93 



engender confidence, must be verified by the event in a substantial 

 majority of cases. 



These difficulties urite to make this a field in which fertile 

 results may be soonest expeoted from the "case system" of investiga- 

 tion, while the extieme complexity of the basic problem makes it 

 essential that the simplest cases be the first attacked, thus approach- 

 ing as nearly as possible to the laboratory method. Furthermore, 

 the impossibility (if we are to be intellectually honest) of promis- 

 ing direct economical benefits therefrom, m.akes research institutions 

 particularly appropriate centers for certain aspects of such work, in 

 cooperation with the governmental weather bureaux. The very encour- 

 aging progress that has been made in the experiment now being carried 

 out by the Scripps' Institution corroborates this view. 



The results of twelA'^e years" work there, to date, appear to show 

 that in that region a useful correlacion does exist between oceanic 

 conditions in the offing, and the weather ashor:, for (over this 

 brief period of years) when the sea surface near southern California 

 has been cooler than normal from .\ugust to October, but the mid- 

 Pacific warmer, the rainfall of southern California hes been greater 

 than usual during the following winter; and vice- versa . attempts to 

 predict the amount of rainfall have been about 75fc verified. Thus it 

 appears at present that, for southern California at least, tempera- 

 ture departures in the various parts of the Pacific are one of the 

 classes of indicators that can be combined into cumulative forecasts 

 of seasonal rainfall and perhaps of temperatures. 



Much work yet rem.ains to oe done to uncover the effect of other 

 factors that are undoubtedly concerned, and to place the system on an 

 assured basis. But the suggestive results of this attempt, to date, 

 not only justify the continuation of this line of work in southern 

 California, for which the Scripps' Institution has plans, but point 

 the need of investigations of the same sort in other representative 

 regions chosen on the basis just stated (Page ). The relationship 

 that rainfall in Ecuador and northern Peru bears to ocean tempera- 

 tures off that coast offers a very promising case for study. Other 

 American vantage points that seem favorable, because interpretation 

 promises less difficulty there than m m.ost parts of the world, 

 appear to be Northeastern Brazil, British Colum.bia, Southern Alaska, 

 and the Gulf coast and south Atlantic Soeboard of the United States. 



It is obvious that efforts to work up the great mass of ocean 

 Temperatures already accumulated at several places would be an 

 essential item in any broad-scale research in this general field. 

 And all institutions so doing, whether governmental or private should 

 >e encouraged to follow a comimon plan. 



All such data should also be published promptly, in order to be 

 generally availaole, again according to some general plan. 



