44 



days previous to the taking of the observations. It continued to do 

 do so for several days more. 



The cold water northwest of the berg can be accounted for by the 

 fact that cold Labrador current water existed a short distance from 

 the berg in that direction. It is not likely that any considerable 

 portion of the observed temperature differences can be attributed to 

 the influence of the berg. The temperature wall was well marked for 

 several hundred miles and about the same sort of local variations of 

 temperature could doubtless have been found along it at many 

 places 100 miles or more removed from the nearest ice. 



In the case of the above berg it was noted that the current was 

 swifter in the warm water just south of the berg than in the cold water 

 just north of it. It has very often been observed that when the patrol 

 stops to drift in the warm water south of a berg that is on the temper- 

 ature wall, the ship passes the berg, even though it be calm. On the 

 other hand, when the ship stops in the cold water to the north of the 

 temperature wall, it is usual for the berg to drift to the eastward 

 faster than the ship. 



This year several fragments of glacial ice were picked up by pulling 

 boats from around icebergs for the purpose of preserving them and 

 determining the specific gravity of this sort of ice. The samples were 

 kept in the patrol ships' refrigerating compartments and a very few 

 of them were delivered in apparent good condition to the Jefferson 

 Physical Laboratory at Harvard University. 



At the request of the United States Coast Guard, the Jefferson 

 Physical Laboratory made careful tests of pieces of ice taken from 

 various parts of the specimens. The average density of the iceberg^ 

 ice was found to be 0.8994. This figure is considerably higher than 

 the figure assumed on page 104 of Coast Guard Bulletin 18, though 

 slightly lower than that for air-free fresh-water ice. 



It is hoped that further tests of the density of ice from different 

 bergs can be made in the future. It would be unwise to base conclu- 

 sions on the results obtained to date, because of the possibility that 

 the few samples tested may not have been truly representative of the 

 average sort of iceberg ice. 



