miles per minute. This sudden eruption will be almost immediately succeeded by great expansion and consequent fall of temperature, so that within a few hours the heavy volatile metals begin to condense and rapidly recede into their crater, and the faculæ in front and at the sides will now stream inward to occupy this vacuum with constantly accelerated velocity, pouring over the edges like the rush of waters at the Falls of Niagara. As they sweep downward over the inner rim of the funnel, these streams of faculæ will glow with increased whiteness, and appear to be sharply cut off at their inner ends; but this is only apparently so, and is due to the position of the observer, who looks almost directly downward upon these descending streams. It is for the same reason that the faculæ appear more brilliant when near the borders of the solar disk (see page 109). Any good view of a sun-spot when analyzed will show the streams of faculæ thus pouring inward, and they are among the most peculiar and conspicuous phenomena to be observed. The drawings of Professor Langley, reproduced in the Popular Science Monthly for September, 1874, and July, 1885, are particularly striking in their illustration of these effects, though their significance and interpretation were not then at hand. But while these heavy metallic vapors so rapidly condense and subside in the forward or initial portion of the sun-spot under observation, new depths of intensely-heated faculæ are generated behind, and these operate with renewed energy upon the fresh surface of the solar core in rear of the original seat of eruption; so that each sun-spot, while in an active state, will exhibit two entirely distinct aspects, the forward portion of the crater in a state of rapid condensation and subsidence of the recently erupted metallic vapors, and with inflowing streams of incandescent hydrogen from the front and sides, and the rear portion of the crater up to its rearward wall, and even streaming forth from beneath it, in a state of violent eruption. The large volcanic craters of the Hawaiian Islands exhibit similar partial eruptions and subsidences progressing simultaneously in the same depths. The sudden formation of the great incandescent loops and plumes to which Professor Langley calls especial attention, and which have hitherto been so perplexing, can now be readily understood and explained. If one of these inflowing streams be carried partially down into and across the crater, and then caught, in its advance, by the uprush in the central or rear portions of the cavity, it will be at once swept upward alongside the ascending eruption, and either scattered at its forward extremity into sprays and plumes, or else thrown forward bodily in the form of a more or less complete loop. In a sun-spot fifty thousand miles in diameter, such a loop, having a long diameter of twenty thousand miles, if we give a speed to the faculæ of seven thousand miles per minute, would be formed in about