the first of the rules indicating 1945 and rule 4 indicating 1944 

 as a high-pressure year. Evidently better can be achieved. 

 However, leaving room for the great number of minor influences 

 which may have disturbed the primary variation, it is 

 astonishing to see how few discrepancies between the actual 

 and the purety constructed curves remain which obviously 

 call for special explanations. Such are, for instance, the 

 eruption of Krakatoa in 1883 and Mount Katmai in 1912, 

 which have caused noticeable disturbances in the more regular 

 variation of air-pressure. 



The above supports my view that this outstanding long-periodic 

 terrestrial oscillation of nearly three years is almost completely controlled 

 by solar action. This is an important conclusion as it may open the way 

 towards an understanding of comparable pseudo-periodic variations in 

 other parts of the world which so far have remained one of the most 

 intriguing puzzles of the seasonal forecaster. 



Rules 3 and 4 prove that the wave-length is stretched and compressed 

 between two and three years. Evidently the fundamental oscillation 

 would present a period between two and three years in undisturbed 

 conditions. This is in accordance with theory. Finally, we have to 

 answer the question whj^ harmonic analysis emphasizes, as was shown 

 by H. J. de Boer, a mean period of 3-4 years — the same period that was 

 found by H. P. Berlage from tree-ring measurements back over four 

 centuries — although it has no physical reality. There is every reason 

 to assume that the shortest waves lose their individuality in this operation, 

 two crests or depressions coming in ^'ery quick succession simply being 

 counted as one. Different facts thus lead to one point of view. The 

 above rules of interaction are so simple that we may hope to be able even 

 before long to interpret them together by some physical principle of 

 interaction between earth and sun. 



Discussion 



At the conclusion of the paper an interesting discussion followed, one 

 of the chief contributors being Mr. C. H. B. Priestley, who remarked 

 that Dr. Berlage was the first one to have given a physical explanation 

 invoking the time-lag between air-pressure and sea-temperature 

 fluctuations for the existence of the three-year southern oscillation. 



In answer to a question by Mr. MacKenzie about the periodic devia- 

 tions of the Humbolt current from the Peruvian coast, Dr. Berlage 

 expressed the view that the seven-year period simply resulted from a 

 doubling of the more fundamental 3-4-year period. 



In reply to a question by Dr. Andrew Thomson, Dr. Berlage said that 

 there were very few observations of sea temperature along a line from 

 Chile to Indonesia, so that a check on the theory could not in this case 

 be obtained. 



Mr. Simpson questioned the explanation of the time-lag given by Dr. 

 Berlage and asked whether it could not be the eftect of local sea-water 

 circulations. 



In reply to a question by Mr. Hutchings, Dr. Berlage said that the 

 fluctuations of temperature were of the right order of magnitude to 

 account for the changes of pressure, but no calculations had been made 

 as to the precise amount. 



