track, while on the other theories the probabilities ranged themselves against this fundamental requirement. Every missile projected from a mighty primeval volcano on this earth with the requisite speed will be freely revolving around the sun in an elliptic track, and frequently crossing the earth's path.
We have learned from Tschermak the fundamental fact that in all probability meteorites owe their origin to some volcanic source. Adopting this as a fundamental position, I have given the astronomical grounds which appear to demonstrate conclusively that these volcanoes must have been on our own earth. I am, however, at once met by the objection, that the mineral substances found in meteorites are not those which mineralogists generally speaking recognise as the products of terrestrial volcanoes at the present day. I fully admit this, indeed it is well known that in many cases a meteorite can be shown by chemical analysis alone to be one of those bodies which has fallen from the sky. Not only the actual minerals which are present, but the form in which they are associated, are generally speaking so characteristic that the meteoritic character of a stone can often be unhesitatingly pronounced upon even though the fact of its fall is entirely unknown. This is a very great difficulty, and I should be inclined to regard it as an insuperable one were it not for a particular circumstance on which I must now dwell. No doubt it is quite true that the character of a mineral mass will, generally speaking, enable the experienced mineralogist to pronounce decisively whether the object is meteoritic or whether it is not. There is, however, one notable instance in which the ordinary rules for the diagnosis of a meteorite would have certainly proved
