METHODS OF FOEECASTING THE WEATHER. 159 



A new, and we must at once say a truly empirical method of weather 

 prediction, is that at present in use In* all the official central meteoro- 

 loo'ical establishments in the world. This method has o-raduallv and 

 slowl}^ develo})ed according- to the exact rules of investigation in 

 scientific practical meteorology, and is still far from having reached 

 perfection. It has developed entirely, without any addition of an 

 a priori nature, out of the observations of the weather processes, and 

 is therefoi'e based entirely upon well-established observational data. 

 The most fundamenttd of these facts is that the weather is associated 

 with the distribution of atmospheric pressure. It has been recognized 

 more and more clearly b}^ experience that the weather is determined 

 not by pressure as shown by the barometer at the place of observation, 

 but by the barometric conditions that prevail over vast regions; for 

 instance, those distributed over the whole of Europe. Therefore one 

 must chart and study the distriljution of atmospheric pressure over 

 the whole of Europe if one wishes to understand the weather actually 

 prev^ailing. 



It was necessary, first of all, to determine by extended observations, 

 made as nearh^ simultaneous as possible, the distribution of atmos- 

 phei'ic pressure for a definite hour, in order to perceive to what kind 

 of weather this distri))ution of atmospheric pressure corresponded. 

 It was by this means demonsti'ated that there is an extraordinarily 

 great variety of forms of atmospheric pressure distribution; that 

 these, however, can be classified into a certain number of types b}^ 

 having regard to the form as well as to the weather conditions given 

 in these forms. * * * Xhe thorough and persevering study of 

 the weather that prevails on the occurrence of each t\ pe has led to 

 the definite and certain recognition of the following theorems: 



1. The weather, in all its details, depends upon the distribution of 

 atmospheric pressure, and the same weather always corresponds to 

 the same location relative to this distribution. 



2. The weather of any place is, therefore, determined b}' its position 

 in and relation to the various styles of pressure distribution. 



3. If we succeed in knowing in advance what distribution of atmos- 

 pheric pressure will prevail on a certain day or on a series of succes- 

 sive days or a longer , season, then the weather of the day or of the 

 period of time is therel)y determined in advance. 



4. The modifications introduced by reason of geographical conditions, 

 the configuration of the ground — as, for example, the location of a 

 place in the Alps, etc.- -are constant for the location in each style of 

 pressure distribution. 



By means of these theorems, which were deduced from exact obser- 

 vations, the foundation Avas laid for a careful method of weather pre- 

 diction. Two things were now necessary: (a) The perfecting of our 

 knowledge of the typical distributions of atmospheric pressure and of 



