1912. No. 12. 



THE SEA WEST OF SP1TSBERGE^". 



Tiz. 6. Vertical Temperature- Curves of Sect. VI. Readings of Nansen- 

 thermometer are not corrected. 



Om 



700 



500 



temperature-ob- 



seni-ations at Stat. 

 3 on June 25th 

 and September 

 7th; the increase 

 of temperature 



Fig. g. Vortical Temperature-Curves of 

 June and September 1910, for Stat. 3. 



Readings of Nansen-thermometer are not 



corrected in broken and unbroken lines. 



The dotted line gives the corrected 



temperatures of June. 



2' 3' 4* 5' e- 



600- 



900 



1000 



1100 



during the interval 

 would have been 

 astonish in gl}- great 

 if the readings were 

 quite comparable 



^1 



without correction ; it is more probable that the readings ot 

 the first date, in depths between the surface and 400 metres. 

 were too low as we have suggested above, while the ob- 

 servations of the latter date were fairly correct. The differ- 

 ence between the cur\-es oi Fig. 8. for Stat. 4. is not so 

 great; but the shape of the curves indicates a similar con- 

 trast between the observations of June 25th and those of 

 September 9th. It is a striking fact that at both stations the 

 observations taken with the reversing thermometer in June agree very well 

 with those of September, while there is such a great difference between 

 all observations taken with the Nansen-thermometer in the ti^o months. 

 Figs. 9 and 10- representing the temperature-obser\ations at Stats. 29 and 

 30 on July 22nd and September 6th— give a more probable illustration ot 

 the variations between these dates. It is thus probable that the temperature- 

 readings of the Nansen-thermometer have been much too low during the 



1Z00 



fSOO 



1400 



f500\ 



