(Meteorology,' p. 163), The diurnal oscillation is a phenomenon which invariably makes its appearance in every part of the world where the alternation of day and night exists. . . . Within the Arctic circle, however, the diurnal oscillation dies out, or rather merges in the annual.' These curves of variation can only be determined accurately by a long series of hourly or two-hourly observations, and accordingly they are only known for a comparatively small number of stations, so that it is hardly to be wondered at that the phenomenon has never been satisfactorily explained. About fifty years ago Dove proposed the subtraction, from the total barometrical pressure, of the tension of aqueous vapour (p. 110), as he found that by such separation of the total pressure into its two constituent parts-dry air pressure and vapour tension-the curve of the daily variation of the dry air was far simpler than the original curve, especially at continental stations. It will be seen from fig. 17 (p. 89) that at Nertschinsk in Siberia, where the air is very dry, the lesser maximum and minimum almost entirely disappear in May. principle of separation is, however, not generally accepted at present, and we have not yet arrived at the interpretation of this regular barometrical motion.

This

Of late, however, a line of inquiry has been taken up which bids fair to throw some light on the cause of the phenomenon. Mr. Buchan and Mr. H. S. Eaton have independently investigated the problem of diurnal range in connection with the geographical position of the station where it is observed. Mr. Eaton has calculated the curves for the seven observatories in the British Isles for a year, and has shown how the continental character gradually imprints itself on the curves as we travel from

the Atlantic coast towards the more continental stations (fig. 18). He has also shown that the difference in the barometrical curves is related to the daily range of temperature. Where the latter is small, as at Valencia or Falmouth, we have the minimum at 3 A.M. more marked than that at 3 P.M., while at Kew, where the thermometric range is comparatively great, the reverse is the case, and the afternoon minimum is the more important. Falmouth, however, as will be seen, ex

Diurnal Range of Pressure and Temperature (1876).
The temperature curves are those showing a single oscillation.

Kew.
Falmouth.
Valencia.

hibits curves of pressure and temperature which are intermediate in character between those of Valencia and Kew. When we go to a really continental station like Nertschinsk, as has just been said, we find the morning minimum almost entirely obliterated in some months, and the curve apparently exhibiting a single oscillation.

Mr. Buchan's inquiry has been on a much more extensive scale than Mr. Eaton's, and he has not con

fined his discussions to stations for which the complete range-values have been calculated. Indeed, if he had, material would have failed him. In part of his investigations he has taken the simple difference between the forenoon maximum and the afternoon minimum, and he finds that the amount of daily amplitude so given is modified by proximity to the sea, or even to sheets of water like the great lakes of North America. This shows us that the amount of vapour present in the air is a factor which we must not disregard when dealing with this question, and yet our knowledge of the distribution of vapour in the atmosphere is very imperfect, as we shall learn in the next chapter.

In fact, when we begin to study the barometrical curves closely, we find that we want curves of equally minute accuracy for all the other elements for each station, before we can attempt to offer an explanation of the phenomenon.

In order to throw light on the geographical distribution of barometrical range, Mr. G. Harvey Simmonds calculated its elements for every station for which he could find a sufficiently continuous series of observations, taken at two-hourly intervals at least. The results were published by Mr. R. Strachan in the 'Quarterly Journal of the Meteorological Society,' vol. vi. p. 42. The list of stations in the table is, as might be expected, but small, only numbering thirty in all, but they cover a wide extent of latitude, from Hobarton in 42° 52′ S. to St. Petersburg in 59° 57′ N. The results, as far as they go, furnish evidence in support of Herschel's idea that diurnal range disappears in very high latitudes.

Fig. 19 exhibits the annual curve of pressure for London, for the equatorial part of the Atlantic, and

for Central Siberia. former curves do not vary much from month to month, that for Barnaoul shows a very extensive reduction of pressure of more than 0·8 in. from winter to summer.

It will be seen that, while the two

Barometrical pressure has a yearly as well as a daily range, and this annual curve presents a totally different appearance at continental stations from that which it exhibits, for instance, in these islands. All these changes are, however, connected with climate,

and with the seasonal changes in the distribution of the various meteorological elements. An examination of the charts of barometrical pressure for January and July, Plates VI. and VII., Chapter XIII., will give the reader a general idea of the regions where the barometer is highest and lowest respectively in each of these months, and of the extent of the annual change in different parts of the earth.

CHAPTER VI.

THE MOISTURE OF THE ATMOSPHERE.

WE have already said in Chapter II. that aqueous vapour is the most abundant element in the composition of the atmosphere, next to its principal constituents, nitrogen and oxygen. It is, moreover, the most important of all, meteorologically, inasmuch as it is that which, by its sensitiveness to the action of heat, and by its liability to change of condition from gaseous to liquid and solid, and vice versa, is capable of exerting a great influence on almost all meteorological processes. The subject of this and the next chapter will be the determination of the amount of water passing into the atmosphere, present in it in the vaporous state, and finally passing out of it in a liquid or solid condition.

The subject, therefore, divides itself into three heads:

1. Atmometry, or the determination of the amount of water passing into the air by evaporation.

2. Hygrometry, or the determination of the amount of water present in the air in a vaporous form.

3. Hyetometry, or the determination of the amount of water condensed out of the atmosphere in the form of rain, hail, or snow.

George Stephenson's saying that it is the sun which works railways,' is well known, and is perfectly