Hydrographical Observations from the Danmark Expedition. 375 



Both causes led to an increase in the density of the water-par- 

 ticles in the surface layer, these sinking down and warmer particles 

 from the deeper layers coming to the surface, where these were cooled 

 down and so on; in other words, convection currents were set up, 

 by means of луЬ1сЬ the cold лvas spread deeper and deeper down. 



In the first part of the winter 1906—07 (until January 1907, 

 series XV) these convection currents reached down to a depth of ca. 

 50 m. Here there was always a break in the sahnity gradient, this 

 in the layer 50 — 75m. remaining unchanged about 32i/2 p.m. 



Below 75 ra. the conditions during this period were normal (that 

 is, increasing temperature and salinity greater than 33 p. m.) and 

 chlorides and heat were carried from there to the mixed layer on 

 the barrier. 



The warm and saline bottom-water was drawn out of Øresund 

 in the course of the \\'inter 1906 — 07 and became replaced by greatly 

 cooled surface water with salinity about 32^4 P- m- coming in from 

 the sea. The boundary between the upper and lower layers in Øre- 

 sund was thus smoothed out and the homogenous, ice-cold layer 

 formed (with a depth of 140m., salinity 32 yi p.m. and temperature 

 — 1.78°), which was found in the middle of May 1907 (series XX). 



As soon as the average density rose in the sea (May 1907), the 

 bottom-water flowed in and a transference of warmth and clorides 

 occurred, leading again to the formation of a division between upper 

 and lower layer at a depth of 75 m. This movement of the bottom- 

 water was independent of changes caused by the influence of the 

 atmosphere in the upper layers, for in the middle of May the mean 

 air temperature was still — 8° and no changes of density occurred 

 in the upper layer in May — June, which could give rise to movement 

 of the bottom-layer. 



It was not until the months of June, July and August, that the 

 mean air temperature became higher than that of the surface of the 

 sea ( — 1.7°) and then on the one hand direct transference of heat 

 took place, and on the other mixing with the warmed up water from 

 the melting snow and ice was going on. 



The average salinity of the layer — 50 m. decreased to ca. 31.3 

 p. m. and the temperature rose; the great break in salinity at 50 m. 

 in the series XLII shows, that there was constantly a boundary at 

 this depth between two layers of different character. 



In September 1907 the air temperature again grew lower than 

 that of the surface of the sea and in the course of the winter 1907 — 08 

 the surface layer — 50 m. was cooled down to nearly freezing point, 

 while the salinity was increased to ca. 32 1/2 P- rn. 



Less ice was formed this winter than in the preceding, neverthe- 

 less the increase in salinity was somewhat greater and the temper- 

 ature of the upper layer was on an average somewhat higher (which 



