The earth-throe on that occasion was felt in Peru about five minutes past five on the evening of August 13. Twelve hours later, or shortly before midnight, August 13, Sandwich Island time (corresponding to 5 p.m., August 14, Peruvian time), the sea round the group of the Sandwich Isles rose in a surprising manner, “insomuch that many thought the islands were sinking, and would shortly subside altogether beneath the waves. Some of the smaller islands were for a time completely submerged. Before long, however, the sea fell again, and as it did so the observers found it impossible to resist the impression that the islands were rising bodily out of the water. For no less than three days this strange oscillation of the sea continued to be experienced, the most remarkable ebbs and floods being noticed at Honolulu, on the island of Woahoo.” The distance between Honolulu and Arica is about 6300 statute miles; so that, if the wave travelled directly from the shores of Peru to the Sandwich Isles, it must have advanced at an average rate of about 525 miles an hour (about 450 knots an hour). This is nearly half the rate at which the earth’s surface near the equator is carried round by the earth’s rotation, or is about the rate at which parts in latitude 62 or 63 degrees north are carried round by rotation; so that the motion of the great wave in 1868 was fairly comparable with one of the movements which we are accustomed to regard as cosmical. I shall presently have something more to say on this point. Now in May, 1876, as we have seen, the wave reached Hawaii at about a quarter to five in the morning, corresponding to about ten, Peruvian time. Since, then, the earthquake was felt in Peru at half-past eight on the previous evening, it follows that the wave, if it travelled directly from Peru, must have taken about 13½ hours—or an hour and a half longer, in travelling from Peru to the Sandwich Isles, than it took in August, 1868. This is unlikely, because ocean-waves travel nearly at the same rate in the same parts of the ocean, whatever their dimensions, so only that they are large. We have, then, in the difference of time occupied by the wave in May, 1876, and in August, 1868, in reaching Hawaii, some confirmation of the result to which we were led by the arrival of the wave simultaneously at all the islands of the Sandwich group—the inference, namely, that the observed wave had reached these islands after reflection from the Californian shore-line. As the hour when the direct wave probably reached Hawaii was about a quarter past three in the morning, when not only was it night-time but also a time when few would be awake to notice the rise and fall of the sea, it seems not at all improbable that the direct wave escaped notice, and that the wave actually observed was the reflected wave from California. The direction, also, in which the oscillation was first observed corresponds well with this explanation.