Value of `eta` in gauges not recorded correctly

[mandli]

I noticed that if num_eqn > 3 that the value of eta is not recorded correctly in the gauge files despite being calculated correctly. I do not off-hand see the problem, but can reproduce it with just setting num_eqn = 4. Note that there are assumptions in the Riemann solver that may not like this so be careful if using this.