366 



RECORD OF SCIENCE FOR 1887 AND li 



the following table, which gives the iiiimber of charts that show such 

 minima: 



Only the first of these series, depending upon twelve years of records, 

 can have any pretense of accurately presenting the normal annual rate. 

 The remarkable double maximum of frequency with a diminution in the 

 middle of winter will probably remain even in long series of observa- 

 tions, and has its origin in the development of the continental area of 

 high pressure in January and February. The small number of depres- 

 sions under 29 inches in the Pacific Ocean is doubtless in part due to 

 the insufficient data from this ocean for the first year, but Loomis shows 

 that even in the last three years (summer of 1881 to the summer ot 

 1884), where the number of observations was nearly sufficient, only 

 nineteen cases below 736 millimeters occurred in the year, whilst on the 

 Atlantic Ocean twenty-eight cases occurred annually below 725 milli- 

 meters on the average of four years ; in the Pacific Ocean the barome- 

 ter sank below this latter limit only five times in three years, so that 

 such deep depressions occur sixteen times more frequently over the At- 

 lantic than over the Pacific Ocean. This result is confirmed by another 

 consideration: On the average of five years' observations at Stykkis- 

 holm, in Iceland, the barometer fell below 725 millimeters on G.8 days 

 annually, but in three years' observations in the Aleutian Islands only 

 on one day annually. The lowest depression reported from the Pacific 

 Ocean during seven years is 719 millimeters, but in the North Atlantic 

 such depressions occur on the average about thirty times annually. 



As concerns the location of these barometric minima, Loomis finds 

 that of one hundred and thirty-one in the first {a) series, one hundred 

 and twelve occurred on the Atlantic coast, and only nineteen in the in- 

 terior of North America ; that, therefore, the neighborhood of the ocean 

 is an almost indispensable condition for these deep depressions. He 

 presents the location of the one hundred and thirty-seven minima of 

 series (h) on a chart that shows that three fourths of the total number 

 occurred on the Atlantic Ocean or its coast, and that of the remaining 

 cases ten occurred within 100 miles distant from the coast; seven at 

 100 to 150 miles; and only one, January 15, 1881, at Moscow, occurred 

 more than 150 miles from the coast. Within the area of the ocean more 



