1 87 1.] Chemical Science. 403 



of specimens of antique glass and metal-work. The silver specimen with 

 which the present note is concerned cannot have an antiquity of less than 

 1500 years, but the date of its manufacture may prove to have been even more 

 remote. It was shaped like a fibula. It resembled a crescent in that the ends 

 of the bow were small and pointed and its middle much stouter; sections 

 throughout were nearly cylindrical, and in the widest part about |-inch in 

 diameter. The specimen was nearly covered with a dark sub-metallic crust, 

 and underneath this a soft powdery grey substance was observable. Both 

 these layers, which together presented an average thickness of i-30th of 

 an inch, could be readily removed by the finger-nail. Below, the substance 

 appeared white, metallic, and uniform, but yet was remarkable for its extreme 

 brittleness, the object being easily snapped in numerous pieces by a very 

 slight pressure. If the brittleness had pertained to the outer crust or layer, 

 obviously altered as it was, it would not have been surprising. But to find 

 that an apparently continuous metallic substance of great lustre, and, cer- 

 tainly once, at the time of its manufacture, in a perfectly ductile condition (for 

 portions of it had been twisted into fillets), to find this metallic substance as 

 brittle as a stick of chalk was certainly difficult to explain. And a further 

 difficulty was offered by the observation that in the thicker part of the cres- 

 cent, and only there, a kind of core existed similar in colour to the main mass 

 of the object, but possessing great tenacity and ductility. This core could not 

 be broken short off, but might be wrenched out of the surrounding brittle mass, 

 though with some difficulty. It tapered off towards the narrower ends of the 

 crescent, and terminated where those parts did not exceed £th of an inch in 

 thickness. The outer crust consisted of the corroded and chemically altered silver- 

 alloy. It was made up of finely-divided metallic silver, with traces of the 

 sulphide and chloride of that metal, and it gave also a faint trace of iodide ; 

 copper, apparently in the form of a basic carbonate, was likewise present, with 

 a trace of gold. The main part of the specimen consisted of the second and 

 extremely brittle layer referred to before. On analysis it was found to have the 

 composition: — Silver, 94*69 per cent; gold, 0*41; copper, 3*43; lead, 0*28; 

 antimony, with trace of arsenic and bismuth, 1*21. The central ductile 

 and malleable core had precisely the same composition as this brittle 

 layer. To what, then, are the physical differences between them to 

 be attributed ? Not to chemical differences or chemical changes, but to a 

 molecular change which, in the course of ages, has occurred throughout 

 all the thinner parts of the piece of silver, but had left unchanged in the 

 thickest part a central fusiform core. A crypto-crystalline structure had been 

 produced in the previously homogeneous alloy, which caused the peculiarly 



Fig. 8. 

 ^*»" '-" -^ ^'a Core of unaltered silver. 



Crystalline brittle silver. 



~C Outer altered crust. 



Section (in the plane of the paper) of the silver object referred to. 



easy fracture of the metal. A smart blow with a hammer was capable of 

 shattering it to powder, while by rolling or gentle hammering, the brittle mass 

 could be restored readily to its pristine condition. The density of the brittle 

 silver was 9-06, but, by rolling, it became io"20. There was not, therefore, a 

 chemical change, but a physical one. However produced, whether by minute 



