92 THE GLASS 



save the rude and problematic "Mithras" scratches (page 29) have 

 appeared in the camp. 



It is worth while also to record the statement of Professor William 

 Ridgeway, the author of "The Early Age of Greece," "The Origin 

 of the Thoroughbred Horse" etc., who visited the camp in 1905, 

 that he couU recall no other extant model of an ephippion. — Ed.] 



Glass. 



Unlike the pottery, the glass at Melandra is well pre- 

 served. It therefore lacks the iridescent beauty of decay- 

 ing glass and retains the colours given to it in the process 

 of manufacture. These colours are either various shades 

 from brown to yellow or pale translucent greens and blues. 

 In one case a deep, almost opaque, blue is used. Like most 

 Roman glass the fragments from Melandra contain 

 numerous small air-bubbles, flaws which cannot be 

 avoided in the use of small furnaces such as those found 

 at Warrington,^ where it is likely that much of the local 

 glass was made. 



The different forms of glass found at Melandra are : — 



(1) Window glass. This was evidently cast by pouring 

 the molten material on a flat stone, for the under side of 

 the sheet of glass reflects the roughness of the stone, while 

 the upper side has a smooth and somewhat wavy surface 

 and a naturally bevelled edge. 



(2) Small button-shaped discs of glass. These too are 

 made by pouring a small quantity of molten glass on a 

 flat stone so that the lower side is flat and slightly 

 roughened, whilst the upper side is rounded and smooth. 

 Most of the discs are of either black or white opaque glass, 

 but there is one example of clear green glass. The discs 

 may have been used as counters in some game, or else for 

 ornament (as they are used on mule harness in Greece at 

 the present day). 



9. Cf. Warrington's Roman Remains by T. May, p. 37 seq. 



