Category	Mark	Comments
Overall Presentation	7	Navigation could be more imaginative in a multi-paged report: it is difficult/slow to jump to a specific page. There are an excessive number of sub-pages. Variables should be given in italics (as done throughout the PD course). Many of the small gifs should be put in a line and not one above the other: and many are small enough to be in the main body of the text. Your starting parameters are not required in the report, though it did make the errors stand out for the monoclinic refinements.
HTML	9	Some 8-bit characters used, but otherwise fine.
Results A	18	Very good table of results except for missing esd's on geometry. You should constrain B-value to single value. Figures good size with nice use of expanded regions.
Results B (P 1 21/n 1)	7	Extremely serious error made when generating the 4th oxygen position. You must put it on the other side of the pseudo mirror plane so that you generate a tetrahedral sulphate anion. As a result, you ended up in a false minimum in which your S position is actually one of the oxygen atoms. You should check before and after the refinements that the anion geometry makes sense (since you know the values from the room temperature structure). It does not make sense to refine anisotropic temperature values for a sample at very low temperature. You should give proper figure captions, especially when large figures are in sub-pages. Despite the coordinate errors, you do get the correct distortion of the unit cell. You also have the fractional occupancies all wrong by a factor of two, i.e. the values should all be 1.0, and not 0.5. This is not critical for single phase refinements since the scale factor compensates, but it would be serious if you were using Rietveld for quantitative analysis. A nice touch was to include the number of parameters refined in your tables, but this was then spoilt by including scale factor and zero-point error, neither of which should be reported in a scientific paper. You should not be using databases with the project: you have all you need from the room-temperature structure regarding geometry and interatomic distances.
Results B (P 21/b 1 1)	8	You have made the same serious model errors here, but at least you have set the refinement up correctly in P2(1)/a with y unique. This refinement went into a different false minimum and you should have found it easier to spot since the S atom is still approximately in the correct positiom.
CIF	7.5	CIF should be a plain text file (not HTML to appear as text). Some selected structural information could be included.
Discussion & Final Conclusions	6	Good project and well written, but unfortunately spoilt by poor crystallographic understanding in places. You should never try refining the fixed 90° angles when monoclinic: some programs will even overide you if you try this one. However, it was a good idea to see the effect of the b angle being either < 90° or > 90°.