50 



A more accurate method of measuring the value of the meteor- 

 ological factor was found by our investigation to be an equation 



Bergs = 4.8-0.08(c)-0.12((?) 



where the number of bergs is expressed as a value 0-10; c represents 

 the pressure difference in millibars between Belle Isle and Ivigtut; 

 d represents the pressure difference between Stykkisholm and Bergen. 

 (For details of the values see Bull. No. 15, p. 48.) 



Fig. 18. — March, 1927, anomaly of atmospheric pressure. Conditions 

 when reflected in the spring spell more ice than normal 



A forecast was made in a letter, March 11, 1927, to the chairman 

 of ice patrol board, that about 396 bergs would drift south of New- 

 foundland during 1927, or field ice and bergs somewhat in excess of 

 that during a normal year. As a matter of fact we now know after 

 a careful check that the number of bergs for the danger season, 

 March to July, was 367, and about 390 for the entire year, and thus 

 the forecast was quite accurate. (See iceberg table, p. 16.) The 

 number of bergs may vary within wide hmits; for example, in 1912 

 there were 1,200 and in 1924 only 11. The forecast in 1927 was 'pur- 

 posely kept from publication because it is desired to test the work 

 with a few more trials, but it all is most encouraging and the service 

 will be continued for 1928. 



