The saltpetre refinery occupied the right central portion of the front, being sixty-five feet long, fifty-five feet broad and thirty feet high, open from the floor to the ventilated roof. At the east end were four of the large evaporating iron pans, placed side by side, and elevated three feet above the floor by the brick work which surrounded them; five similar pans were in a corresponding position at the west end, and the large copper drying pans occupied forty feet along the north side at the same height. Each evaporating pan had a separate furnace, and the heated air from the whole passed beneath, and in contact with the bottoms of the drying pans on its way to the great chimney; the furnaces opened into side rooms communicating with the outside open space in the rear of the building. Thus the refining room was entirely free from ashes, dust and smoke. The centre space of the floor, about thirty-six feet square, was sunk four feet to allow water from the canal to pass around the bottoms of two of the large evaporating pans, which were placed therein near the centre of this area, and nine feet apart; these were used for a special purpose. The best quality of gunpowder can only be made from the purest saltpetre; the impurities of the crude material are mainly deliquescent salts, which rapidly deteriorate the strength of the powder by the moisture absorbed. To refine more or less the rough saltpetre of commerce is then a necessity even in producing an inferior article. To carry the refining process to the extent of nearly absolute purity, required several successive crystallizations and washings, involving a large amount of manual labor in the manipulation, and consuming much time. This was particularly the case in the very large amount of saltpetre, eight to ten thousand pounds per day, used by the Works, the refining of which would demand extended buildings and apparatus, as well as requiring a large number of operatives. Hence, it became desirable to devise methods by which hand labor could be superseded by motive power and machinery; in this I was entirely successful. Thus, in the operations of filling the various boiling pans with water or mother-liquor; the transference of the boiling solution of saltpetre to the draining trough, and thence to the crystallizing machines; the cooling down of the solutions, and their constant agitation to break up the forming crystals into fine particles, and transferring of these to an adjoining tank; the washing of the crystallized mass, and the subsequent removal of the mother-liquor and wash-waters, were all accomplished by machinery, with the assistance of two or three workmen only.