142 FRIDTJOF NANSEN. 



to inaccuracies in the samples, there has probably been some 

 evaporation through the cork-stoppers. 



Although we have examined, both by titration and with the 

 hydrometer of total immersion, numerous water-samples of very 

 different origin. Polar water, Atlantic water, etc. we have not 

 been able to discover any certain difference in the ratio between 

 the amount of halogen and the specific gravity of the water 

 and it seems as if this ratio is practically constant in the samples 

 examined by us, if no higher degree of accuracy be desired. 



PI. I demonstrates the distribution of temperature and 

 salinity at the sea-surface duiing the voyage of the Michael Sars 

 from July to September, 1900, according to the observations made 

 chiefly by Helland-Hansen, and for the northern part of the 

 sea (Sections III — V) by Dr. Hjort. The route of the ship 

 from Geiranger Fjord in Norway to Dyrafjord in Iceland with 

 Stations from 2 to 12, was made from July 18th to 30th, 

 1900. The route from Dyrafjord to Station 13 and thence to 

 Jan Mayen was made from August 3rd to 7th, 1900. The 

 section from Jan Mayen to Lofoten, with Stations from 29 to 

 47, were made from August 9th to 14th, 1900. The sections 

 IV, V, and III from Finmarken to Bear Island and thence 

 back to Lofoten were made from August 28th to September 

 8th 1900. 



The surface temperature seems on the whole to have been 

 somewhat lower than usual at this season in the north-eastern 

 part of the sea i. e. in the region of the Gulf Stream along the 

 northern Norwegian Coast and towards Bear Island. But in 

 the western part of the sea, east and north of Iceland, the sur- 

 face temperatures seem to have been higher than usual. This 

 becomes at once evident if PI. I be compared with similar sur- 

 face charts from the summers 1876 — 1878 (of the Norwegian 



