Bug: Trouble fitting the log(Gaussian) distribution.

 Prism has built-in a log(Gaussian) model, which could also be called log normal. If you try to fit a frequency distribution to this model, older versions of Prism probably won't find a fit. The problem is that we built-in an incorrect initial value for the parameter Center (which could also be called the geometric mean).

To fit this model with older versions of Prism:

1. Choose the built-in model, and then clone it to create a new model. You may also want to rename it to "Log Normal" which is a more standard term. 
2. Consider changing the parameter name from Center to Geo. Mean.
3. Change the rule for initial value for Center (or Geo. Mean) so it becomes 1 * XatYmax
4. Now the equation should fit properly.

We fixed this issue, and current releases of Prism fit the equation just fine.