ripples, and isobars map out the varying pressure on the earth's surface induced by the uneven flow of air. If we had a series of barometers at the bottom of a river, and observed them all simultaneously, we should find a defect of weight under the eddies, with an excess of pressure under the backwater, and we might draw isobaric lines which would map out the position of these vortices exactly as we do in air.

But it must be specially noted that the depression of a cyclonic eddy is not entirely due to centrifugal force like that of a water eddy. The depression of an ordinary cyclone could not be produced by the centrifugal force of the fiercest hurricane that ever blew; and the cause of the diminution of pressure in a cyclone is at present

unknown.

When we discuss the changes shown on two different synoptic charts, we shall talk of changes in the shapes and positions of isobars as if those lines were graphical abstractions. What we really do is to trace by their means the ever-changing eddies of the atmosphere-the death of old ones, the birth of new ones, or the fusion of several existing systems into a new disturbance of a different type.

To avoid all danger of theoretical errors, we simply define the kinds of eddies by certain abstract shapes of isobars, and then, entirely from observation, we collate certain definite kinds of weather and sky, with different portions of each isobaric configuration.

CHAPTER V.

BAROGRAMS, THERMOGRAMS, METEOGRAMS.

WE have already seen that a series of synoptic charts are, as it were, a series of bird's-eye views, not only of the appearance of the sky, but also of instrumental readings over a considerable area. Owing to the expense of telegraphy, these can rarely be taken oftener than every eight hours, but we all know by experience that very great changes of weather may occur within that time. Every well-equipped observatory is therefore supplied with instruments which record automatically the height of the barometer, of the thermometer, the direction and velocity of the wind, the occurrence and quantity of rain, while a wet-bulb thermometer for moisture and an electrometer are sometimes added.

The trace marked on paper by a barograph is called a barogram; that by a thermograph, a thermogram; both or either of the records left by the wind-instruments are called anemograms; and if all or several of these are combined in one diagram, the whole is called a meteogram, because it is a writing of meteorological instruments. Fig. 25 is a copy of one of the meteograms which are

published by the British Meteorological Office, and we propose to devote several paragraphs to its consideration, so as to explain how, by means of these continuous records, we can fill up the gaps in the history of weather-changes

which are shown at considerable intervals only, on synoptic charts.

the

sequence

It should be remarked that in treating of these we no longer deal with generalities, but with the actual changes and variations of each day. Also that the aim and object of all meteorology is to explain the details of of weather as it occurs at any one place, if possible with great minuteness, and that synoptic charts are only a means to that end. Ever since the discovery of the barometer and thermometer, innumerable attempts have been made to discover the nature of weather-changes by calculations, which were based on the fluctuations of these instruments as observed at any one place, but with

very moderate success. We shall very soon see why this was so; and, in fact, why it could not be otherwise, from the nature of things.

METEOGRAMS.

The

Now for our actual meteogram in Fig. 25. barometer-trace is marked "bar," with an appropriate scale at the edge. Temperature is marked "therm;" while the upper curve of all, marked "vel," gives the varying velocity of the wind, only we must note that for convenience its base is taken from the top of the diagram, and that, therefore, it is measured downwards. Rain is marked near the bottom of the diagram by oblique lines, which are proportional to the amount of rain measured at the end of each hour, so that we can see at a glance when rain fell and how much of it. Below all is the trace marked "direct," which refers to the direction of the wind. The notation requires a word of explanation, as we have to represent an angular scale on a rectangular diagram. It will be seen that the bottom of the diagram is marked N for north; then as we rise upwards we get successively east, south, west, and back to north again at the top of the direction-figure. By this means we can see at a glance whether the wind is veering or backing. When the wind veers, the trace moves upwards; while, when the wind backs, the curve descends. A little practice is necessary to read this easily, but when once learnt it is very convenient.

But to understand this meteogram fully, we must look at the charts given in Figs. 26 and 27, for the 8th and 9th

of December, 1874, respectively. The broad features are very simple. The cyclone which we see approaching in the first chart passes across England to Holland, where we find it in the second chart; and now let us see how

FIGS. 26 and 27.-Charts to illustrate meteogram.

this affected the sequence of weather at Stonyhurst, near Manchester, marked s on the charts.

The meteogram, Fig. 25, refers to the three days, December 7 to 9, 1874, while the charts refer to the two latter days only. The barometer rose till nearly midnight on the 7th, under the action of the wedge which we find over the North Sea on the chart for the morning of the 8th. Then we see by the trace that the mercury was falling fast, owing to the approach of a cyclone. This fall continued till about midnight, when a rapid rise com