beautiful and rare, and the more striking from the absence, to the observer, of the sun or clouds, and the complete insulation of the phenomenon, yet by close observation upon other evenings, it was found that partial effects of the same kind were very common, and, from the manner in which these could be observed, the explanation above given was fully confirmed. On several evenings after, when observing the sun-set from the neighbouring hill of St. Catherine, it was found that if the atmosphere was generally clear, but with compact and distinctly-formed clouds floating in it, the effect was produced. The usual appearance of rays at sun-set, diverging amongst the clouds in the west, from the sun, is well known; but even when these were not visible, upon looking to the eastern half of the hemisphere, and especially to the north or south of east towards the horizon, it was rare that some clouds could not be distinguished with long shadowy projections behind them, always converging to the spot opposite to the sun. Frequently clouds could be selected moving more immediately in the neighbourhood of the observer: of those which passed overhead, the shadows could not be observed close to the clouds; but carrying the eye onwards towards the east, the same shadows became visible, when considerably fore-shortened, and could be observed moving on and changing with the clouds themselves. All these phenomena, with their variations, were easily referable to their causes, and may be observed at almost any sun-set in line weather; but the effect of the first evening, so similar in kind, though so different in appearance, was not again remarked. It is with a view of guarding persons who may observe the same effect, against any mistake as to its origin, that the appearance, with its nature, has been thus particularly described.
On the confinement of Dry Gases over Mercury[1].
THE results of an experiment made by myself, and quoted as such, having been deemed of sufficient interest to be doubted, I have been induced to repeat it; and though the original experiment was not published by me, I am inclined to put the latter and more careful one upon record, because of the strong
- ↑ Quarterly Journal of Science, xxii. 220.
