Table 3. Fitted values for the curves in Figure 1 and Figure 2.

Relationship	Estimated parameters	R-square	p.value
vs		0.97
		0.83	0.0109
		0.96
		0.04
		0.97
vs		0.69	0.01
		0.41	0.0848
		0.67	0.0123
		0.53	0.0104
		0.7	0.0093
vs		0.87	0.0069

The linear and non linear relationships, with and without the intercept, for and are reported here. These models are fitted to the experimental measurements listed in Table 2. For each model the fitting parameters, and the correlation p-value are reported. When and are related to , the non linear model with a fixed intercept and a free exponent (i.e. ) is associated with the best fitting results (bold). By adding one more free parameter (i.e. ) we do not get essentially any improvement (italic). The estimated value for , without any simplification, implies , and (see Materials and Methods). A direct proportionality is observed also for .