The ingredients are well mixed and thoroughly dried. Waste glass from a previous batch is also added. The mixture is heated to about 1200° C. in large pots made of Stourbridge clay, and the heating is continued for as much as sixteen hours, and until the whole of the material in the pot is molten and fairly mobile. Scum or glass-gall is removed, and when gas bubbles have disappeared, the temperature is allowed to fall to 700°-800°, when the glass becomes sufficiently viscous for subsequent working. The semi-fluid mass is then blown, moulded, or drawn, according to the kind of article that is required. The physical properties of glass will now be considered in order that we may be able to account for its extended use. Such an inquiry as this, especially in the case of materials in common use, is often interesting, because it frequently happens that the special property upon which we set so much value is an abnormal one and, consequently, the feature which we take for granted is precisely the one into which we should inquire most closely. The most striking feature of glass is its transparency. This property is abnormal, if glass is a solid. Consider what happens in most cases. A substance like nitre melts easily and in the molten state is perfectly transparent; when it cools, crystals form and, though these individually may be transparent, yet the solid mass is opaque. The reason for this is that the solid is not optically homogeneous, and therefore a ray of light cannot pass through it in a straight line. At each facet of a crystal light is deviated and reflected, and in the end is almost wholly scattered. Consequently, an object, even if it can be seen at all, can be discerned only in a blurred and indistinct fashion through such a medium. There are very good reasons, however, for supposing that glass is not a true solid but an extremely viscous liquid. If glass is heated, it softens and begins to flow very sluggishly at first, but afterwards more readily. There is no abrupt change, as there generally is in passing from the solid to the liquid state. Similarly in cooling, there is no point at which it is possible to say that the glass is solidifying. The view that this substance is really a liquid is perhaps a little startling at first, but it becomes less so when we observe that a long glass rod supported at its ends in a horizontal position sags in the middle and is permanently deformed. To avoid that change which would be technically called solidification by a scientist, the article which has been fashioned in glass is cooled down very slowly and gradually. This part of the process is called annealing; it may occupy some days in extreme cases, and it points to the fact that experience has shown that it is necessary to guard against some change which would normally take place if this precaution were neglected.