TRANSACTIONS OF SECTION E. 715 



2. All Experiine7it on the Melting of Ice in Salt Water. 

 By J. W. Sandstrom. 



At the request of Professor Pettersson the author repeated one of the ice- 

 melting experiments mentioned in the preceding paper on a larger scale in the 

 Central Laboratory for Oceanic Research at Ohristiania, Professor Nansen kindly- 

 placing at his disposal the large tank, 3-5 metres in length by 0-4 metre in height 

 and 0-7 metre broad, which had served in his own investigations. This tank was 

 filled to a depth of 35 cm. with salt water at7°0. and 30 per cent, salinity. 

 The tank was divided into two compartments by a partition 30 cm. high. Above 

 this wall there was fi-ee communication between the water in the two com- 

 partments. 



In the right compartment (which represents the Norwegian Sea) a rectangular 

 ice block 50 kilogrammes in weight was introduced. While melting it assumed 

 a very peculiar shape. 



In the left compartment (Atlantic) an extra current was set going, carrying 

 twelve litres per hour of water of the same salinity and temperature in and out 

 of the vessel at a constant level. The movement of the water in the other 

 compartment was studied by means of fine jets of KaMuO^ solution which were 

 made to ascend from capillary tubes at the bottom. The deflection from the 

 vertical of these jets was measured every ten minutes. Thus a good estimate of 

 the velocity at different levels was obtained. 



Before the introduction of the ice block in the right compartment the jets 

 indicated an almost entire absence of movement of the waters of this compartment, 

 although the extra current was circulating constantly in the other (Atlantic). 

 With the introduction of the ice the conditions changed. Three different currents 

 could be discerned to the right of the partitiou wall, which represented the Faroe- 

 Shetland or Faroe-Iceland ridge. From the left (Atlantic) an under-current was 

 attracted towards the under-side of the ice, where it reached its maximum 

 velocity of 023 cm. per second. From the farthermost side of the ice block, where 

 the under-current impinged, cold diluted water arose to the surface and formed an 

 outgoing ' polar ' current the maximum velocity of which was found to be 

 0-0.3 cm. per sec. From the ice, cold and dense water descended to the bottom, 

 where it formed a powerful outgoing current with a maximum velocity of 0T4 cm. 

 per sec. The bottom current had an upward tendency near the partition wall, 

 over which it flowed into the Atlantic compartment. All measurements were 

 made after the conditions had become established, which was found to be the case 

 after two hours. The volume of water carried by the under-current towards the 

 ice was about twenty-seven times the volume carried by the returning bottom 

 current, and about twelve times greater than that of the surface current.' 



By observation of the motion of small particles of insoluble matter held in 

 suspension by the water the current lines in the middle vertical section of the 

 tank were constructed. 



Accurate observations of salinity and temperature at all depths were made by 

 the chemist of the Swedish Hydrographical-Biological Commission (Miss Palm- 

 quist) and the density in situ of the water computed from Knudsen's tables. 

 From these data the accelerating forces of the circulation were calculated according 

 to V. Bjerknes, ' Ueber einen hydrodynamischen Fundamentalsatz und seine 

 Anwendung besonders auf die Mechanik der Atmosphare und des Weltmeeres.' ^ 

 The result was to show that the accelerating forces counteracted the circulation 

 in the part of the right compartment most remote from the ice, while they aided 

 the circulation in the neighbourhood of the ice. The aiding forces ( = 0'57 c.g.s. 

 solenoids), however, exceeded those counteracting ( = 0'18 c.g.s. solenoids), the 

 difference ( = 039 c.g.s. solenoids) being the resultant of two acting forces. 



' This is according to measurements taken in the central longitudinal section of 

 the tank. At the sides the motion is more retarded by friction, which accounts for 

 the fact that the volume of the ingoing and outflowing water was not found to be 

 equal, 2 K. D. V. Aks. Ilaiidl. Bd. 31, 1898. 



