It might be of interest to note that the blisters on the
obverse and those on the reverses are in different places. The reason is that the pressure did not originate centrally in the
flan (that would also have caused larger bumps, affecting the coin on both sides), but that whatever expanded was situated just underneath the non-crystallized surface layer. Don't ask me why that layer is there, in the majority of perhaps even all silver "coins" from the late
antiquity (and before and after as well, and certainly not only in
Roman coins).
In the section figured above, the metallurgist (a professional metallurgist) noted the scattered white areas between the crystals, but could not determine their composition. But one thing is certain, they are the result of migration of molecules, resulting in separation of silver and these substances. Whatever they are (AgCl is one possibility), they show that small molecules (and maybe large) migrate in the metal. Accordingly, both oxygen and water can slowly move through the metal. That must be taken into account when we discuss the possibility of impurities and how they can expand, either over time, i.e. corrosion, or instantly, as when strongly heated and they gasify.
I know of at least one ongoing study dealing with the surface layer of late antique coins. Perhaps some new insights will be gained.
Meanwhile, the images attached tell a story (they are a small sample of numerous similar instances). The little silver piece shows areas where the surface silver has flaked off, revealing the
crystallized core material. Next, the
face is Constantine's, from the famous
Munich medallion (my photo). Note how the surface layer has partly flaked off, revealing the underlying core material (the
crystallization is not clear due to long time wear and polishing). Finally, another detail of the broken
Constantius II siliqua, courtesy "timka".
The blisters can be explained by the original inclusion or subsequent formation of small specks of impurities under the surface layer, expanding either through corrosion (my preferred hypothesis) or through fire exposure. The idea of trapped air is difficult to reconcile with the evidence of migration resulting in concentration or diffusion of molecules through the silver. Once
crystallization had occurred, any trapped air (and thus under pressure) would simply spread out evenly or evaporate.