meter which accompanies a storm approaching from the west, and it gradually veers to the south and west as the storm advances. The other south-east wind is accompanied with dry weather, a clear sky with haze near the horizon, and a high barometer still rising or stationary. It is the result of a higher atmospheric pressure descending over Russia and central Europe, and it does not veer to the south and west. Since this, our only really settled fine weather, is almost always a consequence of the polar current coming from the north of Russia and spreading itself over Europe, it will scarcely be necessary to point out the great value of daily telegrams of the barometer and winds from Archangel and Riga in giving the first intimation of its probable approach to spread sheets of ice over our ponds and lakes in winter, to dry the land in spring for the reception of seed, and to bring sunshine in summer and autumn for the ripening and ingathering of the fruits of the earth.

657. Quite analogous to the above is the north-east wind in its relation to weather-changes. When the centre of a storm passes to the eastward at some distance south of any place, the wind which points to its approach is not a S.E. but a N.E. wind. This is particularly the case when the pressure has been for some time high in the north, which in winter occurs with protracted frosts. Hence one of the surest signs of the breaking up of a severe frost is the setting in of the N.E. wind, if it be accompanied with a falling barometer, and a green or yellowish-green tinted sky; and it is all the more certain if the sky becomes gradually overcast, and the wind veers from N.E. to N. and N.W. It is accompanied with a fall of snow, which passes into sleet as the wind shifts into the west, and finally into rain. On the other hand, if the north-east wind be accompanied with a clear sky with haze near the horizon, and a high barometer rising or stationary, and if the wind does not increase in strength, and tends to veer rather in the direction of E. and S.E., the weather will remain settled for some time-there being nothing as yet to indicate any change.

658. It is thus by a careful observing and recording of the lesser changes of the weather from day to day that the approach of the greater changes included under storms of wind, rain, or snow, may to some extent be known beforehand. It is not by a single observation made at one time that an isolated observer can best draw conclusions regarding the weather likely to happen, but from the character of the changes which have been silently going forward for some time. Hence it is desirable to set down, in a pictorial form, the different observations as they are made, so that the eye may take in at once the whole of the changes which are going forward; and thus the observer may with less trouble and with greater certainty draw his conclusions. The Rev. R. Tyas's handy little annual, 'How to use the Barometer,' is calculated to be popularly useful for recording, pictorially and in tables, the successive changes of the weather from day to day.

STORM-WARNINGS.

659. It is in tropical and sub-tropical countries that an isolated observer may, with the greatest degree of certainty, predict the approach of gales and hurricanes. In these regions barometric pressure and the other meteorological elements are so constant from day to day, that any deviation, even a little, from the average of the hour and season in respect of pressure, the direction and strength of the wind, and the direction and amount of cloud, supposes the existence of a storm as the disturbing agent at no great distance. This important practical problem has been worked out with great success by Mr Meldrum at Mauritius. A few words will explain the method of proceeding.

660. At Port Louis, Mauritius, in July, the mean pressure at 32° and sea-level is 30.274 inches at 9 A.M., from which it falls to 30.186 inches at 3 P.M., it then rises to 30.236 inches at 10 P.M., and again falls to 30.192 at 4 A.M. The

Y

mean direction of the wind for the same month is from S.E. to E., generally E.S.E.; and its force about 0.41 lb. to the square foot, subject to an increase or decrease at different times of the day. Suppose now that the barometer was observed to fall after 9 A.M. with a greater rapidity than was due to the usual daily range; or that in the afternoon it did not rise to the second maximum, or that it fell instead of rising; or suppose, in short, any deviation from the observed daily march, then it is certain that there is somewhere an atmospherical disturbance, near enough to Mauritius to influence the pressure.

661. The direction in which the disturbance is from Mauritius is readily known from the wind. Thus, during the gale which occurred from the 11th to the 17th January 1860, the wind veered from S. E. to S., S. W., and W. This was a great revolving storm, the centre of which passed on the east side of the island, its nearest distance at any time being 129 miles. Here it will be observed that the storm with its low pressure being in the east, and travelling to the south-westward, the wind at Mauritius veered from E.S.E., its usual direction, into the west; or, to express the relation from the opposite point of view, the veering of the wind pointed to the east and south as the seat of the atmospheric disturbance. In all cases the wind is found to veer round so as to blow approximately at right angles (60° to 80°) to the bearing of the storm. The distance of the storm may be closely approximated to by observing the rate and amount of fall of the barometer, taken in connection with the observations of the wind and the clouds. The rate and direction of the progressive motion of the storm are known chiefly from the veerings of the wind.

662. Mr Meldrum worked at the solution of this problem for several years, and took a very effective course to test the justness of his conclusions. He laid down the path of the storm from the Mauritius observations alone; and on afterwards receiving, from captains of vessels who had encountered the storm, a note of the latitudes and longitudes,

stated, to their surprise, when and where they had the storm and the direction and veerings of the wind during its continuance. After sufficient experience had been acquired, a note was sent to the daily newspapers when it was concluded from the Mauritius observations that a storm was abroad, stating its position and probable course from day to day. These notifications, inferences (or predictions, as they may be called when the storm afterwards reached Mauritius), were carefully compared with the logs of ships which afterwards touched at the island. No case of failure has occurred since these notices began to be sent to the daily press.

663. This gratifying result is of great value in various ⚫ ways. It shows what may be done at an isolated station in the ocean; in other words, what may be done in ships at sea. Since these deductions regarding storms are made at Mauritius, from the observed deviations from the average height of the barometer and the direction and force of the wind, the reduction of the ocean statistics of meteorology, and the charting of the results for the use of sailors, becomes a question of pressing importance. Maury's Sailing Charts have already been referred to for the signal service they have rendered to navigation. The Isobarometric Charts given at the end of this book will, it is hoped, prove a valuable auxiliary to navigation; and when the whole thirteen have. been prepared, their value will be much increased.

664. In discussing Meteorological Ocean Statistics referring chiefly to tropical and sub-tropical regions, the most important points to be determined are the following: 1. The daily range of the barometer for the different months, as applicable to different parts of the ocean; or, better still, the law of its variation according to season, latitude, and proximity to large masses of land. 2. The mean monthly barometric pressure of the different seas chiefly traversed by ships, given for every 0.050 inch, or, if possible, every 0.025 inch of mean pressure. 3. The deviation from the mean pressure of the month which may be regarded as an amount of disturbance sufficient to be considered as indicative of a storm at no great

distance. 4. The mean direction and force of the wind for the different months. This is no doubt a vast problem which it will take many years to work out. Mr Meldrum's paper on the Meteorology of Port Louis in the Island of Mauritius, printed in extenso, with its elaborately prepared 42 Tables, in the Report of the British Association for 1867, is one of the most valuable contributions towards its solution yet made. Much may in the mean time be done by shipcaptains to enable them to steer clear of storms, if they would carefully note the occurrence of deviations from the average daily barometric range, and barometric fluctuations generally, taken in connection with the direction, force, and veering of the wind.

665. In regions such as Great Britain, an isolated observer cannot, with a like certainty, draw conclusions from his own observations regarding the approach of storms and other weather-changes. But he may almost, if not altogether, attain to the same degree of certainty, if he be assisted by a sufficient staff of observers, well distributed over Western Europe; and be able, when he considers it necessary, to communicate with them through the telegraph. The possibility of this is evident from what has been already stated in the chapter on STORMS, and its practicability is seen from what was done by Admiral Fitzroy, and his assistant, Mr Babington, previous to December 1866.

666. In his letter of 15th June 1865 to the Board of Trade, General Sabine states that he had examined the warnings given on our coasts during the two years ending 31st March 1865, and found that in the first year 50 per cent, and in the second year 73 per cent, were right; and he adds, "That it seems not unreasonable to attribute to increased experience the marked improvement of these results upon those of the preceding year, and to anticipate still further improvement." The above percentages of warnings which turned out correct, were formed from the warnings sent to all the seaports. But, since a number of these places