cover the whole time of the survey or whether tlie change was much 

 more rapid. Figure 23 has been constructed assuming a gradual 

 change and the error in closure of the triangle distributed amongst 

 the shallow water stations according to distance assuming the dis- 

 tance to be proportional to time. Figure 24 has been constructed 

 assuming the change took place rapidly from one steady state to 

 another and occurred between the occupation of stations 4147 and 

 4148. Thus the similarity in current pattern at the two levels is 

 really better than indicated by the two figures. 



The interval between stations 4147 and 4148 has been selected as 

 that during which the rapid change took place, because of the unusual 

 behavior of the von Arx current meter during this interval. It 

 recorded very swift currents, banded but uniformly east-southeasterly 

 in direction, with a maximum speed of the order of 5 knots. Using 

 the recorded currents, the integral of the elemental products of current 

 and distance between stations 4147 and 4148, expressed as an equiva- 

 lent difference in dynamic height between the two stations, amounted 

 to about 33 dynamic centimeters. It is evident from the magnitudes 

 involved that, if the current meter were not in the midst of some 

 aberration, we were not dealing with a steady state but with a transient 

 condition which probably included a peiiodic movement (such as a 

 tidal current) combined with a sizable change from one steady state 

 to another. The transition between two steady states involves 

 accelerations whereas the formulae developed from the Bjerknes 

 circulation theorem deal only with steady states in which the accelera- 

 tions are negligible. 



It would seem from these measurements that changes in the current 

 pattern do take place in this region and that significant changes in 

 dynamic height may occur in an interval of less than 3 days and possi- 

 bly in as short a time as a few hours. It is also suggested that very 

 swift currents, such as were recorded by the von Arx current meter, 

 may be produced locally in the readjustment which takes place in the 

 change from one steady state to another. As to the possible causes 

 of such a change, it has been noted that for several days prior to this 

 survey the winds had been steadily from the northeast quadrant. 

 More detailed comment would lead at once to pure conjectui-e. 



Returning to (lie speculation that the change from one steady state 

 to another steady state took place rapidly and most of it occurred 

 between the occupation of stations 4147 and 4148, the dynamic height 

 originally comi)utcd would then be valid for all stations except 4145 

 through 4147 and tiie dynamic topogra})hy would then indicate that 

 nearly all the beigs passing Cape Bonavista would be carried into the 

 eastern branch of the Labradoi- Cui-rent and only those passing the 

 49th parallel west of about 52°25' W., would strand on the northern 

 cxh^o of the Giand Banks oi- follow the western branch along the 

 Avalon Peninsula. The subsequent drift of late season bergs lends 



80 



