As to the apparent irregularity in the progress of the superficial decay, it would seem that, apart from differences in the chemical composition of the glass, much depends upon the preservation of the original smooth ‘epidermis.’ Once this is impaired, whether by accidental scratches or by the growth of fungus or lichen, the carbonic acid or the ammonia salts contained in the air or soil find, in the presence of moisture, a secure lodgment, and the work of decay proceeds rapidly. Thus in the case of the little flasks of primitive glass of which I shall have to speak in the next chapter, in one example it may be found that the smooth skin of the glass has for more than three thousand years remained absolutely intact, while in another specimen from a neighbouring tomb the glass not only on the surface, but far into the interior, has taken on a talc-like or porcelainous consistency, and the brilliant colours have for the most part disappeared. There is no need to enter into the details of the chemical processes involved in this process of decay. Suffice to say that the action is one of the same nature as that which has played so important a part in the geological changes of the earth’s surface, especially in the disintegration of the granitic rocks. It depends upon the power possessed by carbonic acid, in the presence of moisture, of decomposing the silicates of the alkalis. The soluble carbonate of soda or of potash thus formed is then quickly washed out from the surface of the glass. There remains, in the form either of iridescent scales or of an opaque pearly crust, a layer consisting not perhaps of pure silica, but of an acid silicate of lime, alumina, or lead as the case may be. Now a piece of clear glass may appear to the eye to be devoid of internal structure. But the ‘metal’ has, we know, in every case been subjected during the manufacture to a complicated series of involutions and doublings, to say nothing of the subsequent inflation if the glass has been subjected to a blowing process. When decay sets in—something similar may at times be seen in the case of a piece of wrought iron—this complicated formation is in part revealed, for it is evident that upon it the lines taken by the decay are in a measure dependent. On blown glass especially, the disintegration of the surface tends to result in a scaly formation resembling that of the shell of an oyster. As a result of the decomposition of light in its passage through these fine superficial films, and of the partial reflection from the back of the scales at various depths, we get those unsurpassed iridescent effects that we associate above all with the glass of the Romans. That these brilliant hues are dependent entirely upon the physical structure is well shown by the total disappearance of the colours when the surface of a piece of iridescent glass is moistened, as well as by their reappearance when the glass is again dried.