Files that need to be present with program:
        -Volcal are gastherm.dat and gas.sto.
        -Solvgas are gastherm.dat and gasrun.dat
        -Gasworks are gastherm.dat, workrun.dat, and workox.dat. 

   Some notes on making a Workrun.dat input file for Gasworks that will and converge on a solution. Pardon the horrible pun. In this example for a gas-rock titration as in the Mt. Spurr example. Every run is a little or a lot different.

   First run Volcal to convert the gas analysis to Gastherm basis components and in units of moles. Even if a gas analysis is already in units of moles and in basis components only, Volcal will properly format the block of columns that contain the gas composition data.

   The user tells Volcal to make either a Gasrun.dat or Workrun.dat file. The block of columns that contain the gas composition are identical in these two files, and the Volout.dat file in either case will be identical. For the purposes of example I did run it twice for Spurr. In practice Volcal needs to be run only once for a gas. It's faster to simply copy and paste the composition data blocks from the Volcal output file into a file previously used for something else. But don't forget to update the header!

   I did not run Solvgas initially, and tried Gasworks.

00:
   In the workrun file from Volcal
     -change ipun=3 to get the correct plot files
     -change lims=1 to invoke solid solutions (I failed to do this for the initial Spurr calc.)
     -change mins=1 for gas-rock titration in grams

   It crashes.

   In hindsight I should have got this one to run by adjusting mtrys BEFORE proceeding-
__________
01:
   I copied (from the Kilauea workrun file) tiny placeholder values for the gas components in the rock that has not yet been added. Components 6-30.

   I put in the rock composition you sent, adjusted so that all are in units of wt%.

   This too crashes.
__________
02:
   Use Solvgas to get better mtrys. Solvgas initially exceeds the default loop limit of 90. Set limit higher. Loops used turns out to be 94

   Edit Workrun.dat to use the mole values for derived gases from gasout.dat as new improved mtry values. Loops used is reduced to 7. I see now I could have got better mtrys for rock components by including them here.

   This too crashes.
__________
03:
   Sideline temporarily four trace metals Sr, Zn, Cu, and Pb, placeholders and rock composition data, to simplify things. While it will converge initially, it crashes at the first titration increment.

   As an aside, whenever you see "SOLIDS AND GASES CARRIED" in the output, the calculations are successful up to that point.
__________
04:
   Tinker with mineral suppression. Finding by suppressing coesite, corundum, and talc-prl, that it runs. Run for 10 increments, and then the three minerals can be unsuppressed and they will not return.

05: Adjust the mtrys a bit more for a cleaner look and make sure it still runs.

06: Put back the four trace metals, and away we go.