This combination would give a good permanent glass, whether the story is true or not, and very long before Pliny’s time glass was made in great variety of composition and color. In fact in default of porcelain glass was used in Roman times relatively more than now. But without knowledge of optics there was no need for glass of optical quality, it was well into the Renaissance before its manufacture had reached a point where anything of the sort could be made available even in small pieces, and it is barely over a century since glass-making passed beyond the crudest empiricism.
Glass is substantially a solid solution of silica with a variety of metallic oxides, chiefly those of sodium, potassium, calcium and lead, sometimes magnesium, boron, zinc, barium and others.
By itself silica is too refractory to work easily, though silica glass has some very valuable properties, and the alkaline oxides in particular serve as the fluxes in common use. Other oxides are added to obtain various desired properties, and some impurities may go with them.
The melted mixture is thus a somewhat complex solution containing frequently half a dozen ingredients. Each has its own natural melting and vaporizing point, so that while the blend remains fairly uniform it may tend to lose some constituent while molten, or in cooling to promote the crystallization of another, if held too near its particular freezing point. Some combinations are more likely to give trouble from this cause than others, and while a very wide variety of oxides can be coerced into solution with silica, a comparatively limited number produce a homogeneous and colorless glass useful for optical purposes.
Many mixtures entirely suitable for common commercial purposes are out of the question for lens making, through tendency to surface deterioration by weathering, lack of homogeneous quality, or objectionable coloration. A very small amount of iron in the sand used at the start gives the green tinge familiar in cheap bottles, which materially decreases the transparency. The bottle maker often adds oxide of manganese to the mixture, which naturally of itself gives the glass a pinkish tinge, and so apparently whitens it by compensating the one absorption by another. The resulting glass looks all right on a casual glance, but really cuts off a very considerable amount of light.
A further difficulty is that glass differs very much in its degree of fluidity, and its components sometimes seem to undergo mutual reactions that evolve persistent fine bubbles, besides reacting with the fireclay of the melting pot and absorbing impurities from it.
The molten glass is somewhat viscous and far from homogeneous. Its character suggests thick syrup poured into water, and producing streaks and eddies of varying density. Imagine such a mixture suddenly frozen, and you have a good idea of a common condition in glass, transparent, but full of striæ. These are frequent enough in poor window glass, and are almost impossible completely to get rid of, especially in optical glass of some of the most valuable varieties.
The great improvement introduced by Guinand was constant stirring of the molten mass with a cylinder of fire clay, bringing bubbles to the surface and keeping the mass thoroughly mixed from its complete fusion until, very slowly cooling, it became too viscous to stir longer.
The fine art of the process seems to be the exact combination of temperature, time, and stirring, suitable for each composition of the glass. There are, too, losses by volatilization during melting, and even afterwards, that must be reckoned with in the proportions of the various materials put into the melting, and in the temperatures reached and maintained.