The nitro-glycerine, of course, collects upon the surface, and can be drawn off by a tap placed at a convenient height for the purpose.  The cover of the tank is generally conical, and is joined to a glass cylinder, which is cemented to the top of this lead cover, and also to the lead chimney.  In this glass cylinder is a hole into which fits a ground glass stopper, through which the nitro-glycerine can be drawn off.  There will probably never be more than an inch of nitro-glycerine at the most, and seldom that.  It should be taken to the filter house and treated along with another charge.  The acids themselves may either be run to waste, or better treated by some denitration plant.  This house probably requires more attention than any other in the danger area, on account of the danger of the decomposition of the small quantities of nitro-glycerine, which, as it is mixed with such a large quantity of acids and water, is very apt to become hot, and decomposition, which sets up in spots where a little globule of nitro-glycerine is floating, surrounded by acids that gradually get hot, gives off nitrous fumes, and perhaps explodes, and thus causes the sudden explosion of the whole.  The only way to prevent this is for the workman in charge to look at the thermometers frequently, and at the colour of the escaping fumes, and if he should notice a rise of temperature or any appearance of red fumes, to turn on the water and air, and stir up the mixture, when probably the temperature will suddenly fall, and the fumes cease to come off.

The cause of explosions in this building is either the non-attention of the workmen in charge, or the bursting of one of the water pipes, by which means, of course, the water, finding its way into the acids, causes a sudden rise of temperature.  If the latter of these two causes should occur, the water should at once be shut off and the air turned on full, but if it is seen that an explosion is likely to occur, the tank should at once be emptied by allowing its contents to run away into a drowning tank placed close outside the house, which should be about 4 feet deep, and some 16 feet long by 6 feet wide; in fact, large enough to hold a considerable quantity of water.  But this last course should only be resorted to as a last extremity, as it is extremely troublesome to recover the small quantity of nitro-glycerine from the bottom of this tank, which is generally a bricked and cemented excavation some few yards from the house.

It has been proposed to treat these waste acids, containing nitro-glycerine, in Mr M. Prentice’s nitric acid retort.  In this case they would be run into the retort, together with nitrate of soda, in a fine stream, and the small quantity of nitro-glycerine, coming into contact with the hot mixture already in the retort, would probably be at once decomposed.  This process, although not yet tried, promises to be a success.  Several processes have been used for the denitration of these acids.