Fig. 3. Linear regression analysis of arginine kinase rate constants and flux rates against the extent of the hypo- or hyperosmotic treatment (osmotic change). The osmotic change was determined by subtracting the control osmolarity from the experimental osmolarity. Negative values indicate a hypo-osmotic treatment, positive values indicate a hyperosmotic treatment and zero indicates controls. For example, the-320mosmoll^-1 osmotic treatment results when muscle from an animal exposed to 35 salinity is superfused with a medium of 640mosmoll^-1 (640-960=-320). Regression lines are shown with 95% confidence limits. Regression equations are as follows: k_forward=-4.4x10^-5O+0.10, r²=0.07, P=0.07 (not significant); k_reverse=-3.3x10^-4O+0.41, r²=0.28, P<0.01; forward flux=-1.3x10^-3O+1.43, r²=0.24, P<0.01; reverse flux=-1.4x10^-3O+1.66, r²=0.22, P<0.01, where O is osmotic change.