and it is this office that the finest and highest of the Krakatoa dust is thought to have performed. And here a peculiar cycle of operations, first suggested by Wollaston years ago, and generally neglected since then, may be reconsidered. So long as the water-substance is in the vaporous condition, it acts as a gas, and tends to expand upward; part of it would thereby be condensed, generally in the solid state, and on losing the gaseous condition the frozen particles would at once tend to fall toward the earth, impeded only by the presence of the thin air; but, after a certain length of falling, they would reach air warm and dry enough to allow them to re-evaporate, whereupon their vapor would again expand upward, and the cycle of operations begins anew. Wollaston suggested that the gases of the air might be thus affected by the extreme cold of upper space, and that a limit of the atmosphere might so be determined. There is, however, no experimental evidence yet adduced to prove that oxygen and nitrogen would behave in such a way, and the limitation of the atmosphere must be due to other causes; but the upward extension of water-vapor might be thus controlled. May we not, therefore, imagine that the vapor of the upper atmosphere, re-enforced liberally by steam from Krakatoa and other volcanoes in eruption at the same time, found its opportunity for condensation much improved for several months by the lava-dust from the same sources; and thus explain the brilliant sunsets and the strength of color in Bishop's ring during the winter of 1883-'84? But gradually the dust settles down, very slowly on account of its large ratio of surface to weight; and the vapor also decreases by slow downward diffusion; then the brilliancy of the display is lost, and the moderate residual of vapor, condensing as well as it can alone, produces only a fainter-colored ring and sunset glows that are visible only under especially favorable circumstances.
Be all this as it may, it is well to bear in mind that some such explanation must be found and accepted, for the facts of diffracting particles and their relation to Krakatoa are too well proved to be doubted, unless evidence not yet forthcoming shall appear in great strength.
The ring is doomed to disappear, and hence deserves a close watching. For, as Forel has pointed out, the outburst of Krakatoa must have had its rivals in ancient if not in modern times, and rings like Bishop's must in all probability have resulted from former dusty explosions. But these had all faded long before Bishop's ring appeared, and we must, therefore, conclude that it will fade away also. It should be carefully watched, especially from high-level stations, and those who make a persevering record of it should not fail to inform Professor Kiessling, of Hamburg, about what they see.
