89 



It has often been suggested, however, that at least a partial 

 basis for such a sequence could be found in the sporadic variations 

 that are known to take place in the surface temperature in various 

 parts of the sea, combined with any corresponding expansions or 

 contractions of the ocean currents, and with the rate of evaporation 

 from the surface. This suggestion has been removed from the realm 

 of purely theoretical potentiality to the stage of actual test by the 

 comparisons between the physical state of the sea water and the local 

 weather that are now being carried on, esptcially in thj North 

 A-tlantic, in Canada, in California} and in Java, For example, marine 

 temperatures are now being used in an attempt to determine whether 

 the Weather in the South /Atlantic states or in Europe shows dependence 

 on conditions in the Gulf Stream. /^nd predictions of the weather of 

 southern California developed at the Scripps Institution from the 

 temperature of the adjacent sea during the preceding months (as an 

 index of the strength and permanence of the north Pacific high) have 

 been verified to an encouraging degree for the past twelve y.iars. 

 H'-cent investigations also show a sequence in tem.peratures of pressure 

 and temperatures across the Pacific Ocean, extending over some months, 

 which suggest the effects of a transportation of heat by ocean 

 currents. 



It is obvious that studies of this sort, if looking toward 

 weather prediction, presuppose the occurrence of longer or shorter 

 term fluctuations of temperature in the sea, of a sort that cannot be 

 described as regularly "seasonal". i^nd as pointed out on page ., this 

 supposition is justified, variations of this sort having been ooserved 

 so freauently that they must be accepted as characteristic of every 

 part of the sea where the temp-^rature has been studied in detail. 

 But before the claim that these events can be used as a basis for 

 weather prediction can be upheld, it is necessary to establish, not 

 only that a regular correlation exists between the two classes of 

 phenomena for th- parts of the earth in question, but that the 

 changes in the sea regul-rly antedate the changes in the atmosphere, 

 and not the reverse; also whether the former are so great that their 

 effects are not entirely masked by the complex atmospheric phenomena 

 that immediately control the weather. 



This quantitative aspect of the problem is especially pressing, 

 because meteorologists and oceanographers have to do here mostly with 

 minor fluctuations in the th':rmal state of the sea, seldom with 

 major alterations of a sort that would strikingly be reflected in 

 the weath-r of some part of the v/orld, such as the heavy rains over 

 parts of the Peruvign d-Esert early in 1925, or the droughty and other 

 consequijnces of unusual outbursts of polsr ice. While th^se minor 

 fluctuations are known to occur commonly, little is known about them 

 except in the marginal seas in high latitudes (where they may be 

 expected to reach their widest range), A.nd while a progressive move- 

 ment of such temperature abnorm-alities as develop m.ay be expected to 

 take place along the tracks of the major ocean currents, precise 

 information on this point is much needed. 



In the northern hemisphere, for example, easterly movements of 

 this sort have, for the most part, been traced in high latitudes 

 north of the 40th parallel. But this may partly be because the 

 temperature abnormalities so far actually recorded (not surmised) 

 have been much greater in high latitudes than in low, allowing their 

 progression to be followed more certainly. To illustrate the 



