714 KEPORT— 1903. 



its contact with the ice into ' Arctic ' water.^ One part of this Arctic water 

 consists of Atlantic water diluted with ice-water. This kind of water rises to the 

 surface and contributes to the maintenance of the polar currents. The other and 

 greater part consists of Atlantic water, which has given up its surplus of heat 

 and sunk to the bottom, there to form the great cold bottom layer of the oceans. 



4. The warm undercurrents always follow the trend of the deepest isobathic 

 lines, while the ice-currents only exist over shallow parts of the sea. As soon as 

 an ice-current leaves the coast-bank and takes its way over a deep part of the 

 ocean, its ice is e.tposed to melting by the warm undercurrent, which is attracted 

 by the ice and maintained by the energy set free at the melting process. 



5. The above-mentioned chart shows a remarkable example of how north-going 

 warm currents and undercurrents are deviated to the west in the Norwegian Sea, 

 in spite of the powerful influence of the earth's rotation. 



6. The metamorphosis of Atlantic water into Arctic water involves also a 

 biological change. The foraminifers, &c., of the Atlantic die out (as already shown 

 by Sir J. Murray) in contact with the cold water and sink to the bottom, there to 

 form calcareous deposits. Consequently the course of the warm undercurrents 

 can be traced up to the highest latitudes by a surplus of CaO in the bottom 

 sediments. 



7. The author's experiments with exact measurements in the Skagerrack and 

 the Baltic show that the motion of the deeper layers there is by no means 

 insignihcant or slow, as it is judged to be by the advocates of the 'wind-theory,' 

 but is, as a rule, stronger than that of the surface water. How far this holds with 

 regard to the deeper parts of the oceans remains to be investigated. Therefore 

 current measurements in the Atlantic at depths of 800-4000 m. are a pressing 

 desideratum in oceanography. 



8. As the accumulation of polar ice varies with the season, and is influenced 

 by terrestrial (meteorological) as well as cosmical phenomena (radiation, &c.), it 

 is evident that the current system set in action by the cycle of latent heat must 

 show periodical variations with the seasons and also periodical or non-periodical 

 variations of longer duration. Are there any indications of such variations in the 

 movement of the undermost layers of the sea ? The author's experience is that there 

 are such indications, and this discovery has led him to propound the present theory. 



9. The Antarctic Ocean presents the grandest example of ice-melting and of 

 variations in ice-melting. It must be borne in mind that the energy liberated by 

 ice-melting in the ocean is proportional to the depth of the submerged part of the 

 ice. From an iceberg .500 m. in depth the melting of one kilogram of ice will 

 produce an amount of work equal to 7 kilogram-metres. Great ' outbursts ' of 

 icebergs from the Antarctic are known to happen from time to time (Eussell). 

 Such outbursts, which carry icebergs down to low latitudes in the Indian Ocean, 

 may exercise influence upon the climate of India, Australia, &c.,as thereby part of 

 the warm area of the ocean from which the water evaporates, which ultimately 

 falls as monsoon rain upon the coasts of these countries, may be encroached upon 

 by cold polar water. It is a matter worth notice that the last great ' outburst ' 

 from the Antarctic and the last great droughts of India fall within the same 

 period of years (189I-I898). By means of regular surface observations on board 

 of liners crossing the Indian Ocean and a few series of deep soundings along the 

 60th and 100th meridians such yearly variations in the hydrographic state of the 

 Indian Ocean as can be of meteoroLigical interest might be ascertained. 



In order to put the theory of the influence of ice-melting to a test the author 

 has carried out a series of experiments so arranged as to correspond as nearly as 

 possible to the natural conditions of the Norwegian Sea, and has compared the 

 results with the actual results of the Norwegian and Swedish hydrographic research 

 in this part of the ocean in 1900. A description of the last experiment of the 

 series, carried out by Mr. J. VV. Sandstrom, assisted by Miss A. Palmquist, is 

 given in the next paper. 



' It is evident that in the course of oceanic circulation there must be a transition 

 of Atlantic water into Arctic water, and vice versa. The first-named metamorphosis 

 is effected by the ice-melting process. 



