PRODUCTION AND PROPERTIES OF GLASS CONTAINERS 17 air, while another possibility is that they are formed by the impact of meteors on sand or sandstone, producing a pool of molten material which is shaped by splashing for a great distance through the air. Lumps of silica glass have been found in the Libyan desert, formed apparently by this method. The agency of lightning, apart from producing the fulgurites, may in fact be a possible origin of the commercial manufacture of glass. A French writer in 1878 first drew attention to the not uncommon occurrence of the burning of grain and fusion of the ash as the result of fire caused by lightning. The masses of glass found in the ash were called "lightning stones" (pierres de foudre) by the French peasants, who believed them to be the cause of the fire. For many centuries, plant ash of various types was one of the major constituents of glass making batches and the anhydrous sodium carbonate used by glass manufacturers today is always referred to as soda ash. Another illustration of the possible accidental origin of glass manufacture was related by Pliny in his "Natural History" in the first century A.D. He refers to the banks of glass making sand which were situated at the mouth of the River Belus in Phoenicia and which at that time had been in use for many years. "The story is that a ship, laden with nitre (i.e., sodium carbonate), being moored upon this spot, the merchants, while preparing their repast upon the seashore, finding no stones at hand for supporting their cauldrons, employed for the purpose lumps of nitre which they had taken from the vessel. Upon its being subjected to the action of the fire, in combi- nation with the sand of the seashore, they beheld transparent streams flowing forth of a liquid hitherto unknown this, it was said, was the origin of glass.'• This story is a plausible one, though it has been doubted if sufficient heat could be developed by an open fire as described. An American investigator has found that a wood fire kept burning for two hours in the open developed a temperature of 1200øC, which would be sufficient to melt not only a soda- silica glass but also a soft soda-lime-silica glass. (The lowest-melting eutectics being respectively at 793øC and 725øC.) The discovery of glass manufacture could have been made in the way related by Pliny, but the dating of the oldest piece of glass indicates that it took place hundreds of years before the time of the Phoenician traders. Where and when glass manufacture had its beginning is an open question, but the oldest piece of glass actually bearing a date, now in the Ashmolean Museum, Oxford, was made between 1,551 and 1,527 B.C.E. Some authorities have given dates as early as 12,000 B.C.E. for glazes, and 7,000 B.½.•. for pure glass, though the difficulty arises of establishing whether many of the early pieces are glass at all and not natural stones. As regards written records, a chemical text of Babylonian origin of 1,700 B.C.F•. or earlier, contains recipes for four glazes based on a "master glass", unfortunately unspecified, while the "Nineveh

18 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Tablets", 668--626 B.C.E., appear to be factory records and give not only the Assyrian names for various kinds of glass, but also directions and for- mulae for manufacture. We may note that dictionary definitions are numerous, but in general are too restrictive and fail to bring out the essential continuity of condition from the fluid melt at high temperatures to the rigid glass at room tempera- tures. One of the better of these definitions is that given in Thorpe's Dictionary of Applied Chemistry, i.e., "an amorphous, hard, brittle substance, usually transparent but sometimes translucent or even opaque, breaking with a conchoidal fracture and consisting of a material obtained by the fusion of one or more of the oxides, silica, boric oxide, phosphoric oxide and the metallic oxides, followed by sufficiently rapid cooling of the fused mass to prevent crystallisation of the components. The glass may be colourless or coloured either by one or more of the constituent oxides or from the effects of neutral bodies such as carbon, sulphur or selenium. According to definition, glass, being amorphous, shall be isotropic." (In other words, being uncrystallised, its physical properties are the same in all directions, unlike many other transparent substances used, for instance, in optical instruments.) Transparency is frequently considered an essential characteristic of glass, yet much commercial glassware is either translucent or opaque. It is always true, however, that the lack of transparency is caused by the fact that some material, usually crystalline, is dispersed or suspended in a glassy matrix, itself transparent. Other products may contain much glassy matter, but are not considered as glasses. For example, glazes and enamels are essentially glassy, whilst practically all fired ceramics contain a proportion of glass, but there is no suggestion that these products should be included in the definition. Similarly excluded are the organic substitutes which differ so greatly in composition, methods of manufacture and properties from those substances which have always been known as glasses that they constitute a separate division of chemical technology. Discussion of the definition of glass usually centres around the argument whether, at ordinary temperatures, glass is an undercooled liquid or amor- phous solid, which places the emphasis on the definition of liquid and solid rather than on the properties of the substance to be defined. When a crystalline solid is heated to its melting point, it changes to a liquid, or a mixture of liquid and a different crystalline solid, and on cooling the reverse phase changes take place. The materials which have always been called glasses are characterised by the property that when melted at high tempera- tures and cooled, they do not devitrify, i.e., they do not undergo the dis- continuous change into the stable aggregates of crystalline phases which equilibrium would require. At high temperatures, during melting, they are