ATLANT. DEEP-SEA EXPED. 1910. VOL. i| PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



Norway. A slight minimum is seen about August lOlh, 

 near Shetland. 



In this manner the general variations in salinity are 

 fully explained as a consequence of the variations in 

 regions. The casual meteorological variations (e. g. in 

 wind-conditions) exert quite a subordinate influence upon 

 these general variations demonstrated by observations 

 from a ship moving at considerable speed and covering 

 great distances. 



The more rapid casual, or periodic, variations may 

 be studied by means of the B-curves, pp. 76*, 78*, 80*, 

 and 82*. They are drawn on a much larger vertical 

 scale than the A-curves. The differences o — ni 12,2 

 have, as already mentioned, been smoothed by the calcu- 

 lation of continuous means for every 3 successive values. 



Great variations were met with in the transitional 

 regions mentioned above: on June 26th, 27th, 29th, July 

 9th - 12th, 27th, 28th, and August 14th. Otherwise the 

 variations were relatively small. There are no conspicuous 

 daily variations to be seen. The maxima, for instance, 

 occurred at any time of day or night. Nor is there any 

 quite definite rhythm in the variations, although a mean 

 period of some 16 hours seems to be possible. 



Comparison with the curve W shows that there is 

 in many cases some apparent coincidence between the 

 variations in salinity and in wind velocity. On many days 

 a maximum of wind velocity corresponded to a maximum 

 of salinity, but on other days a maximum of wind corre- 

 sponded to a minimum of salinity. This discrepancy might 

 be accounted for by variations in the direction of the 

 wind, but our observations do not seem to give a clear 

 and definite solution. 



It must also be borne in mind that the salinity in 

 many or most cases shows horizontal variations down to 

 considerable depths which are similar to those at the 

 surface, as will be seen from the vertical sections and 

 will be further described below. A casual variation of the 

 wind cannot create a momentary displacement of a deep 

 water-stratum. 



Our observations seem to prove that there are many 

 but rather small variations in salinity due to a heler- 

 ogeneous or irregular distribution of salinity. Patches 

 of water with relatively high salinity may evidently 

 alternate with patches of less saline water. These varia- 

 tions are, in many cases, probably connected with peculiar 

 dynamic conditions. 



Near the border of a current or, generally, in places 

 with a great horizontal gradient of salinity a persistent 

 departure from the average wind conditions will evoke 

 variations of salinity at a fixed point (relatively to the 

 sea-bottom). Such variations cannot be properly studied 

 by means of our observations (cf. section 7). 



7. The Surface Temperature. 



The A-curves for surface temperature (ig ) on pp. 76*, 

 78*, 80*, and 82* show clearly the same general variations 

 as we find in the salinity curves. An increase in salinity 

 corresponds to an increase in temperature. The only 

 exception to the rule is that the coastal surface waters in 

 August (the last series) were as warm as, or even warmer 

 than the Atlantic water outside. Near the Newfoundland 

 coast the surface water was also warmer, and at the same 

 time less saline, than further out in the Labrador current. 

 This corresponds to the well known fact that in summer 

 the surface water near a coast is very often warmer 

 than out at sea. This was not the case, however, along 

 our route in British waters on July 27th and 28th. 



When the exceptions mentioned are left out of account, 

 a study of the A-curves will show that the variations in 

 salinity coincided with similar variations in temperature. 

 On the 11th, 12th and 20th of June, not far from the 

 Azores, the surface temperature had a maximum not 

 accompanied by a similar maximum of salinity. Other- 

 wise there is a perfect positive correlation between salinity 

 and temperature as far as these smoothed values are 

 concerned. This correlation will be dealt with generally 

 in section 18. 



The great variations in the mean surface temperature 

 in the areas in question are, therefore, due to the regional 

 variations. The smaller variations are more easily recog- 

 nized in the temperature curves than in the salinity curves 

 (on account of the scale used). They are evidently due 

 to local variations in the direction of the currents or in 

 the track of the ship. A small deviation from the general 

 direction of the current or of the ship may produce a 

 marked effect in places where the horizontal gradients of 

 salinity and of temperature are considerable, as will always 

 be the case when the water flows with great velocity 

 relatively to the surrounding water-masses, e. g. near the 

 northern border of the "Gulf Stream" between America 

 and Europe. 



The positive correlation between salinity and tempe- 

 rature is very conspicuous even in the details shown by 

 the B-curves. With the exception of the series from August, 

 the various series show with remarkable clearness that 

 almost all the rapid variations in salinity were accompanied 

 by quite similar variations in temperature. The daily 

 period of temperature is not so evident as might be 

 expected: it is very often entirely blurred by the local 

 variations. It is, therefore, a very unsatisfactory task to 

 determine the daily variations of temperature from obser- 

 vations made on board a ship which passes through heter- 

 ogeneous water-masses. Reasonably trustworthy results 

 can, as a rule, only be obtained by measurements covering 



