actual sea water. In fact, data for triple concentrated sea 
water could not be obtained even at 100°C, due to slow 
precipitation of CaSO, during the run. Another apparent 
reason for variations with Lindsey and Liu's results is 
that their method of fitting data to a polynomial shows a 
standard deviation of 0.017°C. Lindsey and Liu fit their 
data at first to a second degree polynomial in concentration 
at fixed temperature. However, the maximum number of data 
points at any temperature is only three. After the coef- 
ficients of power series in concentration were determined, 
the coeffictents were f1eted toa polynomial jin temperatuner 
VISCOSITY OF SEA SALT SOLUTIONS 
Preliminary work has been reported previously (2). The 
viscometer used was of the glass capillary type and is shown 
in Figure 22. Figure 23 shows the viscometer above the 
constant temperature silicone oil bath. Complete details of 
the system, its operation and results have been recorded (18). 
There were two main difficulties in measuring the 
Viscosity Ob sea salt solutions. First, much of, the 
temperature range to be investigated is above the boiling 
point of water and hence the system has to be pressurized. 
Second, natural sea water contains enough calcium and sulfate 
ions that precipitation may occur at high temperatures and 
concentrations. If this were allowed to occur the results 
would be erratic. The problem was avoided by precipitating 
and filtering the solution. 
Measurements of the viscosity of synthetic sea water is 
contained in work by Fabus et al. (19) and Grunberg (15). 
Kinematic Viscosity from Experimental Data 
The equation used to calculate the kinematic viscosity 
from the measured times of flow (t) is 
(21) 
where C and E are constants nearly independent of flow rate 
and temperature. The complete derivation is given by 
Ghenw@lsie 
70 
