directly effecting density distribution are of greater relative importance 

 than in many ocean areas where the winds play a dominant part in 

 determining the currents, the wind does have an important effect in 

 modifying the Labrador Current. The magnitude of the component 

 of the goestrophic wind shown in figure 21 is not spectacular but its 

 seasonal increase is in qualitative agreement with the northeastward 

 diversion of a part of the Labrador Current and with the observed 

 behavior of late-season bergs. 



The 14 usable traces obtained from the TD and mentioned earlier 

 in the description of instruments and methods are reproduced in 

 figures 22 to 35. Superimposed on the traces are the temperatures 

 obtained with reversing thermometers attached to the wire during the 

 TD cast. The reversing thermometer temperatures are connected by 

 broken lines. The solid lines are drawn through values of o-, computed 

 from the observed salinities and the reversing thermometer tempera- 

 tures. While there has accumulated a considerable mass of data from 

 bathythermograph casts and although the BT is a rapid response 

 mstrument it is not very stable and to date has been limited in depth 

 range to about 275 meters. It is considered doubtful that the BT is 

 reliable to the desired precision of 0.1° but we are hopeful that even 

 with the experimental model of the TD used such a precision was 

 attained. Figures 22 to 35, then, represent the first reliable subsurface 

 temperature data obtained with a rapid response instrument between 

 depths of 275 and 1,500 meters and probably between the surface and 

 1,500 meters. The precision with which the TD recorded depth was 

 not as satisfactory, not because of any inherent weakness in design 

 but simply because the windings of the Schaevitz transformer were 

 mounted on a shaft with a press fit instead of being rigidly pinned and 

 the friction was not sufficient to prevent accidental rotation of the 

 windings with respect to the instrument thus producing fortuitous 

 shifts of the atmospheric pressure position of the drum carrying the 

 trace. 



The traces can be grouped as coming from four general geographical 

 areas: figures 22-24 are from the area southwest of the T^U of the 

 Banks ; figures 25-28 are from the area near the offshore corner of the 

 Bonavista triangle; figures 29 and 30 are from south of Flemish Cap; 

 and figures 31-35 are centered around the 45th parallel east of the 

 eastern slope of the Grand Banks. The first gi'oup involve multiple 

 lowerings: to approximately 100, 300, and 1,500 meters wire length in 

 the case of figure 22; to 300, 500, 1,170, and 1,596 in the case of figure 

 23; and to 300, 1,506 and 300 in the case of figure 24. Thus there 

 should be in figure 22 for example, 6 traces in the upper 100 meters, 

 4 traces from 100 to 300 meters and 2 traces from 300 to 1,500 meters. 

 In the other figures there are only the downtrace and the uptrace 

 except for figure 25 where various lowerings were made preliminary to 



67 



