between 0.857 gm/cm* and 0.923 gm/cm^, a difference of 0.066 gm/cm^. The dif- 

 ference in density could contribute to a variation in the longitudinal plate-wave 

 velocity of the order of 150 meters/sec. The measurements presented in this re- 

 port do not show such a large variation of density. It is evident that although 

 density variation contributes to a change in the longitudinal plate-wave velocity, 

 it is not the major contributor. In 1954 and 1955, measurements were made in 

 smooth and in rough ice. An indication of the smoothness of the ice is shown by 

 the probable error of ice thickness measurements from tables 1 and 2. 



One set of geophones was placed in an area of rafted ice in 1955. The plate 

 velocities from this geophone set were lower. The size of the charge which was 

 used appeared to have no eflfect on the plate velocity, provided it was sufficiently 

 large to generate the waves. 



Attempts to measure shear wave velocities were unsuccessful, as it was im- 

 possible to identify positively the shear wave from the complex waves recorded. 



Tables 1 and 3 show the results of ice-thickness measurements determined from 

 air-coupled flexural wave frequencies. When compared with measured ice thick- 

 ness, the results for an area of smooth ice show, in general, an ice thickness ap- 

 proximately 10 centimeters less than the measured value. The difference probably 

 is due to the effect of the bottom layer of the ice sheet which is a weak slush type 

 of ice. For data taken in an area of rafted ice, the results from calculated ice thick- 

 ness are less accurate. The probable error in these tables gives an indication of 

 the unevenness of the ice-sheet surface. 



Efforts were made in 1955 to collect data on the independent parameters which 

 would aid in defining the state of sea ice. State of the ice is defined here as being 

 a function of a set of independent parameters, such that when these parameters 

 have some given value, the physical properties of the ice are fixed; that is, the ice 

 has a given state. If one of these parameters should vary, then the ice would assume 

 a different state. If the sea ice state can be defined by some dependent parameter 

 p; and further, if p is some function of n independent parameters such that 



P = /(^i. 12, qz, • ■ ■ qn) 



then any infinitesimal change in one or more of the independent parameters re- 

 sults in a change of the sea-ice state. 



At NEL, studies are being made on sea ice in an effort to determine which are 

 the independent parameters. In the work covered in this report, the parameters 

 which have been considered are density, salinity, air-bubble distribution, crystal 

 grain size, crystal grain distribution, and the elastic moduli. 



.The two salinity profiles taken at Site 2 are in agreement. Site 1 and Site 2 did 

 not appear to be located in the same ice sheet, and the variation in the salinity 

 profiles at these two sites seems to indicate that the thermal history of each sheet 

 was quite different. 



A density profile was made only at Site 1. The profile shows a distinct variation 

 in density but there appears to be no correlation with any of the other parameters. 

 The average density through the ice sheet is 0.90 gm/cm^. This value is used in 

 all calculations in the report. 



The profiles of air-bubble distribution, as presented in figures 13 and 14, show 

 that the size of the maximum number of air bubbles per unit area is below 0.1 

 mm. The general shape of the curves does not appear to have any correlation with 

 depth in the ice sheet. 



20 



