Poor Toulouse indeed. Speaking of fire, I used to make fluxes for fire assay, a type of refining technique. This was not cooking, although a 900 C furnace was used, and at this temperature, small amounts of moisture are inconsequential (large amounts were!). It was important to weigh the ingredients (components) rather than measure them volumetrically. The correct ratios could be assured this way. Significant density variation could occur amongst the batches, due to the measuring-container packing techniques used by different technicians.