and air-tight. The crucible having been put into a suitable furnace and gradually heated, a small platinum pipe communicating with the upper part of the crucible is also connected with an exhaust pump, so as to remove every particle of air from the crucible and from between the granules of glass while these still retain their granular condition. As soon as the glass becomes fluid, it forms a homogenous mass, the law of diffusion equalising any minute differences in composition of continuous grains, while wholly avoiding those long "wreaths" or streaks so fatal in large masses of glass. On the strength of these crude notions a number of various-shaped clay crucibles were ordered to be made, with a view to carry on an elaborate series of experiments on the lines indicated; but as these crucibles required at least three months to dry, I had ample time to pursue some other interesting investigations relative to the production of glass for ordinary commercial purposes.

In going over a glass-works some years previously, I had noticed what I, at the moment, thought was a great oversight in the mode of proceeding. The materials employed, viz., sand, lime, and soda in ascertained quantities, were laid in heaps upon the paved floor of the glasshouse, and a labourer proceeded to shovel them into one large heap, turning over the powdered materials, and mixing them together; a certain quantity of oxide of manganese was added during the general mixing operation, for the purpose of neutralising the green colour given to glass by the small amount of oxide of iron contained in the sand. The materials were then thrown into the large glass pots, which were already red-hot inside the furnace. What appeared to me to be wanting in this rough-and-ready operation was a far more intimate blending of these dry materials. A grain of sand lying by itself is infusible at the highest temperature attainable in a glass pot, and the same may be said of a small lump of lime; but both are soluble in alkali, if it be within their reach. These dry powders do not make excursions in a glass pot and look about for each other, and if they lie separated the time required for the whole to pass into a state of solution will greatly depend on their mutual contact. In such matters I always reason by analogy, and look for confirmation of my views to other manufactures or processes with which I may happen to have become more or less acquainted. I may here remark that I have always adopted a different reading of the old proverb "A little knowledge is a dangerous thing"; this may indeed be true, if your knowledge is equally small on all subjects; but I have found a little knowledge on a great many different things of infinite service to me. From my early youth I had a strong desire to know something of any and all the varied manufactures to which I have been able to gain access, and I have always felt a sort of annoyance whenever any subject connected with manufacture was mooted of which I knew absolutely nothing. The result of this feeling, acting for a great many years on a powerful memory, has been that I have really come to know this dangerous little of a very great many industrial processes.

I have been led into this long digression because I meant to illustrate my observations on the extreme slowness of the fusion of glass by an analogy in the manufacture of gunpowder. I have shown how impossible it is for the dry powdered materials employed in the formation of glass to chemically react upon each other when they are lying far apart. Now, if we take the three substances -- charcoal, nitre, and sulphur, of which gunpowder is composed, and break them into small fragments, then shake them loosely together, and put a pound or two of this mixture on a stone floor and apply a match, the nitre will fizzle a little briskly; the sulphur will burn fitfully or go out, and the charcoal will last several minutes before it is consumed. If, instead of this crude and imperfect mixture, we take the trouble to grind these matters under edge- stones into a fine paste with water, and then dry and granulate it, we have still the precise chemical elements to deal with as we set fire to on the stone floor; but they now exist in such close and intimate contact as to instantly act upon each other, and a ton or two of these otherwise slow-burning materials will be converted into gas in a fraction of a second. The inference I drew from this analogy was simple enough, viz.: grind together the materials required to form glass, and when the heat of the furnace arrives at the point at which decomposition takes place, the whole will pass into the state of fluid glass much more quickly, and will yield a more truly homogenous glass than is obtained in the usual manner.

I was at this time engaged in constructing a large reverberatory furnace for the fusion of glass on the open hearth, and I may forestall what I have to say respecting this mode of founding glass, by stating that when I employed a mixture of raw material merely shovelled into the bath as practised in ordinary glass-making, it took from ten to twelve hours to fuse half a ton of sand and lime in my new furnace; but