In analyzing the Internal consistency of the iceberg record, it 

 is of interest to understand how the various sections of, the record 

 were collected. The original compilation made by Mecking for 

 1880-1899 was of iceberg sighting reports. Because of duplicate 

 sightings Mecking 's counts commonly show several thousand iceberg 

 sightings each year. Smith, in contrast, tried to eliminate duplicate 

 reports, and his counts for the years 1900-1929 represent the actual 

 number of icebergs which drifted past 48°N. Smith, through comparison 

 of the two compilations with associated winter surface atmospheric 

 pressure patterns, was able to establish a correspondence between the 

 two different compilations, and published a summary table of iceberg 

 intensity on a scale of 0-10 for the period 1880-1929 (Smith, Un- 

 published notes) (Smith, 1931). After 1929, although the data were 

 collected and analyzed by different persons, it probably was done 

 rather uniformly until 1942. During 1943-1945 the International Ice 

 Patrol was suspended, and iceberg counts were based on the reports of 

 convoys and military air patrols. It is believed that the counts for 

 these years are too high (Schell, 1962). When the International Ice 

 Patrol recommenced after the war, regular aerial ice patrols replaced 

 ship reports as the primary means of collecting ice information. 

 Because of the relative regularity and more systematic coverage of 

 aerial reconnaissance, the data collected since 1946 is undoubtedly 

 the most accurate. 



The iceberg counts used in this paper are given in Table 1, and 

 were based on data presented by Smith (1931), Murray (1969), and 

 Kelly and Morgan (1970). 



The environmental parameter used in evaluating the internal 

 consistency of the iceberg record is the December-March atmospheric 

 pressure difference between Belle Isle, Newfoundland and Ivigtut, 

 Greenland. This pressure difference is an indicator of the strength 

 of northwesterly winds along the southern Labrador coast since the 

 pressure gradient it measures is approximately normal to the Labrador 

 coast in a northeasterly direction. Many authors (Mecking, 1907; 

 Smith, 1931; Soule and Challender, 1949; Schell, 1962) emphasize the 

 relationship between the strength of northwesterly winter winds and 

 the drift of icebergs south toward the Grand Banks. Smith (1931) and 

 Schell (1952,1962) made use of Newfoundland/Greenland pressure 

 differences in deriving iceberg prediction formulas. Assuming then 

 that a relationship exists between iceberg intensity and the Belle 

 Isle/ Ivigtut pressure difference, if the iceberg data have been 

 collected with relatively uniform accuracy there should be no sub- 

 stantial difference in the iceberg/pressure difference relationship 

 when it is derived for various periods between 1880 and 1969. Such 

 a derivation is made for the periods 1880-1903, 1919-1942, and 1946- 

 1969. These three periods more or less cover the periods of Mecking 's 

 data. Smith and pre-aerial patrol data, and aerial patrol data. 



16 



