WEATHER TYPES ON THE PACIFIC COAST. 79 



In April we had almost constant rain from the 1st to the 

 17th, then followed suddenly clear weather to the month's 

 end and after. The fact that the change from one type to 

 another is so very sudden is what causes the difficulty on 

 this coast in forecasting the weather. These phenomena, 

 as aids to forecasting, I call weather types. 



This study is only possible by reference to the reports of 

 the observations taken three times a day simultaneously at 

 4 a. m., noon and 7 p. m., Pacific time, telegraphed to San 

 Francisco and charted by entry on outline maps. Isobars 

 and isotherms are drawn showing the belts or areas of like 

 pressure and temperature, and symbols are added marking 

 stations where rain has fallen or cloudiness exists. It is 

 seen that map after map, day after day is almost identical. 

 A persistence of some one barometric characteristic covers 

 the same region. Applying the principle of composite pho- 

 tography, taking a transparent outline map of the same 

 scale as the weather map and drawing lines enclosing like 

 areas, and continuing this process on the same transparent 

 map, we have represented a great number of like areas su- 

 perimposed upon each other. 



We thus find the high or low barometer regions to cor- 

 respond with certain characteristic conditions of cloudiness 

 and rain, which remain stationary and hover over the same 

 locality during the continuance of the high or low. For 

 instance, grouping all the charts that have high pressure 

 over Oregon, and the low over southeastern California, it is 

 noticed that remarkably fine warm weather with northwest- 

 erly winds continues for a succession of days, while this con- 

 dition lasts. When the barometer changes, it does so sud- 

 denly, and the weather changes with equal rapidity. The 

 greater the number of these like features of barometer and 

 weather found, the greater, of course, is the frequency of 

 the type. Illustrating in the case of February last, it is 

 found that a persistent high overlay the district embracing 

 Oregon with parts of Nevada and Idaho. Plate II illus- 



