time of about 7 ' 2 hours, were affected by internal waves and could 

 not be treated s\nopticalh- without adjustment. 



In figure 23 the departures of the dynamic heights from their average 

 value have been plotted against the times of observation and are repre- 

 sented b>' the solid line. It appeared that little would be gained by con- 

 sideration of periods longer than the total elapsed time. Of shorter 

 periods there seemed to be an undulation of the order of a 6-hour j^eriod. 

 The introduction of such an harmonic curve of 4 mm. amplitude is 

 shown as the curve of long and short dashes. Comparison of this curve 

 with the original curve shows the same characteristic departures from 

 the mean and as a consequence the general pattern of the topography- 

 would remain with gradients much the same as those in figure 22. 

 It would seem, then, that if the fluctuations in dynamic height are the 

 result of internal waves, the waves of importance in the present instance 

 are of short enough period to be of the order of magnitude of the time 

 interval between stations and thus to require the conclusion that 

 the series of observation is not sulTficienth' great in number nor closeh' 

 enough spaced to lend itself to harmonic anahsis with promise of worth- 

 while results. 



The possible causes of the variation among the 12 stations are not 

 limited to errors in temperature and salinity and to the effect of internal 

 waves. Figure 20 suggests the possibility of spatial differences. The 

 results have been discussed thus far with the location of the observations 

 referred to the berg on the assumption that if the berg were in motion 

 with respect to bottom the motion would be that of the current and that 

 there would be little relative motion between the berg and the surround- 

 ing water. If, however, one assumes an appreciable effect of the wind 

 on the exposed portion of the berg in contributing to its motion, the 

 conditions which existed were favorable for the berg to move across 

 current lines. The wind of force 3 to 5 was blowing from 187° true. If 

 the berg's movement with resepct to bottom be assumed to have been in 

 a direction of 052° at 3^ knot and the direction of the Labrador Current 

 135°, the corollary conditions are reasonable and are shown in the small 

 diagram in figure 24. 



In this figure and in figure 25 the locations of the stations have been 

 plotted with respect to bottom on the above assumptions which include 

 a berg movement of \i knot in a direction 052°. The dynamic topog- 

 raphy of the surface has been redrawn in figure 24 to accord with these 

 station locations. The general circulation pattern and the maximum 

 gradient are much the same as those features shown in figure 22. Thus 

 the assumption of berg motion does little to explain the indicated 

 relative dynamic topography. If the assumption of a general south- 

 easterly flow of the Labrador Current is maintained the topographic 

 pattern at the 140-decibar level must have a clockwise eddy beneath 

 the edd>' shown in figure 24, and this clockwise eddy must have equal 

 and opposite gradient components in a northwest-southeast direction 



29 



