138 Dr. Cutbush on the Chinese Fire, &c. 
composition by charcoal, &c. were not investigated; and 
infers, as the only certain mention of salt-petre is to be 
found in the manuscript containing directions for props 
gun powder, that our salt-petre was not known to the an- 
cients. The knowledge of gun powder, however, Pain Fe 
a new era in pyrotechny. 
The fire works of the ancients consisted for the principal 
part of illuminations, and the use of some particular compo- 
sitions, in which certain oils, as naptha, were employed. 
Alexander the Great, witnessed some experiments with 
Senn at Ecbatana. 
their sa and ceremonies. Of their rites, whether sali 
gious or segiedatin they appear to have made use of fire 
under some particular form. 
The representation of figures in ithe wil age by the 
Chinese, was common ory I. 
entered a ag was an sadanen of that: kind, in 
honor of his en 
e ancients, a beeen, had two descriptions of fire 
works; one of which they set off by hand, and threw amovg 
the people, and some were merely illuminations. . Of this 
description were cardons, stars, and fire balls. A writer of 
antiquity mananeTER in speaking of these exhibitions, that 
“he has seen a great many of these artificial machines, but 
to speak vt truth, few which have succeeded, and it is com- 
monly after acclamations of joy, the spectacle is finished by 
the destruction of some, and the wounding of a great number.” 
The other description of fire works was calculated for 
theatrical exhibitions, which consisted of illuminations, 
transparencies, various figures of man and beast clothed 
with fire, &c. The art of wattage figures in fire, seems 
tohave been the most pertect. e figure according to 
modern art, is first covered with clay, or — scr 
to prevent the action of the fire, and then deco with a 
multitude of small cases charged with saslley: le 
that produce, if so required, variously coloured fires. These 
cases are so connected by means of quick match, as to com- 
their fire to each other, in suecession or otherwise: 
