58 MARION EXPEDITIOX TO DAVIS STltAIT AND BAFFIX BAY 



changes in the intensity of the North Athuitic atmospheric circula- 

 tion, coming to the conchision that a weak circnhition of air from 

 Aiignst to February is followed by a relatively small amount of 

 pack ice off Newfoundland during the succeeding spring. After a 

 strong circulation during the fall and early winter months more 

 pack than usual spreads over the Grand Bank. His studies are 

 especially important because they are based on a series of yeai-s, 

 running back to 1S60. The scale is +'2 for a year of much ice. and 

 -2 for little ice, a closer comparison probably being impossilile due 

 to the nature of the material for the early years. 



flecking (1906 and 1007) has published two important papers as 

 a result of a carefid study of ])ack-ice conditions in the northwestern 

 North Atlantic, and an investigation into the possible causes of the 

 annual variations for the period of 1880 to 1900. His data are 

 obtained from the following sources : United States Army Signal 

 Service, United States Hydi'ogra]ihic Office. United States Weather | 

 Bureau, and Deutsche Seewarte. 



The records of these offices consist of the rei)orts of ice sigiited bv 

 trans-Atlantic ships on their regular voyages through the ice regions 

 off Newfoundland. The adoption of prescribed lane routes acros.> 

 the Atlantic in 18T5, their modihcation in 1898, and the present 

 method of seasonally shifting the tracks wdienever ice conditions are ' 

 :i serious menace, are all modifying factors which must be given due , 

 consideration in arriving at an accurate ice record over a period of ' 

 many years. The fact that many reports often refer to the same ' 

 Held or floe may result in duplication and so caution is needed for a i 

 correct compilation. Mecking has conclusively shown that the factor 

 chiefly controlling the variations in the limits of the spring pack ice 

 in the northwestern North Atlantic is the barometric gradient dur- _ 

 ing the previous winter across the ice stream in the vicinity of the 

 Labrador coastal shelf. The assumption is that favorable winds 

 and currents during the colder months of the year over Labrador 

 will drive more pack ice than normal past Newfoundland in the 

 following spring. The agreement between the values of the iee ^ 

 curves and the pressure gradients on his graphs is close. The sprint: i 

 of 1887 was, however, an exception when a great quantity of pack^ 

 appeared off' Newfoundland, although the ])ressure gradient had aver- 

 aged weak during the preceding winter. This inconsistency. Meckiui: : 

 thinks, Avas due to the scarcity of icebergs which normally tend to 

 break up the pack ice, allowing it to drift freely. Also the year of, 

 1889 was ])eculiar in that j)ractically no pack ice drifted south of, 

 Newfoundland during the spring despite a favoi-able pressure gradi- 

 ent. INIecking attributes the inconsistency to the extremely warm 

 sunnner of 18.S.S which melted so much ice as to produce a deticienov 

 in the following spring. And in this respect this is the only case 

 recoi'ded when the temperature in one sunnner was noticed in the 

 crop of i^ack ice the succeeding spring. 



Schott (1904, p. 305), with the aid of ship reports contained in 

 the files of the Deutsche Seewarte. icviewed the period 188(^-1891. 

 comparing each year of the series with icgard to (juantity of pju'k 

 ice off' Newfoundland. He agrees with Mecking that 1889 was an 

 umisual year, but lie points out that the pack appeared in Septem- 

 ber, to remain for the balance of the vear. 



