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



31 



difference between 100 and 400 metres being 3 hours. The 

 amplitude displays at first a decrease, and then an increase 

 to a very considerable nia.vimum in the "discontinuity- 

 layer" at 300 metres. Thence the amplitude decreases 

 rapidly to almost nought (B — 0-03° C. only) at 500 met- 

 res, where k., has nearly the same value as at the surface 

 or at 100 metres. At the Scottish station the conditions 

 are different: ^., shows irregular variations, and B has its 

 maximum at 400 metres with but a slow decrease towards 

 the bottom-layers. As to the semi-diurnal variations we 

 have data for comparison from the Danish station worl<ed 

 3 months earlier (table p. 27). At this station tlie am- 

 phtudes were much smaller in the upper 200 metres tiian 

 at either of the other stations. A large maximum is found 

 at 400 metres, and a considerable amplitude at 500 metres, 

 while at 600 metres the semi-diurnal temperature variation 

 is but small and much less than at the Scottish station 

 from the same region in August. At most levels the va- 

 lues of k^ at Stat. Da. do not differ much from those 

 found at Stat. 115, but more from those at Stat. Sc, in 

 spite of the short distance to the latter station. 



When converting the temperature-variations into ver- 

 tical displacements we find at all levels except 300 metres 

 greater semi-diurnal oscillations at the Scottish station 

 in the central part of the Faeroe-Shetland Channel than 

 at Stat. 115. At Stat. Sc. an absolute maximum of no 

 less than 58 metres' total amplitude is calculated for the 

 variations in depth of the isothermal surfaces at about 

 200 metres. It seems very likely that comparatively large 

 vertical displacements have really occurred at this station 

 at 200 metres, and at 100 metres too. A minimum is 

 found at 400 metres, while in the deeper layers the ampli- 

 tude is fairly great again. At Stat. 115 a maximum oscil- 

 lation (24 metres) is found at 300 metres, with a rapid 

 decrease downwards to some 3 metres only at 500 metres' 

 depth below the surface. At the Danish station the cor- 

 responding variations increase slowly from 8 metres at 

 100 metres' depth to 12 at 300 metres, and then quickly 

 to a maximum (31 metres) at 400. At 500 metres the 

 oscillation is found to be of the same magnitude as at 

 the Scottish station (24 metres), while at 600 metres it is 

 much less again. 



We thus find a great variability in the appearance of 

 the semi-diurnal temperature-oscillations, which seem to 

 be different not only in amplitude but also in phase in 

 vertical direction, to change from one station to another 

 within the same area, and also to be subject to varia- 

 tions with time as regards the harmonic constants. 



26. The Ratio between the Diurnal and 



Semi-Diurnal Variations. The Combined 



Variations. 



in column 10 of the table on p. 27 the ratio between 

 the amplitude of the diurnal variation {A) and that of the 

 semi-diurnal (B) is recorded, hi the rise and fall of the 

 sea surface with the tides the diurnal variations are gene- 

 rally of but secondary importance compared with the 

 semi-diurnal. The ratio between the amplitudes expressed 



K,-^,- O 

 by .. „ should, according to the equilibrium theory 



of the tides, be equal to 0-68, but is in the great majority 

 of cases much smaller [Krummf.l, 1911]. In the Atlantic 

 area it is mostly 0-1 — 0-3 only. In some regions, espe- 

 cially in low latitudes, the ratio may attain comparatively 

 large values, as is also the case in the Baltic. At Thors- 

 havn it has the relatively high value of 0-73, while along 

 the west coast of Norway it is only about 0-1. 



/I 

 The corresponding ratio ^ found for the temperature 



variations here dealt with has remarkably high values. At 

 Stat. 115 it is between 0-9 and 1-5 at all depths except at 

 300 metres where it amounts to 0-4 only. At this station 

 the diurnal temperature-variations seem to be dominant 



A 

 in most water-layers. At Stat. Sc. the value of -r. is 



B 



smaller than at Stat. 115 at all depths from 100 to 500 

 metres. At the Scottish station the semi-diurnal variations 

 are especially predominating at 200, 300 and 500 metres. 

 In the deep water at 600 metres the ratio is above 1 at 

 this station too, and at 100 and 400 metres the diurnal 

 variation is nearly as great as the semi-diurnal. Knudsf.n 

 has computed the constants of the 0-period in tempera- 

 ture at 400 metres at the Danish station. The (double) 

 amplitude was found to be 0-48, so that the ratio between 

 the amplitudes of and M is 0-6, which agrees fairly 

 well with the number 0-8 found for the same depth at 

 the Scottish station not far away. 



The ratio between the diurnal and semi-diurnal 

 temperature-oscillations thus seems to vary both in ver^ 

 tical and horizontal direction in the Faeroe-Shetlaitd Chan- 

 nel. At some depths and in some regions the semi-diurnal 

 variations in temperature seem to predominate in much 

 the same way as is usual in the tides, while elsewhere 

 the diurnal periods appear to have more bearing on the 

 variations within the sea than on the tidal movements 

 observed at the coasts. 



By a summation of the departures due to the diurnal 

 and semi-diurnal variations we obtain the variations il- 

 lustrated by the curves in Fig. 3. The differences between 



