the synthesis; it rebuilds indeed the twelve carbon atoms' moleculeunder one condition only, that the product of the synthesis should be eliminated in proportion as it is formed. But such an elimination is, in all probability, what really takes place in all life processes.

The wide consequences of this new step can easily be foreseen. As Professor Maquenne, a specialist himself in this branch, truly says in an enthusiastic article upon Mr. Hill's discovery, 26 it marks an epoch in biological chemistry.'

III

Is it possible to foretell weather several days, or maybe weeks, in advance? This is the question which now engrosses the attention of many practical meteorologists. Popular wisdom has always said 'yes' to this question, and there are in the weather-lore of each nation many sayings to this effect. Some of them belong, it is true, to the same domain of superstition as astrological predictions. Such is, for instance, the once so famous Bauern-Praktik whose origin Dr. Hellmann has traced as far as ancient Greece and the Vedas. But there circulates also, amongst mariners, peasants, and hunters all over the world, a certain stock of practical knowledge of weather which is based upon a correct observation of nature. When the Greeks say that the autumn and winter months are months of gales; or the North-west Canadians predict a spell of warm and dry weather after a snowstorm of short duration has blown early in autumn; or the Russian peasants remark that when the first snow has fallen upon an already frozen ground the snow will lie late in the spring, and the spring will be cool-there is scientific observation in such prophecies, and recent researches into the seasonal periodicity of gales in Greece, the character of weather in the North-western prairies, and the influence of the snow-cover upon spring temperatures in Russia have decided in favour of these practical observers. The question is consequently this, Cannot science do better? After having succeeded in forecasting weather twenty-four hours ahead, cannot it make a further step in advance?

The means by which meteorologists succeed in issuing the daily forecasts which we now find in the morning papers have so often been described that a few words will be sufficient to refresh in memory the leading principles of these prophecies. In every civilised country of the world there is one or several weather bureaus whereto telegrams are sent, once or twice daily, from a great number of meteorological stations, reporting the state of the weather at each station the height of the barometer and the thermometer, the direction and the force of the wind, the cloudiness of the sky, and so on. As soon as this information reaches the central bureau it is embodied

26 Revue Generale, the 30th of December, 1898, vol. ix. p. 927.

in a weather chart by means of a system of conventional signs. All the spots at which the atmospheric pressure (or rather the corrected height of the barometer) is the same-300 inches, 29-9, 29-8, and so on -are connected on the map by curved lines or isobars,' which show at a glance the distribution of pressure over a wide area. The same is done for temperatures; while the wind which blows at each station, the state of the sky, and the amount of rain that has fallen during the previous day are marked on the map by comprehensive symbols. A true picture of the different sorts of the weather experienced in the region which is covered by the weather chart is thus obtained. Every one knows these weather charts which are exhibited at different places and are printed in some leading paper in each country-the Times for the British Isles. They are so comprehensible that even the uninitiated reader, if he sees in the morning that the isobar curves are sinuous and contorted, and run close to each other, is tempted to predict that the weather will be boisterous and unsettled during the day.

The atmosphere of the earth is never at rest. It is involved in a general circulation during which masses of air, hot and cold, are carried at different levels from the equator to the poles, and back to the equator.27 But local depressions, or local eddies-similar to those which are seen in a swift current of water-are also formed here and there. And it is these eddies, or centres of low pressure, which determine the wind that will blow at a given spot, the clouds that will obscure its sky, and the amount of rain that may fall upon it. The weather will be different to the east and to the west, to the north and to the south, of a local depression of atmospheric pressure.28 Besides, these eddies continually shift their positions, and the main difficulty is to foretell whereto this or that centre of low pressure will move, and how its dimensions will be altered within the next twenty-four hours. If such a centre of depression has made its appearance on the western coast of Ireland, it will shift eastwards as a rule; but it also may be shifted to the south-east, or, after having described a U-shaped curve, it may creep next towards Iceland; and the weather at, let us say, London will depend entirely upon whether London is now in the centre of the depression, or in its front or rear, to the left or to the right of its path.

The tracks followed by each of these centres of low barometric pressure for the last thirty years (they are still named 'cyclones,' although no real cyclone storm is implied) have been the subject of most laborious investigations. For every separate region-the British Isles, Western and Eastern Europe, South Russia, India,

27 The laws of the general circulation have been discussed once in these pages, Nineteenth Century, April 1893.

28 By using the word 'eddies' it is not meant that real eddies be formed. The word is only a convenient description of an analogous feature.

Japan, and so on-we have now detailed descriptions of the different types of depressions, and atlases of the tracks which they follow at different seasons. The result is that when an experienced forecaster looks on this afternoon's weather-chart for Europe, or the States, or India, and compares it with both the weather-chart of the previous day and his atlas of cyclone-tracks,' he can foretell whereto the centre of low pressure will be shifted by next morning, what will be the probable distribution of isobars, and consequently what sort of weather will prevail next day in the different sections of his own country. He issues his forecasts, and in nearly nine cases out of ten they are correct, although the forecast as regards rain is beset with great difficulties, especially in these isles, on account of the importance of local conditions. 29

It has often been objected that although meteorologists have undoubtedly succeeded in placing weather predictions twenty-four hours in advance upon a scientific basis, the practical value of their forecasts is not yet great.30 However, this last depends entirely upon the methods of bringing the forecasts to the knowledge of the population; the rapidity with which warnings of changes of weather are issued and disseminated; the degree of confidence that has been won by the local meteorologists; and finally upon the average level of popular education. Even in the British Isles, for which weather predictions are beset with more difficulties than anywhere else, the storm-warnings and partly the weather forecasts are taken notice of by the population. But it is especially in the United States that one sees how much the meteorological service may become part of the daily life of a nation.

The daily weather-charts and the forecasts are prepared in the States with wonderful rapidity. The forecasts are ready one hour and forty minutes after the observations have been made (at eight o'clock of the 75th meridian) at from 2,700 to 2,960 stations scattered over the States and Canada, as well as at the auxiliary stations of Mexico and the West Indies. Immediately the forecasts are telegraphed and spread broadcoast, reaching nearly 30,000 persons and institutions. The local and the auxiliary bureaus, as well as the post offices, spread them by all means, including free postcards and telephone messages. The warnings of frosts and blizzards in February, of night frosts in the spring, of storms on the coasts and squalls on

"This is why the tendency is now to decentralise the weather forecasts. In the United States twenty-six weather bureaus have lately been created: they receive all the telegrams (on the circuit system) and issue independent forecasts. Experience has shown that they are of an immense value. Besides, the best meteorologists (Abercromby, Bebber, Woeikoff) encourage by all means individual forecasts, and invite all persons interested in weather to practise in that art.

*The first part of this statement has also been contested lately by no less an authority than Dr. H. Klein, but with little success. See the most instructive discussion on this subject which took place between Dr. Klein and Dr. Bebber.

the lakes, of inundations, and so on, are the subject of a special care. Thus, last winter, when a cold wave and a blizzard were expected in the West, 650 points in twelve ranching States, as also all the railway and steamboat stations, and thousands of private persons were warned from the Chicago weather bureau. Immediately most ranchers took their flocks of sheep under shelter (200,000 head of sheep and cattle in one single small spot), and masses of both sheep and cattle were saved from an almost certain destruction by an awful blizzard.31 In April last most valuable crops of strawberries were saved in the same way. The strawberries were covered with straw, or artificial clouds were made.32

The squalls which are going to blow on the great lakes; those which are foreseen to sweep over the Columbia River during the salmon season; the storms and rains that may be fatal to crop operations in Dakota; the rains which may damage the drying of raisins in California; and the coming floods of the Mississippi are telegraphed in the same way to the respective regions, either from Washington or from the local forecasting bureaus. Moreover, great numbers of private telegrams, to inquire whether next day will be favourable for salmon-fishing, or to what height the Mississippi or such a river may rise during the next days of hay-making, or when a big raft of timber ought to be floated, are received in numbers at the weather bureaus and immediately answered. Nay, the meteorological service has so much won the confidence of the population that last year it was very seriously urged by the Press to issue forecasts of increase of crime,' it being known that such an increase really takes place during some sorts of hot weather.

IV

At the present time the weather forecasts which are issued every evening cover only twenty-four hours in advance. Thirty-five years ago even such forecasts were described as an awful self-conceitedness on the part of the meteorologists, sufficient to discredit them. Maury

31 Monthly Weather Review, vol. xxvi., March 1898.

32 Same publication, April 1898, p. 139. Mr. Willis L. Moore has made a special study of the 'cold waves,' and is most successful in their prediction. In the above mentioned case twenty-four stations were warned from Washington. The warnings were also widely distributed by mail from Raleigh, Tarboro, and Parmele by the logotype system. A number of display men, besides posting warnings at the post offices and depôts (shops), also distributed them by telephone. . . . They were also, where opportunity offered, sent into the country and circulated verbally. . . . The average time that the warnings were received in advance of the frost was fourteen hours.' The words of warning are also attached to the stamps of the receiving postoffices, and they are printed on all letters, postcards, and papers issued from the offices for distribution. The weather warning is thus printed by the same movement which the post officer has to make in order to stamp the letters.

himself lectured Fitzroy on this theme. But now such forecasts are already found insufficient, and on all sides the desire is expressed to know the coming weather several days, if not weeks, ahead. Meteorologists have thus to face a new problem, and they approach it in two different ways. On the one side assiduous researches are made in order to see whether there is not a certain periodicity, or certain cycles, in the recurrence of hot and cool, dry and wet weather; and on the other side research is directed towards ascertaining the different types or spells of weather, their duration and the succession in which they follow each other.

It is now certain that the number and the size of the dark spots which we see on the surface of the sun are in some way connected with the weather which we have on the earth. Charles Meldrum, Sir Norman Lockyer, the Indian meteorologists, and especially Dr. W. Köppen in his great work, have proved that there is a certain periodicity in the temperature, the rainfall, the number of cyclones, &c., which corresponds to the eleven years' periodicity (11.1 years) in the number of sunspots.33 However, the amount of variation which may be due to this cause is so small in comparison with the non-periodical irregularities of weather that it is often masked and obliterated by them. Moreover-to say nothing of the connection which exists between the sun-spots' period and the magnetical forces in our atmosphere-the whole matter, as has been shown by Polis, is more complicated than it seemed to be at first sight. It appears that when the sunspots are at a minimum, mild winters and hot summers prevail, while cold winters and cool summers seem to characterise the maximum periods of the sunspots; while Mr. A. McDowall points out that not only the seasons and fractions of the year, but different days as well, must be treated separately in all discussions upon the influence of the sunspots' periods. Years of sunspots' maxima are, in his opinion, years when the monthly and daily extremes of temperature are greater as a rule.35 In short, our weather is undoubtedly influenced by the eleven years' periodical variation of the Sun's radiation which is indicated by the sunspots. But this influence is only now studied in such detail as to be taken into consideration in weather predictions.

34

Another weather period, which perhaps has not yet been taken sufficient notice of, is the thirty-five years' period discovered by the

"Henry F. Blanford summed up this question a few years ago in Nature, vol. xliii. 1891, p. 583.

Das Wetter, vol. xi. 1894, pp. 73, 169.

"Meteorologische Zeitschrift, 1896, vol. xxxi. p. 431. As to the quantities of rain and snow during the maximal and the minimal sunspots' periods, the whole matter is too complicated to be expressed in one sweeping sentence. Local conditions at different seasons must be taken into account, as may be seen by comparing the researches of Dr. H. Klein (Meteorologische Zeitschrift, 1897, p. 145) with the abovementioned papers.