CORROSION OF METAL ANTIQUITIES — GETTENS 561 



Copper is now only a minor constituent. This type of patina forma- 

 tion seems to occur principally on bronzes buried in sandy soils. Here, 

 both carbonic and humic acids play an important role in dissolving 

 out the copi^er of the alloy, leaving the tin constituent as stannic acid. 

 In other occurrences the tin oxide is powdery and friable. The 

 hydrous tin oxide patina alteration product is similar to cassiterite but 

 the lines are broader and more diffuse, which suggests that it is 

 cryptocrystalline and more finely divided than natural cassiterite. 



It has recently developed that there is a close resemblance between 

 the diffuse X-ray diffraction pattern of tin oxide alteration product 

 on bronzes and that of the yellow tin mineral varlamoifite discovered 

 in recent j-ears in the Maniema region of the Belgian Congo by the 

 mineralogist N. Varlamoff. It was named after the discoverer by 

 S. Gastellier (1950) who made the first analyses. These show that 

 the material is not a definite mineral species but a complex mixture 

 containing: Metastannic acid, HaSnOs, average 59.22 percent; SnOa 

 25.55; FesOs 9.45; SiOa 1.68; and H2O 2.12; total 98.02. Tlie yellow 

 color is caused by the ferric oxide. Another occurrence of varlamof- 

 fite from Cornwall, England, was subsequently described by Russel 

 and Vincent (1950-52), who showed that the X-ray powder patterns 

 of varlamoffite and of hydrated stannic oxide (metastannic acid) pre- 

 pared in the laboratory are identical in spacing and intensity with 

 cassiterite, but that they are broader and less well defined. In tests 

 made in the Freer Gallery Laboratory, the diffuse cassiterite patterns 

 given by several specimens of stannic oxide from ancient Chinese 

 bronzes were found to match perfectly the diffuse lines and spacings 

 given by a specimen of varlamoffite from Cornwall, England (speci- 

 men USNM E8886). A specimen of varlamoffite from the Belgian 

 Congo (USNM 115558), however, gave lines only slightly less sharp 

 than cassiterite. Although the investigators of varlamoffite mentioned 

 above speak of hydrated stannic oxide or metastannic acid, HsSnOs, 

 Weisser and Milligan (1932) many years ago showed conclusively 

 by thermal differential analysis and by X-ray diffraction analysis 

 that no true hydrates of stannic oxide exist. They maintain that the 

 diffuseness of lines of the cassiterite pattern is caused by tlie small 

 particle size of the stannic oxide crystals and that any water involved 

 is adsorbed water. 



F. Lihl (1962) as well as Plenderleith and Organ (1953) have 

 reported evidence of stannous oxide as well as stannic oxide among 

 corrosion crusts on tin objects, but there is no mineral of stannous 

 oxide listed in the current edition of Dana's System of Mineralogy. 



Much must be done before the mechanism of the solution of copper 

 and its replacement by tin oxide on bronze surfaces is completely 

 understood. 



