116 



near the edge of the shelf. Because of other considerations, however, 

 this section was located near the middle of the shelf. No canyons 

 were found to intersect this section but because of its location all that 

 can be said is that if canyons do exist on the Labrador shelf they 

 do not extend as far as the line of this section. Other incidental 

 data collected on the General Greene, both during the season and 

 post-season cruises, consisted of barograms and sea-water surface 

 thermograms. 



The equipment and methods of measurements were the same as in 

 1937. Routine salinity bridge measurements and calibration titra- 

 tions were performed by the oceanographer's assistants, Bernard S. 

 Loebig, yeoman, first-class, and James F. Flannery, seaman, first-class. 

 Other assistants in the observational work were Thomas C. Crabe, 

 yeoman, third-class, and William George, seaman, first-class. 



At the 268 stations occupied during the season the extent and 

 nominal depths of the serial observations of temperature and salinity 

 were the same as during the season of 1937 except that at the stations 

 comprising the section across the Gulf Stream occupied during the 

 third survey the observations extended to a deptli of between 2,000 

 and 2,500 meters. On the post-season cruise the temperature and 

 salinity measurements extended to as near bottom as was practical at 

 most of the stations. In general the oxygen samples were taken at 

 depths from 100 meters down and the helium samples only in the 

 deeper layers, from 400 meters down. 



As a check on the fimctioning of the reversing thermometers, a pro- 

 gram of intercomparison through periodic shifts in thermometer pairs 

 was carried out as in previous years, A total of 1 ,735 individual com- 

 parisons were investigated. The probable differences between the 

 corrected readings of a pair of thermometers varied among the dif- 

 ferent groups from 0.003° to 0.017° C. and the probable difference for 

 the entire 1,735 comparisons was 0.010° C. An additional 267 com- 

 parisons were made in which one thermometer of the pair was unre- 

 liable. The probable difference for these 267 comparisons was 0.034°C. 

 However, as the readings of these thermometers were known to be 

 unreliable they were not used in determining the observed tempera- 

 tures which are considered to be accurate to 0.01° C. 



Near the end of the season a progressive error in the standardization 

 of the salinity bridge was suspected. This was later verified by the 

 careful titration of 38 actual sea-water samples from various depths. 

 The tabulated values of salinity have been corrected for this error, the 

 original measurements before correction being too low by the following 

 amounts: Second survey 0.019%o; third survey, 0.029%o; fourth 

 survey and post-season cruise, 0.054%o. As these corrections are 

 equivalent to differences in dynamic height of 14, 21, and 39 milli- 

 meters when referred to the 1,000-decibar surface and are, for practical 



