BISULPHIDE OF CARBON PRISM. 488 



should be scraped from some pieces of looking-glass (preserve the 

 substance scraped off for electrical experiments). The plates 

 need not be more than about 1*5 or 2 mm thick, but it does no harm 

 if they are thicker, except that very thick glass is difficult to cut 

 with pastille ; in any case, however, it is best to have the plates cut 

 square by a glazier and simply to round off the edges upon a grind- 

 stone. The width of the plates should be about l cm greater than 

 the diameter of the cylinder, and the length at least l cm greater than 

 the longest diameter of the elliptical opening which they are to 

 close ; but, as the length needed to make their edges exactly meet 

 at p depends on the width of the narrowest part of the ground 

 cylinder, it is best first to cut pieces of cardboard ta the proper 

 shape, and fit them so that their edges may touch one another 

 properly, as in fig. 277, jB, and then have the plates cut to the same 

 size. 



The well-cleaned plates are placed one on the other and joined 

 along one of their short edges by a strip of thin paper glued upon 

 both ; this forms a hinge convenient for giving to the plates their 

 proper position, and may afterwards be removed again. When 

 this band of paper has become pretty stiff, one plate is raised by- 

 taking hold of the other short side, the second plate remaining on 

 the table, and the cylindrical piece is inserted between them ; the 

 plates remaining in contact with the ground edges of the cylinder, 

 the proper position is found, and holding the cylinder with one 

 hand, the upper plate is turned back, the edge of the cylinder 

 covered with glue, and the plate pressed rather firmly but cau- 

 tiously upon it, taking care not to displace the cylinder. The glue 

 should not be too thin, so that it may become firm as soon as 

 possible ; a very thin layer should be laid on the edge with a fine 

 camel's-hair brush ; care must be taken that nothing is smeared 

 inside the cylinder, nor much squeezed into the inside when the 

 plate is pressed upon the edge. When the glue has become firm, 

 after a few hours, the whole is cautiously raised, turned over so 

 that the plate that has been fixed now rests on the table, and the 

 other plate is fixed in the same manner. The thin layer of glue 

 generally contracts in drying, and leaves chinks between the edge 

 and the plate. The joint should therefore be made perfect by the 

 following cement, which never becomes quite dry and is im- 

 penetrable for the carbon disulphide. Break 10 grammes of glue 

 into small pieces, and soften them for a few hours in cold water ; 

 pour the water off, and add 10 gr. of common brown treacle ; heat 

 the whole cautiously in a small tin can, stirring continually until 

 the glue is dissolved and both substances are thoroughly mixed. 



i i 2 



