46 



effect of the weather upon the distribution of icebergs. It is desired 

 therefore under this section devoted to weather to give a brief account 

 of the results so far of this research work. The period embraces 

 47 years, 1880-1926, a series of sufficient length to permit mathe- 

 matical correlation, and in this respect it has an advantage over 

 previous works. 



The results differ somewhat from those previously obtained by 

 Mecking in that the chief importance is assigned to the variations 

 of the pressure difference between Belle Isle, in Newfoundland, and 

 Ivigtut in southern Greenland, during the period December to March. 

 The pressure difference directly affects the amount of field ice, and 



Fig. 8a. — The atmospheric pressure map constructed by averaging the pressures for the months of 

 December to March in the years 1881, 1891, 1895, 1900, 1902, and 1917. These years were all characterized 

 by a lesser amount of Arctic ice drifting into the western North Atlantic thamisual. (See fig. 23.) 



it has been found that there is a very close relation between the 

 amount of field ice and the number of bergs south of Newfoundland. 

 The field ice tends to act as a fender along the shoreward side of the 

 Labrador current, and thus more or less prevents the bergs from 

 stranding as they are borne southward. The truth of this state- 

 ment w^as curiously revealed during the 1924 patrol, when the un- 

 usual absence of field ice left the season's crop of bergs to strand in 

 northern waters. When the sea ice recedes northward, due to 

 melting in May, the coast line becomes more and more exposed. 

 Stranding takes place on a great scale, and the consequent supply of 

 bergs to the Grand Banks is cut oft". The iceberg menace to steam- 

 ships in the North Atlantic would be greatly diminished, or prac- 



