200 Long Waves in Canals and Standing Waves in Closed Basins 



average depth of 235-7 m. Merian's formula gives a period of 18-39 h and, 

 with a correction for the mouth of 1 167, a period of 21 4 h. It fits quite well 

 the observed value of 22 h. The Japanese method, however, gives a negative 

 correction for both the width and the cross-section, so that the theoretical 

 value is reduced to 15-62 h. It is questionable whether, in view of the rather 

 large correction due to the form of this sea, the Japanese method can still 

 be used. The narrowing of a bay at the opening must lengthen the period 

 and not shorten it (see Defant, 1911, p. 126). If this oscillation is not the 

 uni-nodal period of the free oscillation of the Adria, it could not be explained 

 by its dimensions. Sterneck (1915, p. 44) finds, by integrating step by step 

 the equations of motion, for the Adria a period of 21- 12 h, which is increased 

 to 23-34 hours by a correction factor for the opening 1105. 



The 1 1 h wave can only be the uni-nodal seiche of the Adria, if the strait 

 of Otranto is assumed to be closed. Oscillations of this kind are perhaps 

 not entirely excluded, on account of the great narrowing of the bay at the 

 opening; the unusual wind conditions which accompany this oscillation 

 might also be the cause. The 8 A oscillation then is the first harmonic oscill- 

 ation of the Adria with a node at the opening and a second one inside, 

 whereas the short, 3 h waves which often occur might be transverse oscill- 

 ations. The average cross-sections of the Adria give as period of such trans- 

 verse oscillations approximately 2£ h, which is in agreement with the ob- 

 served values. 



The Black Sea can be considered as a completely closed basin with regard 

 to its seiches. Both Sterneck (1922) and Kurtschatoff (1925) have only 

 been able to find short-period oscillations from tide registrations, which are 

 to be considered as local bay and shore oscillations. A careful analysis of 

 the tidal curves made by Endros (1932, p. 442) has shown that a small 

 seiche of a definite duration can be read from the curves after elimination 

 of the small tides. However, on account of the little development of these 

 seiches, the Black Sea, despite its large dimension, must be regarded as 

 a badly tuned oscillating basin. Endros found periods of 74, 6-4 and 5-5 h. 

 The first period seems to be essentially an oscillation of the north-western 

 shallow section of the sea (Bay of Odessa). Endros believes that the two 

 other oscillations are of the entire water-mass of the Black Sea along the 

 main deep basin towards the southwest (Burgas, 5-5 h), and towards the 

 middle of the western shore (Constanza, 6-4 h). These two periods are very 

 similar, and when they occur simultaneously, it is very difficult to separate 

 them by analysis. The 5-5 h wave must be attributed to the uni-nodal 

 longitudinal oscillation of the main basin in the direction Poti-Burgas. 

 Sterneck found for the theoretical period of this longitudinal oscillation, 

 using the Japanese method, 4-98 h, Defant (1918, p. 78) using the residual 

 method, 512 h. In both computations the Bay of Odessa was omitted. If the 

 water-masses oscillate more from the middle of the lake towards the south- 



