- No. 2. SPITSBERGEN WATERS. 23 



The intermediate layer, with temperatures below o C., may differ 

 much in thickness and situation, as well as in temperature, from time to 

 time. On July i6th I found the upper boundary at about 70 metres (cf. 

 the isotherm of o in Fig. 21). On July i8th this boundary was probably 

 at a depth of about 65 metres, and the layer had a thickness of about 190 

 metres (Fig. 23). On July 2ist the upper boundary of this layer was at 

 about 35 metres, and had a thickness of about 120 metres (Fig. 22). On Au- 

 gust 24th the upper boundary of the cold layer was at a depth of about 

 85 metres. Its lower boundary was at about 225 metres, giving a thickness 

 of about 140 metres (Fig. 24). On August 29th the upper boundary was at 

 a depth of about 70 metres, and its lower boundary at about 215 metres 

 (Fig. 25). It is chiefly the conditions at the stations midway in the mouth 

 of the fjord that have here been considered. 



On the whole the temperature of this cold intermediate layer had risen 

 appreciably from July to the 24th and 29th August, showing that this 

 cold water is gradually heated during the course of the summer by contact 

 and intermixture with the overlying and underlying water-strata. It is 

 evident, however, that the temperature of the layer may change a great 

 deal from one day to another at the same place (cf. our observations at 

 Stations 44 and 48, and also at our Stations 7, 8 and 9 compared with 

 the observations at Station 10, and also Station 5). 



These rapid changes are obviously due to the horizontal circulation 

 caused by the tidal currents, which -flow in and out of the fjords, often 

 with considerable velocity. In this manner the water in the outer part of 

 the fjord may sometimes be carried out of it, and on to the shelf outside 

 the coast, at other times it is carried into the fjord again (cf. Fig. 26). Owing 

 to the variations in the velocity of the current in the fjord, the thickness 

 of the intermediate cold layer may also vary, as it is sometimes extended 

 and made thinner by an increased velocity, or vice versa. 



It will thus be understood that this intermediate cold water of the 

 fjord cannot be quite so local as described in our paper of 1912. It may 

 to some extent be influenced by the nature of the water over the shelf 

 outside the coast of Spitsbergen, in any case where the fjords have no 

 well-developed sill, and are fairly open. This seems to be verified by my 

 observations in the different fjords. In Cross Bay 1 also found an inter- 

 mediate layer with temperatures below o C. (Fig. 12), it was thinner in Cross 

 Bay than in Ice Fjord. In Wood Bay, and Wijde Bay there was no such 

 intermediate layer with temperatures below o C. (Fig. 37 and 28). Onl^ 

 at one station in Wijde Bay, Stat. 29, did I observe temperatures below 

 o C. at 100 and 150 metres; but the two other stations in that fjord gave 



