band auroras are less frequent and when visible are usually seen to the southward of the observer.

the belt of maximum frequency more than eighty auroras are seen annually. The course of the belts of equal auroral frequency bears some general resemblance to that of the isoclinal lines or curves of equal magnetic dip.

The aurora is by no means confined to the northern hemisphere. It is rarely seen close to the equator, or in low latitudes generally, but it appears just as frequently in the southern sky as in the northern, and of course there receives the name of Aurora Australis. Several especially brilliant displays, of late years, have been noted simultaneously in Europe and in Australia, and in fact there is every reason to believe that all northern auroras must have their corresponding southern displays.

As to the connection between the appearance of the aurora and weather nothing very positive can be asserted. That there is some relation appears from the circumstance that at the time of a brilliant display the sky is constantly being covered with cloud and clearing again. In these islands, at least, an aurora is rarely seen except at times of atmospheric disturbance; in fact it not unfrequently accompanies severe storms. When, however, we pass to the region of greater auroral frequency no such agreement is observable, and most Arctic travellers deny its existence. In fact, the late Lieut. Weyprecht, the Austrian explorer, stated that after discussing carefully the observations taken during his long sojourn in high latitudes, on the voyage in which: Franz Josef Land was discovered, he could find no trace of a connection between the aurora and the weather.

As to the connection between the aurora and the phenomena of terrestrial magnetism much remains to be explained. It is generally stated that during the continuance of an aurora the indications of the magnetic instruments are much disturbed, that, in fact, what is called a 'magnetic storm' occurs. This would be accounted for if the aurora were a discharge of an electrical current in the upper regions of the atmosphere, for, as is well known, the flow of an electric current near a magnet at once affects its magnetic condition. However, it appears from the continuous magnetic records at Kew that all auroras are not coincident with disturbances of the magnetic instruments; and Dr. Wijkander, in his discussion of the magnetic results of the Swedish expedition to Spitzbergen, in 1871, states that, as to intensity, no relation is traceable between the magnetic disturbances and the brilliancy of auroral phenomena. The subject, therefore, demands fuller inquiry.

Ozone.-Before leaving the subject of electrical phenomena we must not omit all mention of ozone, although in the opinion of most meteorologists at present, the less said about that substance the better.

Ozone was discovered by Schönbein in 1848, and he gave it its name from its peculiar odour (from the Greek word oo, I smell). We often recognise a peculiar odour in the atmosphere after a thunderstorm, and this is popularly taken, in the case of lightning striking near a person, for the smell of burning sulphur! One of the modes of production of ozone is the passage of electric sparks through a confined portion of air, and this accounts for its presence in thunderstorms.

Ozone is nothing else than oxygen in what chemists call an allotropic condition; it still remains an invisible gas, but is endowed with properties different from those of ordinary oxygen. It has become much more energetic in its chemical action, oxidising objects exposed to it with great rapidity. It is a good disinfectant of the atmosphere, inasmuch as it hastens the oxidation of decomposing animal and vegetable matter: hence it is important to life and health.

It is more abundant on the sea-coast than inland, in the country than in towns, and it is stated to accompany South-westerly winds as distinguished from North-easterly.

The mode of observing it is to expose slips of white blotting paper, soaked in a solution of iodide of potassium and starch, for a definite period, in a cage allowing free access of air, while excluding direct sunlight and rain, and to note according to a definite scale the degree of discoloration the paper exhibits.

Having said thus much about the substance, I give the reasons for the opinion expressed above-reasons so well founded that at the recent International Congresses at Vienna in 1873, and at Rome in 1879, it was unanimously agreed that no mode as yet suggested of observing ozone could be recommended for adoption.

In the first place, a substance, called the peroxide of hydrogen, is occasionally present in the atmosphere. This exerts an action similar in many respects to ozone, for which it is often mistaken, but it does not possess the same beneficial qualities. Not only this gas, but other oxidising agents, such as nitric acid, may affect the ozone papers. In fact, in the course of a recent discussion on ozone, before the Meteorological Society,

Dr. J. W. Tripe, an experienced observer of ozone, stated that the highest degree of discoloration which he had ever noticed had occurred after a grand display of fireworks in his neighbourhood. This, therefore, was not caused by normal atmospheric action.

Again, it is evident that if we are to compare the proportion of ozone in the atmosphere on different days we must cause precisely the same amount of air to pass through the cage on each day. But simple exposure to the air will not effect this, as the wind varies in force continually.

Lastly, the discoloration of the paper is not permanent; the tint once developed may be again discharged by atmospheric action, so that we can have no certainty that we find recorded on the paper the real amount of ozone present in the air.

Note to page 181.-Mr. De La Rue and Dr. Müller repeated their experiments with a battery which was gradually increased to 15,000 cells, and they then found that after a certain number of cells had been reached the law of the squares ceased to hold good, and that the intensity of the force required to produce a discharge at a certain distance became sensibly proportional to the distance, and that it required 9,700 volts to produce a discharge through 1 centimètre (3937 inch). Starting from these data, the electromotive force necessary to produce a flash of lightning one mile (63,360 inches) long, at ordinary atmospheric pressures, is 1,480,570,000 volts.

199

CHAPTER XI.

OPTICAL PHENOMENA.

THERE are four classes of phenomena depending on the action of light: the rainbow, the corona, the halo, and the colours of the sky and clouds.

A rainbow is produced whenever the observer is in full sunshine, while rain is falling on the side opposite to that in which the sun is situated. The bow is a portion of a circle whose centre lies on the prolongation of a line from the sun to the observer's head, the radius of the red bow being nearly 42°. The amount of the circle visible at any time depends on the altitude of the sun. If the sun be actually on the horizon, the bow will be a semicircle. If the sun be at an altitude of 42°, the bow will be depressed, and the top of the red arch will just touch the horizon. With greater altitudes of the sun the bow will be beneath the horizon altogether, and therefore invisible. For this reason the rainbow is more commonly seen in the morning and evening than in the middle of the day, especially in summer.

When the observer is situated on a mountain top, the bow sometimes forms an entire circle, and we all know that complete circles are occasionally seen in the spray of a waterfall.

It is hardly necessary to describe the bow. It exhibits the prismatic colours in their well-known