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



39 



taking 5, — 102749, 5o ~- 102771, z, = Za = 400 metres 

 (the bottom-depth a short distance N of the station is 

 800 metres). We then obtain: 



and 



c :^ 0-65 metres per second 

 2 a ^= 618 /\ f metres 



If, for instance, ^\ v be taken as equal to 0-1 m sec. 

 this computation gives a boundary wave witli a height 

 of more than 60 metres from crest to througli, advancing 

 at a speed of 1-3 knots. Tliis would mean oscillations 

 of the same order of magnitude as we have found by the 

 harmonic analysis of our observations of temperature. 



If we have two layers, each homogeneous with a dis- 

 continuity at the boundary, the amplitude of the oscilla- 

 tions decreases upwards as well as downwards from the 

 boundary, the phase along one and the same vertical 

 having everywhere a definite value k° or k° -f 180°. In 

 stratified water the conditions are altered in several 

 respects, and further complications arise when the upper 

 edge of the ridge is not horizontal, but slopes so that 

 boundary waves may be created at different levels in dif- 

 ferent parts of the ridge. With our present knowledge 

 it is impossible to make even a fairly trustworthy compu- 

 tation of the system of boundary waves which may occur 

 in the Faeroe-Shetland Channel. 



Boundary waves created at the Wyville Thomson 

 Ridge will probably take some days in reaching the places 

 where the "Michael Sars" station 115 and the Scottish 

 station were taken. The waves would probably appear 

 especially at 300 — 400 metres' depth at Stat. 115 and at 

 400—500 metres at Stat. Sc, provided that the waves 

 were substantially confined to the water layers in which 

 they were originally created. On this assumption we 

 should then have a marked decrease of amplitude upwards 

 and downwards from these levels, with a certain constancy 

 of phase. This does not seem to correspond with our 

 results as regards the temperature variations with tidal 

 periods. It must be noted that the waves are damped on 

 the way, so the amplitudes decrease if the difference of 

 density {s^ —5,) is maintained. 



We may certainly expect to find boundary waves in 

 some parts of the sea, such as the Faeroe-Shetland 

 Channel, and it is probable that some of the observed 

 temperature-variations with tidal periods, as well as some 

 of the residual variations, are caused by such waves. In 

 my opinion most of the periodic temperature variations 

 found at Stats. 1 15 and Sc. are, however, due to horizontal 

 displacements by tidal currents. The mixed effect of tidal 

 currents on the spot and boundary waves makes the varia- 

 bility within such a sea-area all the greater. 



}'. Stationary zcavcs. 



It has been suggested that stationary waves (standing 

 waves) may occur in the sea in such a way that perceptible 

 vertical oscillations may arise. In the case of stationary 

 waves the vertical amplitude is nil at the nodal line and 

 increases to a maximum at a horizontal distance from the 

 node of V* of "'t^ wave length. Between two nodal lines, 

 or in one half of the basin in case of a uninodal oscilla- 

 tion, the phase is exactly the same in all places and depths, 

 and exactly 180^ different from the phase at the other 

 side of the node. It is readily seen that the variations 

 of temperature observed in the Faeroe-Shetland Channel 

 cannot be explained by ordinary waves of this kind, possibly 

 with the exception of a few of the residual variations. 



A special kind of stationary waves may arise when 

 boundary waves are reflected on reaching the slope of 

 the sea-bottom in such a way that an interference between 

 the primary waves and the reflected occurs. As an example 

 of such stationary waves may be given the internal seiches 

 (or "temperature seiches") observed, for instance, in Scot- 

 tish lochs and studied by E. M. WF.nnKRHURN. Provided 

 that possible boundary waves in the Faeroe-Shetland 

 Channel are created at the Wyville Thomson Ridge and, 

 like the tidal wave, proceed N. E. in the direction of the 

 Channel, the reflection will probably be of but subordinate 

 importance in this area, even if some reflection takes place. 



Boundary waves casually created by transient atmos- 

 pheric disturbances may be transformed into internal 

 seiches, in which case the period of oscillation would de- 

 pend upon the extension of the sea basin and only rarely 

 coincide with the tidal periods, while the phase should 

 be the same everywhere between the nodal lines.— 



Vertical oscillations of the water layers, whether due 

 to really vertical motions or to horizontal displacements, 

 are of great interest in several respects. The problem is 

 in itself important for our understanding of what is taking 

 place within the huge water-masses of the Ocean, and 

 further investigations by means of detailed observations 

 may even lead to progress in general hydrodynamics. 

 For physical oceanography such studies will be of great 

 significance if the question of the representative quality 

 of ordinary oceanographic observations (cf. section 19) 

 be thereby more definitely established than is now the 

 case. It is also worth mentioning here, that the pro- 

 blem of vertical oscillations has a bearing upon some 

 questions in marine biology and fisheries. Short-period 

 variations in the distribution of the water layers with such 

 amplitudes as may possibly come to light, may perceptibly 

 affect the local occurrence of plankton, and in that case 

 perhaps of fish-shoals as well. 



