194 FKAGMENTS OF SCIENCE. 



sand was connected with a reservoir of compressed air, 

 the pressure being variable at pleasure. The hopper 

 ended in a long slit, from which the sand was blown. 

 A plate of glass was placed beneath this slit, and caused 

 to pass slowly under it ; it came out perfectly depolished, 

 with a bright opalescent glimmer, such as could only be 

 produced by the most careful grinding. Every little 

 particle of sand urged against the glass, having all its 

 energy concentrated on the point of impact, formed 

 there a little pit, the depolished surface consisting of 

 innumerable hollows of this description. 



But this was not all. By protecting certain portions 

 of the surface, and exposing others, figures and tracery 

 of any required form could be etched upon the glass. 

 The figures of open iron-work could be thus copied; 

 while wire-gauze placed over the glass produced a reti- 

 culated pattern. But it required no such resisting 

 substance as iron to shelter the glass. The patterns of 

 the finest lace could be thus reproduced ; the delicate 

 filaments of the lace itself offering a sufficient protection. 

 All these effects have been obtained with a simple 

 model of the sand-blast devised by my assistant. A 

 fraction of a minute suffices to etch upon glass a rich 

 and beautiful lace pattern. Any yielding substance 

 may be employed to protect the glass. By diffusing 

 the shock of the particle, such substances practically 

 destroy the local erosive power. The hand can bear, 

 without inconvenience, a sand-shower which would 

 pulverise glass. Etchings executed on glass with suit- 



Jr o o o 



able kinds of ink are accurately worked out by the sand- 

 blast. In fact, within certain limits, the harder the 

 surface, the greater is the concentration of the shock, 

 and the more effectual is the erosion. It is not neces- 

 sary that the sand should be the harder substance of 

 the two ; corundum, for example, is much harder than 



