and the coincidence of this result with the uniform position of auroral arches, naturally suggests the idea of a real connection between the two phenomena."

607. The aurora is of great extent, having been sometimes observed simultaneously in Europe and America. From observations made at Hobart Town, and in the United States of America, Professor Loomis thinks it probable that an exhibition of auroral light about one magnetic pole of the earth is uniformly attended by a simultaneous exhibition of auroral light about the opposite magnetic pole. The height varies from about 45 to 500 miles above the earth. From observations made on one which appeared in England during March 1826, Dalton calculated its height at 100 miles. Sir John Herschel determined the height of one which was seen on the 9th March 1861 by himself in Kent, and at the same time by Mr Lowe at Nottingham, to be 83 miles. Of auroras which have been seen near the earth, the one thus described by Captain Parry is the most remarkable :-" While Lieutenants Sherer, Ross, and myself were admiring the extreme beauty of the northern lights, we all simultaneously uttered an exclamation of surprise at seeing a bright ray of aurora shoot suddenly downward from the general mass of light, and between us and the land, which was there distant only 3000 yards. I have no doubt that the ray of light actually passed within that distance of us."

608. As regards frequency, auroras have a daily period which reaches the maximum about midnight. The aurora is not often seen in summer, partly, no doubt, on account of the short nights and the clear skies. There is, however, a double maximum and minimum occurrence in the year, which the following table from Kaemtz and Loomis, giving the number of auroras seen in each month in Europe and America respectively during many years, clearly establishes :

In addition to this annual period, there would appear to be a secular period comprising a number of years. Observations appear to indicate a maximum every ten years, and a still larger maximum period recurring every sixty years.

609. The culminating point of the auroral arch being at or near the magnetic meridian, and the centre of the corona in the line of the dipping needle produced, point out an evident connection between the aurora and terrestrial magnetism. The magnetic needle is also much agitated when the aurora is visible. When the arch is motionless, so is the needle; but as soon as streamers are shot out, its declination changes every moment, and this happens though the aurora does not appear at the place of observation, but is seen near the pole. According to Hansteen, the intensity increases greatly a short time before the appearance of the aurora; but as soon as the aurora begins to be seen, it diminishes in proportion to the brilliancy of the display; and it then returns slowly, generally in twenty-four hours, to its original value. During 1857 and 1858, Captain M'Clintock, when in the arctic regions, observed that the aurora in all cases appeared to come from the surface of open water, and not in any case from the fields of ice,—an observation favouring the idea that it is caused by electrical discharges between the earth and the air, and that these are interrupted by the fields of non-conducting ice.

610. General Sabine has discovered that magnetic disturbances of the earth are due to the sun, but not to his heat and light; and are invariably accompanied by the aurora and by electric currents in the surface of the earth. Dr Balfour Stewart considers that auroras and earth-currents are to be

regarded as secondary currents due to small but rapid changes in the earth's magnetism; and that the body of the earth may be likened to the magnetic core of a Ruhmkorff's machine, the lower strata of the air forming an insulator, while the upper and rarer, and therefore electrically-conducting strata, may be likened to the secondary coil; and the sun perhaps likened to the primary current which produces changes in the magnetic state of the core. If this be so, he adds that the energy of the aurora may come from the sun; but this may be considered doubtful, from our ignorance of the way in which the sun affects terrestrial magnetism. The secular periods of the sun's spots, of the variation of the magnetic needle, and of the frequency of auroras, seem to indicate that these phenomena are regulated not by terrestrial but by astronomical causes.

611. R. P. Secchi, Director of the Observatory of the College, Rome, and M. Marié Davy, Chief of the Meteorological Division in the Imperial Observatory, Paris, have for some time given particular attention to magnetic storms, electric disturbances, and auroras in their relations to the weather and the prediction of storms. Marié Davy states* that the perturbations of the magnetic needle are joined inseparably with one or more of the three following phenomena :—1. General disturbances of the telegraphic lines due to widespread auroras, which indicate general movements of the atmosphere in high latitudes and over the Atlantic. 2. Disturbing currents of a more local character occurring over the telegraphic lines some time before the storm appears to which they owe their origin, thus lengthening the distance and time at which the approach of the storm may be perceived. 3. Disturbing currents still more restricted accompanying the electric changes which occur when the storm itself is passing.

612. The cirrus cloud, which is composed of very minute crystals, sometimes appears of a texture so delicate as to elude the eyes of all but the most practised observers; and no doubt it occasionally spreads a screen of microscopic crystals in the

* 'Da la Prévision du Temps,' p. 493. Paris, 1866.

upper regions of the atmosphere so thin, that the eye cannot detect it till it is revealed by electric discharges passing through it. This cloud probably originates in the equatorial current beginning to prevail in the higher regions of the atmosphere, and depositing the watery vapour necessary for the electric discharges, the faint light of which, reflected by the minute crystals of the cirrus cloud, forms the aurora. For the elucidation of the important questions here raised more magnetic observatories are required, so that synchronous magnetic charts might be made for comparison with similar meteorological charts. If this were done, and the relations among these atmospheric elements discovered, the magnetic and electric states of the atmosphere and the aurora might take their place among the most valuable prognostics of the weather,most valuable inasmuch as they would give early indication of approaching storms and changes of weather.

OZONE.

613. In the year 1848 Schönbein discovered a new chemical principle, to which he gave the name of ozone on account of its peculiar smell. Ozone is generally supposed to be oxygen in an allotropic state; that is to say, it is the same substance as oxygen, but in a different form, and endowed with different properties. The properties by which ozone is distinguished from oxygen are the following: It smells strongly and has the flavour of lobsters; readily discharges the colour from litmus paper, oxydises silver, burns ammonia spontaneously, and converts it into nitric acid; burns phosphoretted hydrogen immediately with emission of light; decomposes iodide of potassium, setting iodine free, and hydrochloric acid, setting chlorine free; and is a powerful oxydising and chloridising agent. Oxygen may be transformed into ozone by the electric machine; and there can be no doubt that it is constantly being produced by the electricity which is ever present in the atmosphere, and most copiously by lightning

during thunderstorms. As the most powerful known disinfectant, it most readily unites with the gases which arise from decaying vegetable and animal matter, and, by depriving them of their noxious qualities, is a great purifier of the air. It is this property which brings it within the province of the meteorologist, and accordingly it has been extensively observed of late years in Great Britain and on the Continent.

614. Dr Moffat, Hawarden, has made some interesting observations on the connection between ozone periods and the changes of the weather and the prevalence of certain diseases. From these he concludes that when ozone is largely present in the air it is accompanied with diminished atmospheric pressure, increasing temperature and humidity, and the prevalence of the south-west or equatorial winds; and when it is in small quantities the pressure is increasing, the temperature and humidity decreasing, and the north-east or polar winds prevailing. There is a remarkable coincidence between the ozone periods and storm-telegrams. Also when any marked increase takes place in the ozone of the atmosphere it is accompanied with diseases of the nervous, muscular, and vascular systems.

615. Ozonometer.-The following is the mode of preparing Schönbein's ozone test-papers. Take 200 parts of water, 10 of starch, and 1 of iodide of potassium, and boil together for a few seconds; dip bibulous paper into the solution, and after it is dried, cut it into strips. In observing with it, it is only necessary that a strip be placed in an airy situation, free from wet and the sun's rays. The best position is to attach it to a hook on the inside of the roof of the box for thermometers. After being exposed for twelve hours it is taken off, and dipped in water. The depth of the tint of the paper determines the amount of the ozone, and it is compared with a scale showing the different tints marked according to depth from 0 to 10. The tests generally used are Schönbein's and Moffat's. Care should be taken to estimate the force of the wind during the time, since, in a windy day, more ozone is collected, not because there is more in the atmosphere at the time,