OPTICAL PHENOMENA OF ANCIENT DECOMPOSED GLASS. 197 



of the glass begins, or is ready to decompose, the spherical film then forming will 

 extend itself over the glass, and form part of the general film. If the decompo- 

 sition has begun at different times round different centres, and proceeded with 

 different velocities, the spherical film in every cavity, whether deep or shallow, 

 will extend itself over the general surface. If this is a correct account of the pro- 

 cess, then we ought to find in every cavity a central portion having no connection 

 with the general film, because formed previous to the decomposition of the general 

 surface. When we remove the outer crust from the decomposed glass, we find 

 this to be the case. The central portions are removed with the outer crust ; and in 

 some cases I have found the cavities filled with that portion, in consequence of no 

 decomposition having taken place on the general surface. 



It is difficult to comprehend the nature of the molecular forces under which 

 these singular decompositions assume peculiar forms. There is no difficulty in 

 understanding how, in a homogeneous surface of glass, decomposition round a 

 centre should excavate a perfectly spherical cavity, and how its ultimate form 

 should be changed into that of irregular polygons, by the interference of decom- 

 positions advancing from surrounding centres ; but it is a singular circumstance 

 that the glass should be in any one place in such a state of molecular incohesion, 

 that the decomposition should follow a law which produces a perfectly ellipsoidal 

 cavity. Nor does it seem less remarkable that after a cavity, either spherical or 

 ellipsoidal, or irregularly polygonal, has been formed, spherical cavities should be 

 formed on different parts of the films which compose them. 



The preceding observations have been made on films of glass from Nineveh 

 and Rome, after the opaque or outer films had been removed. Each film, there- 

 fore, has on one side all its cavities concave, and on the other side all of them 

 convex, like a bundle of watch-glasses. Each film consists of many films, which 

 can be separated by the thin blade of a knife or lancet, adjacent films differing 

 from one another only in their thickness and colour. The films, though adhering 

 with some force, are not in optical contact, as they freely admit water, and other 

 fluids between them, and display the fringes of thin plates produced by the dif- 

 ferent thicknesses of the interposed films of air. 



The mode in which the decomposition proceeds round different centres is 

 shown in Plate X. fig. 1, taken from a specimen of glass from Nineveh. The 

 same structure is less perfectly seen in fig. 2, taken from glass found also at 

 Nineveh. When the glass has lain on damp earth, the decomposition has taken place 

 very irregularly round separate centres. It consists of two thick crusts of opaque 

 vitreous matter, enclosing a plate more or less thick of pure glass, with deep ex- 

 cavated cavities on both of its surfaces. A specimen of this kind breaks by the 

 pressure of the finger like the slice of an apple. In a finely-shaped glass bottle 

 found near Cheltenham, and presented to me by Mr Sydney Dobell, a thick grey 

 crust of iridescent films, of uniform thickness, covers the whole of its surface, and 



VOL. XXIII. PART II. 3 I 



