93 



A.lthough surface-terr.peratarss al.nost past counting have been 

 collected in the ppst, it has been rppreciated for many years that 

 one of the difficultiss of such investigation lies in the need for 

 gathering reliable observations at shorter intervals, for various 

 parts of the ocean, for only by such data would it be possible to 

 follow, in detail, just v;ha"t changes do occur in the sea. 



It is pertinent here to consider how far the machinery that 

 would be necessary for analytical investigation in tali- field now 

 exists. So far as physical eauipment roes, the answer would be 

 encouraging for the North Atlantic where steamers are regularly on so 

 many routes that a close net of continuous Oc-anogrsphic data could 

 be obtained easily, if thermographs, barographs, etc, could be 

 installed on a sufficient number of ships and if arrangements could 

 be made for the ships' officers to give these instruments the needed 

 attention; also to care for the records. In fact, continuous sea 

 water thermographs have already been installed on steamers running 

 in various parts of the world under the observations of several 

 different institutions with highly instructive results. The hydro- 

 graphic services also receive a continuous stream of observations 

 from a variety of sources, and the weather bureaux are now developing 

 a scheme of coordinated investigation as noted on Page , In the 

 other oceans data are much needed from regions that lie outside the 

 resrular steamship tracks, hence, cannot be obtained without special 

 arrangement. 



The most serious obstacle to the advance of knowledge as to the 

 general relationship between sea temperatures on one hand, and 

 atmospheric temperatures and pressures on the other has not been any 

 intrinsic difficulty in obtaining the marine observations, but the 

 inability of any existing agency to undertake analysis of the enormous 

 mass of data that has already been amassed, and that will continue to 

 accumulate at an appalling rate if continuous observations are taken 

 on many ships running along as many different routes. For such 

 investigation to be of any practical value whatever, this analysis is 

 essential. I^nd it is necessary to face not only the volume of work 

 entailed, but also its extreme complexity. 



Thr magnitude of such an undertaking, if it were to be applied to 

 any one of the ocean basins as a whole with the fringing lands, is 

 quite beyond the capabilities of any private institution now exist- 

 ing- or likely to be established. At present it is equally bayond the 

 reach of any' of the governmental weather services. The United States 

 Weather Bureau is now appealing to Congress for funds for the task of 

 compiling and analyzing its ocean temp-rature records. But as meteor- 

 ologists", as a body, cannot promise the legislatures th^t such 

 analysis (even if continued for ten or twenty years) will produce 

 commensurate economical results, it is not likely that governmental 

 funds can be secured for large-scale investigations of this sort. 

 Furthermore, there could be no attempt at official long-range weather 

 forecasting based on sea temperatures (except perhaps for some local- 

 ity especially favorable) until a rational basis for prediction be 

 established by the proof thst a correlation exists; until a sound 

 method for translatin:: such correlations into terms of weather be 

 found; and until arrang;:.mcnts be made for the regular collection of 

 the necessary data. 2ven assuming these requirements to be met, 

 official forecasts could hardly be given cut until the methods had 

 been tried out for a long term of years, because such forecasts to 



