16 



OBSERVATIONS AND RESULTS IN PHYSICAL OCEANOGRAPHY 



wire length and the depth increases from the surface and 

 down, meaning that the wire angle decreases. By means 

 of the wire lengths and the actual depths of the different 

 points on the wire, the curve which the wire formed at 

 the time when the water bottles were reversed has been 

 constructed and represented graphically in figure 2. 



The straight line which is entered in figure 2 shows 

 the position which the wire would have had if the wire 

 angle had remained constant. It is seen that the actual 

 depth of any given point on the wire is considerably 

 greater than the depth corresponding to a constant wire 

 angle. This fact is also evident from column 8 in table 

 1. This gives the depths which are derived by means of 

 thermometer depths and wire length and the depths which 

 are computed on the basis of a constant wire angle. In 

 the table the observed tf Tiperatures and salinities have 

 been entered. According to the values in the table, the 

 temperatures and salinities were observed at greater 

 depths than those which are obtained when assuming the 

 wire angle to be constant. The discrepancy increases 

 with depth and reaches an amount of 400 meters at a 

 depth of about 2900 meters. Representing graphically 

 the vertical distribution of the temperature, the temper- 

 ature curve is displaced upward if the depths are de- 

 rived from wire lengths and wire angle only. This ex- 

 ample can be used for illustrating the importance of ac- 

 curacy as to depth even when the vertical variation of 

 the temperature in vertical direction is small. Station 

 71 is situated within an area where the temperatures are 

 very uniform below 1800 meters. In figure 3 the tem- 

 paratures at stations 68 to 79 have been plotted against 

 depth. For station 71 double values have been entered, 

 corresponding to the adopted depths, and corresponding 

 to the depths which have been derived, assuming the 



wire angle to be constant. It is seen that the former lie 

 very nearly on the curve .vhich is derived from the ob- 

 servations at the other stations in this region whereas 

 the latter lie off this curve. 



From the table it is seen that the depths of observa- 

 tions are always less than the wire length and that the 

 difference increases with increasing wire length, reach- 

 ing a value of 382 meters at a wire length of 3345 meters. 

 These figures also demonstrate the importance of the 

 direct determination of the depths at which the tempera- 

 tures were measured and from which water samples 

 were taken. 



A compilation of the differences between wire 

 lengths and actual depths of observation has not been 

 undertaken and in the tables of results (Oceanography 

 1-3) only the actual depths have been entered. At the 

 greater number of stations these have been determined 

 accurately by means of the above method, but in some 

 instances the pressure thermometers have not functioned 

 properly and the depths are, therefore, doubtful. In the 

 tables of results special remarks are entered in each 

 such case. In this place attention shall also be called to 

 the fact that overlapping values of temperature and sa- 

 linity have been obtained at a number of stations at 

 which the greatest depth of the shallow series has been 

 selected slightly greater than the smallest depth on the 

 deep series. These overlapping values do not always 

 fall on a smooth curve. The reason may be that a time 

 change has taken place, but the reason may also be that 

 the depths are slightly in error. An inspection of the 

 temperature graphs shows that errors of + 10 meters in 

 the depth which as a rule would account for the discrep- 

 ancies and errors of this magnitude are not excluded. 



