ship will cross the path of the centre for the second time between the eighteenth and nineteenth hours, but this will be after the centre has passed on, and the storm will end with a S. wind after about thirty hours.

If we suppose the ship in the same place as before, but the rate of advance of the cyclone 1.5 times as great as the rate at which the ship is sailing, and therefore fifteen knots an hour, she will find the wind still holding in the E. and increasing during the first ten hours. In the next four hours the wind will have the normal shift towards SE., and be most violent at this last point, when the barometer is lowest. The cyclone will terminate after about twenty-eight hours, during the last ten of which the wind will have been due S. She will have approached to within ten or fifteen miles of the centre between the fourteenth and fifteenth hours.

A ship which is hove-to in a cyclone, though she may have one or two knots an hour drift, will always have a normal shift of wind, i. e. the shift which would be given by the preceding rules for either semicircle, the right hand or the left hand one. The seaman cannot be too strongly urged to heave-to in the first instance if he is in waters which are liable to these cyclones, and is led by the signs of the weather to expect one. If his ship is once hove-to, he may ascertain accurately his position with regard to the cyclone, and then, either hold on on his former course, or choose any other which may be advisable for him, or, lastly, heave-to on the proper tack.

As regards the period at which these storms are especially prevalent in the different parts of the torrid zone, we may say, as a general rule, that they usually occur when the sun is at its highest altitude, and reach their maximum of frequency at the end of this period; consequently, in September in the West Indies and the

China Seas; in February and March at the Mauritius and in the South Indian Ocean; while in the Bay of Bengal they appear to be most frequent at the change of the Monsoons, i. e. in May and October. In addition to the figures already given, we find in Piddington's 'Horn-Book' (p. 296, third edition, 1860) the following table*:

TABLE OF AVERAGE NUMBER OF CYCLONES IN DIFFERENT MONTHS OF THE YEAR, AND IN VARIOUS PARTS OF THE WORLD.

The storms of the temperate zone in the southern hemisphere have more the character of gales than of cyclones, as is shown by the fact that the predominant

* I have quoted from the last edition of the Horn-Book, and the table is more complete than that which is given in Professor Dove's work. My reason for doing so is that I find the following remark on the table in the Horn-Book:

'In the former editions of this work, I have stated that no cyclones have been known to occur in the month of May in the China Sea; but from a letter from Capt. E. T. F. Kirsopp, commanding the steamer Juno at Manilla, I learn that a severe cyclone was experienced in the Bay of Manilla and in the adjacent China Sea as early as May 4, 1850.

The essential matter, however, for us at present is the fact, that severe cyclones may occur in May in the China Sea, and thus upon the appearance of doubtful weather or an uneasy barometer the careful seaman will take due precaution.'

The importance of this fact will be, I hope, sufficient excuse for my adding it to the text. Trans.

change of direction is in accordance with the Law of Gyration. In the Ondersoekingen met den Zeethermometer' ('Experiments with the Marine Thermometer'), we read at p. 109-The course of the winds in storms is, with few exceptions, from N. through W. towards SSW. Storms which begin with SSE. last longer, and do not change their direction much.'

The barometer stands lower during all storms in the southern hemisphere than when there is no storm, as is shown by the following table :

This shows us that the barometer stands lowest with the equatorial current, as the polar current does not produce so great a depression.

Ordinary tornados and thunderstorms occur on each hemisphere at the time at which the sun has his greatest altitude; so that they occur during our summer in the northern, during our winter in the southern, hemisphere. The hurricanes also assume the character of thunderstorms, i. e. they are accompanied by heavy rain and

violent electrical explosions. The appearance of the sky, as a cyclone is coming on, is characterised by masses of clouds which are continually changing their form, and frequently by a bank of clouds, in the distance, of extraordinary blackness.

On dry land, in certain regions, the Trombs take the peculiar form of dust whirlwinds, of which Baddeley* has given a very vivid description. The amount of electricity which is excited by the friction of the sand is so great that he was able to obtain from an insulated wire, not only bright sparks, but a continuous discharge of electricity.

As to the motion of the waves in a cyclone, they move out from the centre in directions which are more nearly radii of the circle the farther they are distant from the centre. Consequently, they move from the centre towards the circumference in lines somewhat inclined in the direction in which the storm is rotating. Reid has already drawn special attention to this fact.

Hence we obtain the following differences in this respect between the three classes of storms which we have considered.

1. In a cyclone, the waves move at right angles to the direction of the wind, and the more so the farther they are from the centre.

2. In a heavy gale, they move in the direction of the wind.

3. In the case of a wind which has been stopped by another, they move against the wind. (Seamen then say that two winds are fighting with each other.)

All that we have said hitherto has referred to the torrid zone, properly so called, not to the outer edge of the

* Whirlwinds and Dust-Storms of India, illustrated by numerous Diagrams and Sketches from Nature.

Trade-winds, which shifts up and down with the sun like the belt of Calms. Hence the torrid zone will be bounded in all parts, excepting where the Monsoons are prevalent, by a belt, in which calms are of frequent occurrence, and which is called the sub-tropical zone. This zone forms a complete contrast to that which lies close to the equator. At the latter the air is continually ascending, at the former it is descending. At the equator the barometer is low, at the tropics it is high. At the equator the rain falls in summer, in the sub-tropical zone in winter. At the equator the two winds flow towards each other, in the sub-tropical zone they flow in opposite directions away from the tropics. All stations which belong to the subtropical zone are included during the summer in the Trade-wind, on its prolongation backward at that season, and during the winter are outside its area. Such a prolongation in the direction of the poles as that described, is manifested in its most extensive form in the case where a great desert, like that in Northern Africa, prolongs the torrid zone into higher latitudes to a disproportionate extent. Hence, during the summer, northerly winds are prevalent in the Mediterranean, and are known under the name of Tramontane, while the Sirocco attains as exclusive a predominance in the winter, being the return counter Trade which has descended to the surface of the earth. This is the reason that the seaman finds, at the commencement of the Trade-wind, a more northerly wind on the east side of the Atlantic Ocean than on its west side.

2. DISTRICT OF THE MONSOONS.

During the summer months in the Indian Ocean, the SE. Trade-wind is prevalent in that part of the torrid