218 SCIENCE PROGRESS 



changes must first be examined to understand more fully those 

 fluctuations recorded by rain-gauges. 



So far as the investigations with pressure observations has 

 gone at present, there are two conclusions which I think are 

 now beyond criticism. The first is, that all over the world 

 changes of long duration are in operation ; and secondly, these 

 changes are not all alike either in intensity or time of duration. 



In spite of these marked differences, there seems nevertheless 

 to be an underlying connection between them all. It is not there- 

 fore without, but rather within, the bounds of possibility, that 

 the time will come when such apparent divergences will all be 

 found to be very closely allied to each other, and be the natural 

 resultants of one or more primary world atmospheric fluctuations. 



Just as we have been able to place before the reader a map, 

 showing a first approximate relationship between the barometric 

 changes of short duration extending over the whole of the earth's 

 surface, so during the present century it will be possible to make 

 a similar survey for those air fluctuations which require more 

 than three or four years to complete a pulsation. 



The only reason why such a chart is not available to-day 

 is that the length of time over which observations extend for 

 regions well distributed is not yet sufficiently great. 



Our appetites have, however, been wetted by numerous 

 investigations, which are quite sufficient to indicate what a 

 promising field of inquiry lies before us. 



We know, for instance, that in the Indian area there is a 

 gradual rise and fall of pressure, a complete alternation which 

 occupies about eleven years. The same area shows distinct 

 traces also of a variation which takes more than three times this 

 length of time to go through a complete swing. 



In Australia this eleven-year variation is also in operation, 

 but so far as can at present be determined, it appears to be 

 so modified that it presents a fluctuation of about nineteen 

 years' duration. 



Similar to this in length of time, but with a distinct phase 

 difference — that is, the epochs of maxima or minima do not 

 take place simultaneously— is a variation in progress in the 

 southern part of South America. There we have nearly, but 

 not quite, an opposite state of things to that occurring in 

 Australia. Our own islands are not even free from these long 

 barometric swings, for here we can trace a variation which 



