conclude that the system is in quasi-steady state. The waviness of the lines 

 connecting the same layers in the series (180 and 320 meters) is attributed 

 to possible differences in fall rate of the XBT' s as well as internal wave 

 activity. 



During the interval between series one and two, at least one new layer 

 appeared in the profile. It is shown within the shaded section of Figure 4. 

 This "new" layer seems to have forpned from an interface zone much thicker 

 than normal; it occurs within one of the zones showing considerable vertical 

 fluctuations, indicating the possibility that it began with a mlxlnK 

 process initiated by internal wave activity. We do not know the extent of 

 horizontal advection under the island; therefore, we cannot differentiate 

 between the formation of a new layer and the relative advection of a new 

 layer into the sampling area. 



The half- steps which are shown in Figures 3a and 3b could be evidence 

 of spasmodic mixing across an interface as suggested by Stommel and Fedorov 

 (1967). The series of expanded profile sections displayed in Figure 5 contain 

 several half-steps which show up in all profiles over the one-hour series. 

 None of the half- steps exhibit any evidence of growth or decay. It is possible 

 that our series of measurements "captured" the formation of a half- step 

 (indicated in the shaded portion of Figure 4). We computed a mean turbu- 

 lent diffusive heat flux for the T-3 profiles. Using a vertical eddy coef- 

 ficient, A, of 1 cm^'/sec we obtained a flux of 2 x 10"^ g cal/cm''. Using 

 our heat flux calculation and the spasmodic model discussed by Stommel 

 and Fedorov (1967) we calculated a cycle time of about 33 hours during 

 which all interfaces should have been broken and reformed once. The pro- 

 files we measured are much too stable to fit the spasmodic model, but the 

 stepped structure is virtually identical to such a model. 



Arctic measurements taken from T-3 are significant for two reasons. 

 First, T-3 is possibly the world's most stable sea-going platform. Second, 

 both Case I and Case III stability structures occur under the ice. These 

 structural classes are ones which must be investigated in situ before a com- 

 plete understanding of layering phenomena is obtained. 



The acoustic significance of layering in the Arctic . 



It is also interesting to speculate on the significance of the Arctic step 

 structure on the propagation of sound. The influence of the layering struc- 

 ture in the Arctic on sound propagation depends on the boundary gradients 

 and the thickness of the layers, the frequency of the sound, and the number 

 of layers. Since our measurements are only of temperature structure, the 

 sound velocity and density contrasts cannot be specified with certainty. 

 Therefore, the discussion in this section must be considered preliminary. 



462 



