As to the actual cause to which both phenomena are to be ascribed, we must, I think, exculpate Krakatoa from all part or share in producing these strange effects. The appearance of a blue sun at Trinidad, followed two or three days later by a green sun in the East Indies, can not possibly be associated with the occurrence of an earthquake on the Javan shore a few days earlier. Beside, it must be remembered that we should have to explain two incongruous circumstances; first, how the exceedingly fine matter ejected from Krakatoa could have so quickly reached the enormous height at which the matter actually producing the afterglow certainly was; and, secondly, how having been able to traverse still air so readily one way, that matter failed to return as readily earthward under the attraction of gravity. Again the explanation, which at first seems a most probable one, that unusually high strata of moist air, with accompanying multitudes of ice particles, caused the phenomena alike of absorption and of reflection, seems negatived—first, by the entire absence of any other evidence of extraordinary meteorological conditions in September, October and November last; and, secondly, by the entire absence of any of the optical phenomena which necessarily accompany the transmission of sunlight through strata of air strewn with many ice particles. We seem obliged then to adopt a theory, first advanced, I believe, by Mr. A. C. Ranyard, that the phenomena were caused by a cloud of meteoric dust encountered by the earth, and received into the upper regions of the air, thence to penetrate slowly (mayhap not till many months have passed) to the surface of the earth. Mr. Ranyard calls attention to the circumstance that probably the early snows of the winter 1883-’84 would bring down the advanced guard of such meteoric dust; and even as I write I learn that Mr. W. Mathieu Williams has followed the suggestion. He carefully collected the snow which fell in his garden, eighty yards from his chimneys and half a mile from any to windward. Slicing off a top film of the snow with a piece of glass he thawed it, and found a sediment of fine brownish-black powder. Ferrocyanide of potassium added to the snow-water produced no change of color, showing the absence of iron in solution, nor was there any visible reaction on the black dust till he added some hydrochloric acid. Then the blue compound indicating iron was abundantly formed all round the granules, and presently, as their solution was effected, a bluish-green deposit was formed, and the whole liquid deeply tinged with the same color. “It was not,” says Mr. Williams, “the true Prussian-blue reaction of iron alone, but just the color that would be produced by mixing small quantities of the cyanide of nickel (yellowish green) and the cyanide of cobalt (brownish white) with a preponderating amount of Prussian blue.”