104 



BJ0RN HELLAND HANSEN 



[rep. of the "MICHAEL SARS" NORTH 



regarded as stream-lines, we can easily find the vertical 

 variations of the velocity itself when we know the angle 

 which the direction between the stations forms with the 

 direction of the isobaths. This angle being /S we have: 



t'o ~Vp^^ — v' pi sin /*, 

 where Vo is the average velocity of the current at the sur- 

 face between the stations and Vp at the isobaric surface 

 for p decibars, i. e. practically at a depth of m ordinary 

 metres when tn has the same numerical value as p. In 

 order to ascertain the direction of the isobaths is it desir- 

 able to have a number of stations scattered over the area 

 to be examined. An arrangement of the stations along 

 one sectional line is unfit for the purpose. Between Stats. 

 66 and 68 the isobaths obviously go in a direction almost 

 normal to the line between the stations, /. e. /S is a wide 

 angle and the velocity-components computed are nearly 

 as large as the actual velocities. The line drawn from 

 Stat. 85 to" Stat. 86 (curve IV) evidently forms a small 

 angle with the direction of the isobaths (the current) and 

 the components are small in relation to the actual veloci- 

 ties. We should certainly have found much greater values 

 if the stations had been worked straight across the current 

 and not more or less along it. 



Only very few of the "Michael Sars" stations were 

 worked to such depths that we have observations past 

 the lower boundary of the current. Even if the velocities 

 are often small at about a thousand metres below the 

 surface they are in most cases not negligible there. For 

 instance between the stations 66 and 68 the difference of 

 the velocity-components at about 1000 and 1400 metres 

 is about 4 centimetres per second. In the eastern North 

 Atlantic a sufficient number of stations down to 2000 

 metres or more has been worked, so we have been able to 

 construct more reliable charts of the currents here than 

 we can for other parts of the North Atlantic [Helland- 

 Hansen and Nansen, 1926]. 



We need not enter into further detail regarding the 

 numerical calculations of the currents. They are easily 

 made by means of the equations above and the data 

 given in Tables III and IV in Part II. We have here made 

 the calculations by means of equation (f) in section 42; 

 they may also be done by means of equation (e). The 

 numerical results will be a little different in the latter case, 

 because there is a difference between the depth of the 

 standard level surfaces and the standard isobaric surfaces. 



X. CURRENT MEASUREMENTS. 



44. Methods. 



In 1906 a great many current measurements were 

 made in the deep Norwegian fjords, in the Norwegian 

 Sea and the North Sea [Helland-Hansen, 1907]. A rowing- 

 boat was tightly anchored fore and aft with heavy grap- 

 nels and hemp-lines. In the fjords we could obtain very 

 sharp cross-bearings whereby the shifting of the boat's 

 position within a metre or two could sometimes be 

 discovered. The boat kept its position so well that its drift 

 had no effect on the observations of the currents even 

 if the depth to the bottom was as much as 500 metres. 



It was part of the program for the hydrographic in- 

 vestigations on the "Michael Sars" Expedition to have 

 current measurements made in different localities. We 

 intented to use one of our rowing-boats with double 

 anchoring in places where the bottom-depth was not 

 great. The first experiments of this kind were made at 

 Siat. 18 in the Straits of Gibraltar, where a few obser- 

 vations were secured from the rowing-boat. The cur- 

 rent, however, proved to be so strong that the hemp- 

 lines broke again and again, and the plan had to be 



given up. We had not the necessary equipment for re- 

 placing the hemp-lines by steel wires, as we have had 

 when making more recent observations of the currents 

 in the open sea, for instance on the "Armauer Hansen" 

 Expedition to the North Atlantic in 1913. 



Having been forced to give up the work from the 

 small rowing-boat we tried to anchor the steamer by 

 dropping one of the ship's anchors attached to the thick 

 trawl-wire. In the strong and fairly regularly veering 

 current the vessel lay steady enough to allow fairly re- 

 liable measurements to be made. The stout steel wire 

 served to check the ship's movements. In order to obtain 

 an idea of what the latter amounted to the ship's com- 

 pass was read at short intervals. These compass readings 

 did not vary much, even over a considerable length of 

 time. The ship mostly remained lying in one direction, 

 at least during the time needed for carrying out each 

 separate observation. Of course the compass readings 

 do not enable us to judge of the distance which the 

 vessel may have drifted. Naturally the sheering of the 

 ship to or from the anchor does not affect the bearing. 

 Even if a swing of the vessel is immediately indicated 



