs. 11 and 15. In Fig. 14 the solid line reproduces Sayles's line of Fig. 10 directly, while the dotted line gives the mean between Jensen's two curves of Fig. 11. Thus we have two mean seismo-volcanic frequency curves. In the same way, in Fig. 15 the solid line reproduces Fig. 12, and the dotted line gives the mean of the two lines of Fig. 13; and we have two mean seismo-volcanic intensity curves. In these curves, as in the others, percentages are used, so that when diverse phenomena such as frequency and intensity come to be compared and averages computed, each receives the same weight. In Fig. 16 the frequency curves of Sayles and Jensen shown in Fig. 14 are combined into the solid line and the intensity curves of Fig. 15 into the dotted line. In order to make the two curves comparable the maximum in each case has been reckoned as a hundred. As a result of the combination of the data of our two authorities, the personal equation is largely eliminated. It will be noticed that the curves of Fig. 16 are much smoother than those of preceding figures. This is especially true of the frequency curve where there is least liability to errors of judgment. Finally, in Fig. 17 the two curves of Fig. 16 are combined into one, shown by the solid line. This represents the net result of Sayles's data as to the combined seismic and volcanic frequency and intensity of 13 complete sun-spot cycles, and of Jensen's independent data as to the uncombined seismic and volcanic frequency and intensity of 10 complete sun-spot cycles. The dotted line in Fig. 17 is the mean sun-spot curve derived from Fig. 1, and calculated and plotted in precisely the same manner as the six curves of Figs. 10 to 13 from which the mean seismo-volcanic curve is derived. For the sake of convenience in comparison, the sun-spot curve has been plotted with the minimum at the top.
The resemblance between the mean sun-spot and mean seismo-volcanic curves is extraordinary. The maximum of the one occurs at the same time as the minimum of the other, and in both cases there is a steady progress from maximum to minimum and back. If our terrestrial data of earthquakes and volcanoes were as complete as our