ATLANT. DEEP SEA EXPED. 1910. VOL. 



PHYSICAL OCEANOGRXPHY AND METEOROLOGY 



33 



the highest and lowest temperature found in this way are 

 recorded in cohmni 6 (C) of tiie table p. 27. As a re- 

 sult of the relatively ^reat influence of the diurnal varia- 

 tions, the daily disparity seems to be very great at some 

 depths. If further the difference between k^ and A, varies, 

 the interval between two successive ma.xima or minima 

 may be subject to comparatively great changes. In the 

 following table records are given of the interval in lunar 

 hours between the extremes, together with the variation 

 in temperature due to the combined effect of the diurnal and 

 semi-diurnal variations. Tiie first minimum after the upper 

 passage of the moon is numbered 1, and the records start 

 from this minimum. It corresponds to the first "upheaval" 

 of the water-layer in question after the moon's passage. 

 L\t means the time interval between the extremes, and 



10- A/ the temperature variation in 



o^C. 



On the assumption that and k^ — k., may actually 



vary as much as our computations seem to show, the 

 above table gives an illustrating example of the disparity 

 in temperature variations whicii may be encountered in 

 the sea as an effect of diurnal and semi-diurnal oscillations 

 only. We find for instance at Stat. 1 1.') that at 400 metres 

 the interval of time from the first minimum to the second 

 was as much as nearly 15 lunar hours, but the next interval 

 to a minimum was only a little more than 9 lunar hours, 

 while the corresponding intervals at 300 metres were 11.5 

 and 12.5 lunar hours. The disparity which may exist 



with regard to the amplitude of temperature is also demon- 

 strated in tiie table above. 



When examining variations of temperature in relation 

 to the tides we are apt to look for periods of 12 and 24 

 lunar hours. The results now arrived at indicate, how- 

 ever, that in many cases the 12 hour period may be quite 

 distorted and unrecognizable unless the observations are 

 subjected to harmonic analysis. In such cases the ob- 

 servations must embrace 24 hours at least. 



If we have to deal with 12 and 24 hour periods only, 

 the variations will repeat themselves after 24 hours, regard- 

 less of the ratio between A and B and the phase difference 

 A'j — k.>. But it is possible, or even probable, that at least 

 4 different tidal periods have to be considered: M.,, S.^. O 

 and /\;'i. In that case the final result will depend upon the 

 difference of amplitude and phase between these "waves" 

 miutually, and we may expect a still greater variability 

 which, to begin with, may in many cases look very irregu- 

 lar and seemingly without any connection with the tides. 



The combined effect of the diurnal and semi-diurnal 

 variations when converted into vertical oscillations is illu- 

 strated in Fig. 4 (the scale of the oscillations, in metres, 

 to the right in the figure). 



In agreement with what has been said above we find 



at Stat. 1 15 comparatively great oscillations in the upper 



dz 

 300 metres, C J being between 24 and 31 metres (cf. 

 a I 



table p. 27), and a decrease deeper down to 7 metres only, 

 at about 500 metres' depth. 



At Stat. Sc. the corresponding oscillations were very 

 great at about 100 and 200 metres. The isothermal sur- 

 face of 8-16 ' C was on an average found at 100 metres 

 below the surface of the sea at the Scottish station, but 

 owing to diurnal and semi-diurnal variations (other varia- 

 tions disregarded) it was apparently lifted nearly to the 

 level of 70 metres on August 13th between 13 h and 14 !i 

 G.M.T. (19 20 L. H.), and pressed as far down as about 

 140 metres below the sea surface on August 14th between 

 7h and 8h G.M.T. (about 13 L. 11.), the calculated dif- 

 ference of depth being G7 metres. Tidal variations of the 

 same amplitude were found at about 200 metres. The 

 total amplitude was much smaller further down, with a 



dz 

 minimum of C 7- :^ 17 metres at about 400 metres, from 

 a r 



where it increased to 38 metres at about 600 metres below 



the surface. The apparent vertical oscillations of the 



i.sothermal surface caused by tidal variations were much 



greater at this Scottish station than at Slat. 115 except 



at 300 metres, where the ratio was inversed, and at 400 



metres, where the difference was but small. 



