more easterly: the only winds which can preserve their direction unaltered over a large area are due E. and W. winds.

Rotations of the vane against the sun, which extend beyond S. or E., indicate cyclones; if they only extend from NW. to SW., or from ENE. to NNE., they are often only a return of the vane to its original position, indicating, in the one case, that the equatorial, in the other, that the polar, current continues prevalent.

4. SOUTH TEMPERATE ZONE.

The regular rotation of the wind in this zone is with the sun, i. e. from S. through E. and N. to W. and S.; and similarly in cyclones, when the ship is on the NE. side of the path which is travelling from NW. to SE., it is from NNE. through N. and NW. to W. and WSW.; and in both cases the barometer falls until the wind reaches NW., and then rises. The only difference is that in a cyclone the temperature does not vary to any extent, while in the case of the alternation of the regular currents it rises while the barometer falls, and vice versa. If the wind veer from W. through SW. to SE., the barometer in general rises and the thermometer falls. The prevalent wind, when the barometer is highest and the thermometer lowest, is SE., and when the converse conditions are fulfilled, NW., especially in the cold season; the atmosphere also is clear with SE., and thick with NW. If the wind veer from W. through SW. to SE., the weather clears up; if from SE. through NE. and N. to NW., it breaks, and there is rain. If the wind veer against the sun from ENE. through SE. to S., the ship is probably on the SW. side of the path of a cyclone travelling from NW. to SE. The rules for finding of the centre and the course to hold have given.

The only parts of the earth which I ha sidered are the frigid zones. The stormy F seem to be the summer, and the transition fro summer; the winter itself is, comparatively s time of calms. In the American polar sea, the stands at its highest level in spring. Accordin observations, the permanently low level of the which was first observed at Cape Horn by Kr appears to extend far into the antarctic zone. trict of barometrical depression is of far great than that in the vicinity of Iceland. The cold over the ice-fields, seems to stop the most viole winds, which accordingly deposit their aqueous v heavy falls of snow in their attempts to force a hence large floes which are rotating are surround wall of snow. Very little of this snow penetrate interior of the floe, while the different points of t cumference come successively into the area whe contest is going on (Scoresby). In the immediate bourhood of the pole the rotation of the vane be complicated; since the influence of the rotation earth on the wind changes as soon as storms pas pole, from the fact that the velocity of rotation c surface with which the air comes in contact, which been previously decreasing, begins to increase again. dense fogs which arise from the difference of tempera of the sea, and the very cold air lying immediately al it, and the similar fogs which are due to the differenc temperature of the air over the sea and over the ice-fl are the prevailing form which the condensation of moist takes in these regions. This form is also not unusual

spring in the N. Atlantic Ocean, owing to masses of ice which are drifting southwards, and is an indication of the proximity of icebergs, especially off Newfoundland.

There are too few data for the Pacific Ocean to permit me to enter into a detailed examination of the differences between the phenomena on that and on the Atlantic Ocean. The description of the storm in the harbour of Avarua, in Raratonga, which is given by Williams,* shows that the hurricanes are very violent in those seas. He says: 'The whole island quivered to its centre when the waves broke on the coasts. A vessel belonging to the missionaries was carried over a marsh into a wood of large chestnut trees some hundred yards from the shore. The rain fell in torrents from morning to night.'

The practical rules which have just been given are intended to serve a twofold purpose. Firstly, they will indicate to the seaman the conclusions as to approaching weather, which he may draw from the appearance of the sky and the behaviour of the meteorological instruments, especially of the barometer. Secondly, they will show him which of the phenomena are, as yet, imperfectly explained, and which, therefore, demand a more accurate investigation by means of additional observations. It is very satisfactory to find that practical seamen like Fitzroy, Maury, Van Gogh, Andrau, and Jansen, are taking steps to indicate to the officers of the naval and mercantile service what points it is important, not only for science, but also for themselves, that they should ascertain. By this means central stations, like the Board of Trade, the National Observatory at Washington, and the Meteorological Institute of the Netherlands, have beer at which the fragmentary materials are reg worked up into a collected whole.

In my opinion the strictly meteorological such investigations has not been brought fo sufficiently prominent manner in the instruction by these institutions; and I have, therefore, supply the deficiency by the present work.

The theory which has been here propou which has been developed by me in variou since the year 1827, assigns answers to the problems :

1. Why the storms of the torrid zone app frequently in certain districts than in others.

2. Why they take the form of cyclones; and rotation in a cyclone is in a different direction northern to what it is on the southern hemisphere

3. Why they move in fixed directions wit tropics, and turn at a right angle as soon as the the boundary of the torrid zone.

4. Why the cyclone increases in diameter and de in intensity when this change of path has taken pla 5. Why the form of the storms of the temperate presents more variety, * and why in these districts c characters of storms are more prevalent at certain seasons and at certain localities than at others.

In conclusion, I wish to draw attention to the fact, that the theory which has been developed in the foregoing pages is only intended to exhibit the principles from which we may deduce, not only the Trade-winds and Monsoons, but also the regular movements of the atmosphere of the regions of changeable winds, when it is not disturbed by storms. According to the explanation which has been given of them, the hurricanes tend to accelerate the earth in its motion round its axis, whereas the constant Tradewind tends to retard it. The compensating element for the conservation of the earth's rotation, which would otherwise be affected by the great Trade-wind currents, is furnished by the various westerly currents, viz. the predominant equatorial, and consequently westerly current in the temperate zone; the SW. Monsoon of the northern, and the NW. Monsoon of the southern, Indian Ocean; and, lastly, the hurricanes. Despite the destructive action of the last-named movements, they must still be regarded, in the general sense of the term, as agents in the conservation of the vital force in the great organism of the earth. The whole of the phenomena which we have described furnish us with confirmatory proofs, on a great scale, and borrowed from the earth itself, of the great fact that the earth rotates about its axis-a fact whose first discovery is due to the science of astronomy.

In this second edition I have discussed the theories of cthers at greater length than I did in the first. My reason for doing so has not been to bring forward my own theories more prominently, but to show the error of the idea that all atmospherical phenomena may