of the Tessera in a Roman Pavement. 121 



surprised at seeing these two colours intermixed apparently in a 

 most inharmonious manner ; and as the verdigris-green was so 

 different from any other colour I had met with, it suddenly 

 struck me that it was a mere coating to the tesserae, resulting 

 perhaps from chemical decomposition, and on scraping the sur- 

 face with a knife I was gratified to find that the verdigris only 

 covered up a glass of an exceedingly rich ruby tint. I then ob- 

 tained a small portion for a chemical analysis, which was kindly 

 undertaken for me by Dr. Voelcker, the College Professor of 

 Chemistry, the results of which are so interesting that I must 

 beg to lay it before the Club in his own words. 



Examination of red-coloured Roman glass {Cirencester). 



" The red glass which had undergone a partial decomposition 

 was coated with a white crust, which itself was covered with a 

 green substance. The latter on examination proved to be car- 

 bonate of copper ; the white crust dissolved with effervescence 

 in nitric acid, leaving gelatinous silicic acid behind, and was 

 found to consist principally of carbonate of lead and silica. The 

 glass, after having been treated with nitric acid and thus been 

 deprived of the white and green coatings, exhibited a bright red 

 colour ; it was now transparent, not very hard, and easily flexible 

 when exposed to a moderate heat. On analysis the following 

 substances were detected, as — 



Oxide of lead. Oxide of iron. 



Protoxide of copper. Lime. 



Alumina. Silica. 



" The red colour of the glass undoubtedly is due to protoxide 

 of copper, which was present probably in combination with alu- 

 mina in considerable quantities. It is well known that the an- 

 cients were acquainted with the art of colouring glass red by 

 means of copper, for Cooper informs us (Annales de Chimie, 

 serie 1. torn, lxxiii. p. 20) that he detected in an antique red 

 glass protoxide of copper, and Klaproth likewise ascribes the red 

 colour of an antique glass to the presence of copper, which he con- 

 siders to be contained in the glass in the state of protoxide. This 

 gentleman found exactly the same constituent parts as those 

 found by me in the Cirencester glass ; it has further been ascer- 

 tained that all the red glass in antique mosaic church windows is 

 coloured red by copper. Gold, which likewise imparts a beau- 

 tiful red colour to glass, is never met with in Roman glass, and 

 it appears that the property of gold and its combinations was 

 unknown to the Romans, for we do not find any traces with the 

 ancients which could justify the supposition of their being ac- 

 quainted with the art of making the purple and rose-coloured or 



