wave had a height of nearly thirty feet from trough to crest. We read, in accordance with this explanation, that “the regurgitations of the sea were violent and complex, and continued through the day.” The wave, regarded as a whole, seems to have reached all the islands at the same time. Since this has not been contradicted by later accounts, we are compelled to conclude that the wave reached the group with its front parallel to the length of the group, so that it must have come (arriving as it did from the side towards which Hilo lies) from the north-east It was, then, not the direct wave from Peru, but the wave reflected from the shores of California, which produced the most marked effects. We can understand well, this being so, that the regurgitations of the sea were complex. Any one who has watched the inflow of waves on a beach so lying within an angle of the line that while one set of waves comes straight in from the sea, another thwart set comes from the shore forming the other side of the angle, will understand how such waves differ from a set of ordinary rollers. The crests of the two sets form a sort of network, ever changing as each set rolls on; and considering any one of the four-cornered meshes of this wave-net, the observer will notice that while the middle of the raised sides rises little above the surrounding level, because here the crests of one set cross the troughs of the other, the corners of each quadrangle are higher than they would be in either set taken separately, while the middle of the four-cornered space is correspondingly depressed. The reason is that at the corners of the wave-net crests join with crests to raise the water surface, while in the middle of the net (not the middle of the sides, but the middle of the space enclosed by the four sides) trough joins with trough to depress the water surface.[24] We must take into account the circumstance that the wave which reached Hawaii in May, 1876, was probably reflected from the Californian coast, when we endeavour to determine the rate at which the sea disturbance was propagated across the Atlantic. The direct wave would have come sooner, and may have escaped notice because arriving in the night-time, as it would necessarily have done if a wave which travelled to California, and thence, after reflection, to the Sandwich group arrived there at a quarter before five in the morning following the Peruvian earthquake. We shall be better able to form an opinion on this point after considering what happened in August, 1868.