absolute, but relative to the preceding state of the weather. But also these indications are valid for only a short interval before the actual advent of the storm; and in some instances, as in the Hyperborean storm of the 2d and 3d of October, 1860, the interval is too short for any advantage to be taken of the notice. The particulars of this storm, which present in true character the difficulties which the meteorologist must encounter, are too interesting to be omitted, and we shortly recount them from the complete and admirably-conducted investigation published by Prof. C. Piazzi Smyth, in the "Annals of Scottish Meteorology for 1856 to 1871." The term Hyperborean has been employed to prevent confusion with tropical hurricanes; it has also been called, from its essential locality, the Edinburgh storm. We have to consider only the practical lessons to be deduced from the observations of this storm; the account of the actual observations must be read from the before-mentioned report of Prof. Piazzi Smyth. First, then, the barometric notice was insufficient and too local to be of service, while the storm was too quick in its movements. St. Hilda is the most westerly station; and, even if the storm could have been telegraphed thence, the message would have allowed only two hours for preparation, and would have arrived while the eastern men were sound asleep. If a message could have been sent from Iceland the day previous to the arrival of the storm, many wrecks would have been prevented. So that we see the present system of meteorology necessitates not only diligent but earnest watching of the signals that should be afforded by a net-work of cables and overland wires, for it is by a series of connected observations, extended over a large area, that the usefulness of this branch of meteorology is alone likely to be advanced.
But, it may be asked, what definitive knowledge can be gained, say not of storms, but of average weather for some future period? Here we must again refer to Prof. Piazzi Smyth's report on the rock-thermometers at the Royal Observatory, Edinburgh, and to the Proceedings of our own Royal Society for the 2d of March, 1870, in which predictions of the weather during the winters of 1871–'72 are attempted. The rock-thermometers have by their readings shown some well-marked supra-annual cycles, the relation of which to the sun-spot cycles will be known to our readers. And on this point it may be stated that the Radcliffe astronomer announces, in his report for 1871, that the mean azimuthal direction of the wind at Oxford, rigorously computed from automatic records during the last eight years, varies year by year through a range of 58° on the whole, between maximum and minimum of visible sun-spots, the tendency of the wind to a westward direction increasing with the number of spots, and with such west wind, it is to be presumed, the amount of rain also. "The most striking and positive feature of the whole series of observations," continues Prof. Piazzi Smyth, "is the great heat-wave which occurs every eleven years and a fraction, and nearly coincidently with the beginning of
