Printing in the Medieval Islamic Underworld           19

size ofthe script. Given the unlikelihood of a sophisticated etching
press being known in the tenth century, the letters on the print-
block must have been in relief. Hence, a woodcutter would have to
have been a consummate master of his craft to have cut around
 

■'Hmmmr
 

^^f■
 


 

ejt
 

The upper text, apparendy printed from a tin plate, begins with the
first Sura of the Koran.

them so precisely with so few errors. The circles ending the
Koranic passages would have been a particular challenge, but even
under high magnification one cannot discern in them any straight
lines made by a woodcutter's knife. Indeed, the illegibility of the
black-on-white letters in the teardrop design at the top of the
shorter strip, which are so much clumsier than the much smaller
letters on the longer strip, strongly supports the proposition that
the former was printed from a woodblock and the latter from a
metal plate. Specimens in other collections confirm these indica¬
tions that the lower part ofthe Columbia amulet was printed from
a metal plate, as suggested by Safi ad-Din's verse. The likely proce¬
dure would have been for the amulet-maker to inscribe a text with
a stylus in a clay tablet and then harden the tablet by baking.
Molten tin would have been poured on this mold to produce a