162 METHODS OF FORECASTING THE WEATHER. 



pressure for the next day. When this sketch is completed then the 

 predictions for the various portions of tlie kingdom are made upon 

 the basis of our knowledge of the weather conditions at different 

 points of each area of atmospheric pressure. Thus the primary diffi- 

 culty consists in forming a correct conception of the pres.'sure distri- 

 bution for the next day, based on that prevailing on the morning of 

 the da}^ in (juestion, and at the same time a clear idea as to the velocity 

 with which the changes will proceed. In order to facilitate this diffi- 

 cult task the central office receives immediately before the making of 

 the forecast, which takes place at l.;->0 p. m., a short telegram from 

 twelve selected stations in Austria-Hungary, giving the latest infor- 

 mation as to changes in temperature, pressure, and cloudiness that 

 have occurred at these stations since the morning observatioji. From 

 this last item Ave can perceive with more certaint}' whether we have 

 formed a correct idea as to the distribution of atmospheric pressure 

 for the next day or not, and therefore whether to retain or modif}" 

 the forecast. It is only after the data of the midday telegrams have 

 been made use of that the definitive forecast is made. At 1.4.5 p. m. 

 the weather report goes to the printer, and the corresponding tele- 

 grams are sent to thosc^ who ha\e subscribed for the daily telegraphic 

 forecasts. 



The results of this s^^stem of honest weather forecasts are indeed 

 modest, but are such as to show a real and striking progress in weather 

 predictions as compared with other methods. Of course even this 

 earnest scientific method allows us only to consider the general char- 

 acteristics of the weather, as, for example, "fine," " windy," "mild," 

 "line and cold," "cloudy," "rainy," "warm," etc., as the ol)ject of 

 the weather forecast. This method would inmiediately supplant all 

 others if it would undertake to foretell the duration and amount of 

 precipitation, the degree of the thermometer, the exact force of the 

 wind, etc. However, we may at present be A^er}- well satisfied if the 

 general character of the weather is predicted for us. Unfortunately 

 even the scientific method can give us no positive certainty, since even 

 b}^ confining itself to these general characteristics it can at present 

 offer onl}^ a little above 80 per cent of verifications of the weather. 



In this state of the case it is self-evident that our efforts are to be 

 guided in the direction of those studies that will lead us to an ever 

 increasing accuracy in forecasting. These studies of course relate (1) 

 to more and more thorough investigations of the weather conditions 

 at every point and in every phase of the distribution of atmospheric 

 pressure; (2) to the discovery of signs by which to form a judgment 

 (a) as to the rapidity and paths with which each type of pressure dis- 

 tri})ution moves over Europe, (l) into what other forms a given type 

 of distribution transforms itself and the rapidity of such change, and 

 {e) what changes in the weather attend the various modifications of one 



